Trust and Institutions Laurence James University of Alabama at Birmingham

1. Introduction Trust is a vitally important facet of person-to-person relationships. Based on the onslaught of advertising, it is natural to think that trust is a vitally important part of institution-to-person relationships. Nevertheless, there is a tension behind the idea that trust is a vital part of the institution-to-person relationship. This tension is evidenced by what seems to be an overwhelming emphasis on the difficulties of such a relationship.1 This tension implies the following thesis about trust: Erosion thesis:

The nature of institutions is such as to erode/diminish/dilute the trust relationship between individual and institution.

The erosion thesis is the predominant thesis about the possibility of institutional trust. Here, I will argue that though the erosion thesis is on the right track in that it sees there to be tension between institutions and trust, it does not go far enough. I will argue for a stronger thesis: No-Trust thesis:

The nature of institutions is such that it is a conceptual impossibility to have non-derivative trust in them.

My argument will go as follows. I will begin by looking at the moral psychology (and moral phenomenology) of trust. To this end, I will follow Annette Baier’s spectacular Tanner Lecture, Trust (and I will supplement her 1

See, for example, Susan Dorr Goold, “Trust and the Ethics of Health Care Institutions”; Michael Jellinek, “Erosion of Patient Trust in Large Medical Centers”; and Steinar Askvik and Nelleke Bak, Trust in Public Institutions in South Africa.

Trust and Institutions

65

account where necessary). I will illustrate a complex of phenomena that are constitutive of trust. From there, I will argue that anything that is to count as trust (as opposed to some other closely related phenomenon) must at least have the potential to exhibit these phenomena. Since institutions do not even have the potential to do so, it is impossible to enter into a non-derivative trusting relationship with them. I will consider the possibility of a distinct type of trust— institutional trust—that has been suggested and, following an argument of Paul Benacerraf’s (from “Mathematical Truth”), I will give an argument to show that institutional trust, at least in the form conceptualized, ought not be considered a kind of trust. I will conclude by applying these theoretical musings to more practical cases and suggest that the present insight—that trust in institutions is impossible—has important consequences for how we understand the role of the institution and for how we ought to act toward institutions.

2. Trust Annette Baier remarks that “trust is a notoriously vulnerable good, easily wounded and not at all easily healed.”2 Despite the vulnerability one incurs by trusting, trust is a seemingly inevitable facet of human interaction. Most typical cases of trust involve two or more individuals (in a “web” as Baier calls it) that are related through family, friendships, or some other common strand. Other cases of trust relationships may involve individuals in different positions, for example, a potential house buyer and a real estate agent. What is common to all of these cases is the personal acquaintance amongst the trustees. Trust is a varied phenomenon, however, and not all cases proceed in the manner that I have mentioned. Often, we find ourselves trusting by word of mouth: a trusted friend may recommend a mechanic or baby sitter. Other times we find ourselves trusting on presumption: we go on a date with someone we just recently met, presuming them to be trustworthy. And yet, other times, we have no choice but to trust: we get stranded on the side of the road on an icy evening and need a stranger to give us a lift to the nearest gas station.3 It seems, though, that there is another group of putative trusting relationships that cannot be easily classified with the abovementioned cases. The cases I have in mind are those between an individual and an institution. The 2

Baier, “Trust,” p. 110. Here the trust may be either (1) that we trust the stranger to take us where we ask, and where he says he’ll take us, or (2) that we trust the stranger not to harm us. Or, we may trust the stranger on both counts. But, whichever one it is, we have no choice but to trust since the alternative is hypothermia. 3

66

Laurence James

defining characteristic of these cases of trust is the replacement of one of the individuals with an institution, and, accordingly, the lack of personal acquaintance. “Institution” is a diverse concept, and I want to say a quick word about the type of institution that will be the focus of this paper. Here, I am going to focus on the faceless institution and I am going to pick out such an institution by example.4 I want to rule out all those institutions that allow for their representatives to become fairly well acquainted with “customers”. There is a simple reason why we want to exclude such institutions: in order to determine if it is possible to trust an institution, it is a methodological requirement that we remove any confounds that might make such trust derivative. If I trust my bank because I trust my local banker, then the institutional trust is secondary to the individual trust and derivative on it. Thus, by eliminating the possibility of person-to-person trusting relationships (where one of the individuals is a representative of the institution), the conceptual path is cleared, and we can see if it is really possible to trust an institution qua faceless entity. The internet affords us excellent examples where, what is being asked of potential customers, is to trust faceless entities. eBanks, for example, allow an individual to set up a checking account, deposit and withdraw money, take out a loan, in short, do all the things regular banks allow their customers to do. But, and this is the point, the eCustomer never gets any face-to-face meetings with bank personnel. The question that I want to explore is this: can an individual genuinely trust an institution when there is no possibility of that individual forming a trust relationship with a representative of the institution? That is, could there be non-derivative trust of faceless institutions?5 In order to determine if it is possible to trust a faceless institution, we need, of course, to return to our investigation of trust. Trust, for Baier, is a feeling.6 While it might be possible to understand the concept using linguistic analysis, I believe that a phenomenological investigation into the nature of trust will be more fruitful. One must attempt to divine those phenomena that are typically coextensive with and constitutive of trust so as to develop an understanding of what it is. These phenomena can be divided into two rough classes, associated with answers to the following two questions: (1) Why do we want others to trust 4

There is no need to be overly precise about what counts as an institution here. All I need is something that we can all agree is an institution and is faceless. 5 A further issue, which I will not address in the present paper, is whether the results of the present investigation will apply to those institutions where there exists the possibility of forming a trusting relationship with a representative of the institution. 6 Baier, “Trust,” p. 111, “Trust and distrust are feelings…”.

Trust and Institutions

67

us?, and (2) Why do we need to trust others? What distinguishes these two questions with respect to trust is that in the case of (1) it is us, the trusted, who are empowered, whereas in the case of (2), it is we, as the trustees, who are in a position of vulnerability. As the trusted, we are in a position to come through for those that trust us, to fulfill their expectations about how we shall act, but, conversely, we are in a position to let them down and attack them where they are vulnerable. As trustees, we are in a position to have others help us where we cannot so easily help ourselves, and in a position to have others choose for us where we may not have the foresight to choose. But, we are also in a position to be betrayed and let down. These remarks oversimplify. Trustees have some power too. They have the power to affirm. When someone is willing to trust, they affirm by letting the trusted know that they are taken to be dependable, trustworthy, and honorable. To be welcomed as a trusted individual is oftentimes a momentous step in a relationship; a step that can provide a foundation for building a stronger, more intimate relationship. Thus, trusting can be particularly powerful. The trusted, in a like manner, have a weakness. Letting down one who trusts can result in dire consequences, as in exclusion from the relationship, or one’s life. It can result in loss of reputation. Accordingly, the choice to let down one who trusts does not come risk-free. Nonetheless, a sort of intuitive cost-benefit analysis reveals that overall the trusted is in a better position with respect to empowerment than is the trustee. Another phenomenon intimately related to trust is forgiveness. Forgiveness occurs when a trustee overcomes a harm perpetrated by one she trusts.7 Depending on the specific violation of trust, forgiving may be more or less easy, trivial or monumental. Because trust and forgiveness are so intimately linked, we might be able to find significant relations between the purported presence of one and the conspicuous absence of the other. Imagine an individual who claims to trust her dog. Imagine that she, after claiming this, is mauled. If then at a later stage she claims to have forgiven her dog for the mauling, it would not be amiss for us to point out that dogs simply can’t be forgiven. Forgiveness, as it were, cannot take a dog as an object. This is because dogs are not members of the 7

There is a significant literature on forgiveness. For the present purposes, I will go along with the account of forgiveness offered by Jeffrie Murphy in “Forgiveness,” and Laurence Thomas in “Forgiving the Unforgivable”. Briefly, to forgive is to foreswear one’s resentment. For both authors, not all manners of foreswearing resentment are appropriate (forgetting is not, for one). Moreover, the foreswearing of resentment that is implicated when trust is violated is the type that restores the perpetrator to an equal moral footing (Thomas appropriately calls this restoration forgiveness).

68

Laurence James

moral community.8 However, if a dog cannot really be forgiven, shouldn’t we reconsider whether one could really be trusted? The more general point is this: some things can be forgiven, others not, but only those things that can be forgiven can be trusted, and vice-versa. Because I cannot forgive my watch if its battery dies, I am in no position to trust it either. And, when I dote, as a manner of speaking, on my watch and call it the “most trustworthy ever,” this should be recognized as metaphorical, for what is meant is that the watch is reliable.9 We will return to this idea when we discuss institutions and trust. Another phenomenon of trust is, as Baier notes, that “real trustworthiness, like real trust, involves feelings, beliefs, and intentions, which sometimes can be faked.”10 One might feign trust only because not being trustworthy or not overtly trusting would cause a great loss. And, as Baier emphasizes, such faking of trust does not constitute genuine trust. Her point can be made more generally in the following manner. Though trust is a feeling, it must be felt or acted upon for the right reasons.11 The threat of a bad reputation, then, cannot serve as an appropriate motivation to be trustworthy. “Trust, or else!”, as someone might coercively say, belies trust itself and makes trust on the basis of that imperative impossible. All imperatives to trust are doomed to fail.

8

This is not an entirely obvious point. Some are inclined to make the distinction between moral patients (something to which we have moral duties) and moral agents (something which we can hold morally responsible), and they would argue that while dogs are not moral agents, they are moral patients. But, being a moral patient is not sufficient for being forgivable since we only forgive someone who we hold as responsible for his actions (and only moral agents are held responsible for their actions). To hold a dog morally responsible for his actions is a case of incorrectly attributing moral status to him through personifying him. Thanks to Maureen Kelley for suggesting this. 9 A more subtle point emerges here: people are more likely to use the words “trust” and “trustworthy” metaphorically than they are the words “forgive” and “forgiveness.” By testing to see whether something is forgivable, we are able to find those metaphorical uses of the words “trust” and “trustworthy” and banish them from the conceptual terrain wherein lies trust. In some cases where there seems to be parity between the uses of trust and forgiveness, we must rely on our intuitions to test whether these uses are literal or metaphorical. As in the case with someone claiming to trust their dog, we see that that individual has probably personified their pet, attributing to it propositional attitudes that it almost certainly lacks. 10 Baier, “Trust,” p. 112. 11 Baier clearly subscribes to a cognitivist view of emotions.

Trust and Institutions

69

There is another point about trust that the possibility of faking it reveals. This point can be made clear when we ask why one would want to fake trust or trustworthiness in the first place. As with any fakery, the faking is done so as to mislead one into believing that we believe or feel a certain way. Fakery can only be effective when there is some such person to be duped. Faking a feeling of admiration towards my car is impossible and is to be distinguished from faking a feeling of admiration of my car to myself. When the deceit in question is other oriented, it cannot be genuine if the other is not of the appropriate type. Cars aren’t of the appropriate type and what we will need to see is whether institutions are of the appropriate type. Yet another phenomenon of trust concerns flexibility. One type of flexibility concerns what the trusted can be reasonably expected to do in a given situation. A stark example: Suppose that Jack the vegetarian is kidnapped by a group of carnivores, and Jill the vegetarian wants to save Jack. If the only way Jill can save Jack is by getting admitted to the Carnivore Club, and admittance involves eating meat, Jill may choose to eat meat even though her standing principle is to not do so. Being a trusted individual involves being flexible and, when it is called for, even going against one’s principles. Another type of flexibility is tied into forgiveness. If one is completely unwilling to forgive when it comes to a violation of trust, then that may provide a reason not to trust one.12 If Jack and Jill are in a relationship of mutual trust, and Jill does something to violate Jack’s trust of her, and Jack chooses not to forgive, depending on the violation of trust in question, an unwillingness to forgive may indicate that Jack does not have Jill’s best interests at hand. Part of what it is to trust is “to give discretionary powers to the trusted, to let the trusted decide how, on a given matter, one’s welfare is best advanced, to delay the accounting for a while, to be willing to wait to see how the trusted has advanced one’s welfare.”13 If Jack has a blanket policy of not forgiving violations of trust, then it is not possible for Jack to really have Jill’s best interests at hand. Sometimes we violate the trust others have in us for good reasons. I may not attend a friend’s wedding (even though he trusts me to) because I am needed by another friend to help them through a tough time. If my friend were to not forgive me for this, he would not have my best interests at hand.14 Ex hypothesi, blanket policies are incompatible with the flexibility needed to pursue another’s best interests.

12

See Baier, “Trust,” p. 116. Baier, “Trust,” p. 117. 14 This example originates from a seminar on institutional responsibility held at Syracuse University. 13

70

Laurence James

The last phenomenon that I wish to touch on with respect to trust is reliability. An easy first approximation of what trust is, though upon further reflection a conflation, is to liken it to reliability, viz., the trusted person is the reliable person. Baier gives the following example to suggest that trust is not merely reliability, “‘I can trust him to remember my birthday: he has given his bank a standing order to send the same flowers each year on that date. Short of bank collapse, I can count on it,’ would be to speak at least ironically, if not sourly.”15 She goes on to make the further point that the degeneration of a trusting relationship into a mutually predictable one is one of the greatest travesties that could befall such a relationship. I want to push this point further and suggest that reliability proper is incompatible with trust proper, that the relation between the two is not simply one of degeneration.16 This is at first brought out by the fact that sometimes reliability is desired over trust. Consider the following example: Boeing creates an automated pilot that is essentially infallible. The pilot has the ability to adjust flight path for weather disturbances, to ensure that turbulence is almost non-existent for the duration of the flight, to adjust for mechanical problems, and so on. Over a million test runs, the automated pilot does not record one glitch. It is as reliable as can be hoped for. Our regular pilots are not this reliable. They crash occasionally, fly drunk now and then, are unable to think fast enough to avoid turbulence, unable to react in just the right ways to mechanical difficulties even though, for the most part, they are pretty safe. Because human pilots can err in many ways, when we choose to fly with one, there is a presumption of trust that the pilot will do everything in her power to not err. On one hand then, we have the absolutely automated but reliable pilot, on the other, the trustworthy human pilot. Which would you choose to fly with? Almost everyone who accepts the reliability statistic would choose to go with the automated pilot. When it comes to the possibility of losing one’s life, a reliable guarantee that the automated pilot will not crash is, I suspect, far more comforting than the possibility that the human pilot may because she is not perfect. When the potential damage to oneself is inordinately large, faith is more easily placed in reliable safeguards than in trustworthy ones.17 There are other times, though, when reliability is simply not what we want. The man from Baier’s example who ensures that the bank will reliably send flowers on a friend’s birthday belies the trust that the friend has in him. Baier’s 15

Baier, “Trust,” p. 117. Or of erosion, diminishment, or dilution. That it is conceptually impossible to have trust proper in an institution is an important part of my argument. 17 Consider driving down a steep mountain in a bus versus going down that same mountain in an automated train. 16

Trust and Institutions

71

analysis of what goes wrong is that the relationship becomes one of “predictability” or, for our purposes, predictable reliability. Laurence Thomas has suggested that Baier does not have it quite right and that the problem lies not with predictability, but with the lack of agency engendered by having the bank reliably mail flowers each year.18 The flower sender no longer has to put thought into the birthday gift, no longer has to remember that it is his friend’s birthday, no longer has to compose a thoughtful, meaningful card. By passing on these responsibilities he shows a particular disinterest in his friend and not, as one might unreflectively suppose, an interest in his friend’s wellbeing. Part of what makes receiving a gift a pleasurable occasion is the fact that someone took some time from their busy schedule to think about you, your likes, and your wellbeing. Turning a trust relationship merely into one of reliability is tantamount to building in a policy of inflexibility which, as we have already seen, is incompatible with genuine trust. There is more to be said on this issue. Affirmation, one of the surface phenomena of trust, comes about in two ways in a trust relationship. Firstly, when someone trusts us they affirm us by showing a willingness to be vulnerable to us, and by showing that they deem us responsible to take their interests into our hands. Secondly, when we trust someone and they come through in a given situation, we are affirmed by seeing that we are important to them and that they take time to act in our best interests. Insofar as affirmation is tied into trust in two ways, it can be lost in two ways. When we reliably perform for another without tailoring our actions more specifically to the circumstances they are in, we fail to affirm them. And vice-versa for when they act this way to us. The result of this brief discussion is not that trust is superior to reliability, but rather that each is suited to, and better for, different occasions.19 When discussing institutions, this should be kept in mind. It will be instructive to see which is preferred in an institution, and also, what institutions expect from us.

3. The Possibility of Institutional Trust In the foregoing discussion, I touched on five phenomena intertwined with trust. In order these are: empowerment, forgiveness, faking it, flexibility, and reliability. The purpose of looking at these phenomena was to give us a better 18

Laurence Thomas, personal communication. As a final reminder, the general point under consideration here is that though reliability is one phenomenon that makes up trust, it is not trust. Trust is conceptually much richer than reliability alone. 19

72

Laurence James

idea of what trust is. But, something else was suggested: a relationship in which the first four phenomena are absent may not qualify as a trust relationship. In this section, we will examine these five phenomena as they relate to the possibility of trust between an individual and a faceless institution.

3.1 Empowerment No one thinks twice about opening a bank account, except when it comes to considerations of convenience. Yet, by placing our money in the hands of a bank we make ourselves vulnerable. The bank could, hypothetically, freeze our funds, or disallow access to our accounts, or claim that we never opened an account with them in the first place. In short, placing money in bank accounts reduces the amount of economic self-determination that we possess. But, these seemingly important considerations never figure into our decision to open an account. Why is that? Are we stupid? Ignorant of the possible danger of bestowing this power over us on another entity? If someone we had just met assured us that they could store our surplus money, would anyone take that person up on the offer? It is doubtful. What is it, then, about opening a bank account that is different to trusting an individual? To be sure, there are some similarities. In opening a bank account and in trusting another, we empower those entities. And, in opening a bank account and trusting another, we make ourselves vulnerable. Both empowerment and vulnerability are phenomena that figure centrally into understanding what trust is, yet, there seems to be some difference between the empowerment and vulnerability that comes along with opening a bank account and the empowerment and vulnerability that comes along with trusting another. What are these differences? The main difference can be brought out by looking at the proportionality of empowerment to vulnerability in the different cases. In the personal case, the amount of empowerment we give to another is directly proportional to the amount of vulnerability we feel (e.g., trusting a friend to meet us for lunch vs. trusting a friend to not divulge some important secret about us). As empowerment increases, so too does vulnerability on, roughly, a one-to-one scale. In the institutional case, though the amount of empowerment may be quite high, the level of vulnerability stays relatively low. There are many reasons for this. Banks are not required to, nor do they care to, take a personal interest in their average customer. As such, they possess very little information that is of a personal nature. One does not bond with one’s bank. Banks are also required to have certain types of insurance in case of emergencies, and so there is already a guarantee that one’s money will be safe there (contrast this to the personal case

Trust and Institutions

73

where a possible friend comes with no insurance against malevolent behavior). Another interesting facet of person-bank interaction is that there is a veil of anonymity that shrouds the new customer. Banks may be privy to her name, social security number, address, etc., but they are never really privy to her. Anonymity is generally regarded as reducing vulnerability. People constantly frequent psychotherapists who really know nothing about them because such therapists lack knowledge. It is a fact about the human psyche that the more someone comes to know personal details about us, the more vulnerable we are. 20 With the banks not really being in the relevant know, the possibility of being vulnerable to them remains low. The difference in proportionality between the personal trust case and the institutional trust case suggests that we might not be dealing with similar cases; that what is going on with institutions is not, in fact, trust.21

3.2 Forgiveness A consequence of the preceding thoughts is that the issue of forgiveness does not seem to arise between individuals and institutions. I think this is right, but I also think it is important to see why. When it comes to forgiveness and institutions, the direction of forgiveness can go in two ways. An individual may forgive a bank for some travesty, or the bank may forgive an individual. The possibility that I want to explore is whether individuals can genuinely forgive banks. Suppose John is in some financial debt, and he fails to make his credit card payment. The bank promptly charges him a late fee. John calls customer service and pleads his case, but to no avail. The bank will not reverse its late penalty. John knows that this will weigh negatively on his credit record, thus harming him in the future. Even this important concern does nothing to change the bank’s decision. Would it be appropriate for John to forgive the bank based on the fact that they are just playing by the book? 20

And, as I already mentioned, the personal details in question are not social security numbers, but insecurities, say, one’s looks. 21 Laurence Thomas, personal communication, has pointed out that there is an interesting consequence of this line of thought, namely, “that most well-placed people in life have a measure of personal trust with members of institutions, thus making the trust of the institution itself derivative. Not so with poor people. At least not typically”. This is because institutions typically assign representatives to important (i.e., wealthy) customers. By personalizing how it (an institution) acts towards important customers, it increases the chance of keeping that customer.

74

Laurence James

In a trusting relationship, which, as Baier points out, is usually a mutual relationship, it is incumbent upon each of the parties to have the other’s best interests at hand. The bank, in penalizing John shows itself to not have his best interest at hand. A friend would surely let John pay the money back when he could afford it without putting himself in financial detriment. But, the bank does not have this type of interest in John. As Baier states, “It is not really so unaccountable that distrust should be directed not so much at those who once or twice let one down in the most obvious way, who manipulated one or gave one what turned out to be false assurances, but rather at those who prove angrily unforgiving of the letters-down, who do not forgive those who forgive others, who show themselves to be completely reliable punishers….”22 But, along with the distrust that Baier talks of, it is also natural that we should not forgive those who, in choosing to punish us, take no interest in our specific situation. This lack of interest stems from how the bank regards us: not as a partner in some cooperative enterprise, but as a financial commodity in the bank’s moneymaking scheme. The lack of mutual disregard in one another’s interests between individuals and banks is surely one reason why forgiveness is not properly bestowed on banks when they harm us. We may even forswear resentment against the bank for penalizing us because we come to realize that the bank is acting as it said it would from the very beginning. Its action is not directed at us per say, but anyone who commits the violation we commit. Nonetheless, it is this very disinterest that excludes the bank from the list of entities that may properly be forgiven. My remarks in 2 suggested that if something is not forgivable, then it is not trustworthy. Since banks are not forgivable, as I have just argued, they are not trustworthy.

3.3 Faking It No matter how one cuts it up, one cannot fake trustworthiness to a bank, nor can one fake a trusting feeling towards the bank. Fakery only succeeds when the other party is there to be deceived by the faking behavior. Banks, especially of the sort I have limited us to, are simply not in the position to pick up on such behavior. Perhaps this is only a minor point, but, it is evidence that the phenomenology of the “trusting” relationship between a bank and an individual is drastically stripped down in comparison to the phenomenology of personal trusting relationships. Moreover, if I am right in that trust is coextensive with other phenomena, the possibility of faking it being one of them, then the 22

Baier, “Trust,” p. 116.

Trust and Institutions

75

conspicuous absence of these other phenomena would suggest that what we are dealing with cannot be properly classified as trust.

3.4 Flexibility To be involved in a trusting relationship is, in part, to look out for the other’s best interests. This involves a certain amount of flexibility. Consider a couple, one of whom does not speak English fluently. Suppose he asks his partner to correct his English when he misspeaks. Compliance on the part of his partner suggests that his partner is keen on helping him improve his English, thus, his partner is keen on looking out for one of his interests. However, his partner has to know when it is appropriate to correct his English. An inflexible policy of correcting all mistakes will not do. If correcting a mistake involves diminishing one in the eyes of a conversant, then the mistake should not be corrected. A good partner has the wisdom to know when to correct and is flexible accordingly. A bank is a paradigm of an institution that does not exhibit such flexibility. Late payments receive penalties, as do non-payments, overdraws, bounced checks, and so on. Many such violations result in negative checks on one’s credit record. From the bank’s perspective, things should be this way. After all, a bank is a business, and, being too flexible is bad business—especially on the scale at which banks operate. We might say that banks adopt such blanket policies (i.e., such inflexible policies) because they are profit-oriented. However, in a personal trust relationship, this profit-orientation is conspicuously absent. Because these relationships, if optimal, are other-oriented, the way one acts is constrained by the effects it will have on the other. Moreover, since there is a vested interest in keeping the other in the relationship, one does things that provide no benefit to oneself. A good friend does not charge interest on a small loan to another though a good bank does. A good friend defers asking for or accepting payment because of an economic crisis though a good bank doesn’t. A good friend doesn’t issue a penalty because of a late payment though a good bank does. An especially good friend may even give money without ever expecting it back, nay, she may insist that she doesn’t ever want repayment. However, any bank, whether especially good or bad, would never give such a “free” loan. Good friends care, and this is why they show such flexibility to those who trust them. Good banks, on the other hand, don’t care. Each customer is a commodity to be dealt with in a manner that will maximize profit. When, in a bank commercial, the representative touts this or that bank as being the caring or trustworthy bank, we would do well to head Baier’s words, “Healthy trust rarely needs to declare

76

Laurence James

itself, and the mere occurrence of the injunction “Trust me!” or of the reminder “I am trusting you” are danger signals.”23 To be a good bank is to ignore, or, at least, refrain, from doing those things that make friends good. As such, banks are not really the types of entities that can be trusted because they are not the types of entities that look out for one’s best interests. This is best brought out by the blanket policies (the inflexible policies) that they adopt towards their clientele.

3.5 Reliability The point was made above that reliability and trust are to be distinguished, and that far from one being superior to the other, they are best suited for different purposes. As trustworthy as my friend might be when driving me around, it is always reassuring to know that if we should get into an accident, reliably, the seatbelt will lock, the airbag will deploy, and the crumple zone will serve its purpose. As much as John knows that Jane reliably keeps his personal details secret, it is always assuring to know that he can trust her to divulge one of these secrets when it would best serve his interests. There are two questions concerning reliability and trust that we may ask when it comes to banks. First, what are banks like: are they trustworthy entities or reliable entities? Second, what should banks be like: should they be trustworthy or reliable? It is evident that banks, as they are, are reliable. The whole structure that underlies them is in place to ensure reliable operation. If a bank should lose money, it is FDIC insured to guarantee that I will not lose money. When I have some amount in my checking account, the bank is set up so that that money will always be available to me, and at no point, shall I have difficulty withdrawing it. Should someone steal my bankcard and use it for purchases, the bank will investigate the loss and replace all the money when its investigations show fraudulent use of my card. The bank is also constrained in how it acts towards me. It must face big penalties if it defrauds me in any way. Being reliable in how it executes its business policy ensures that it will not become liable to these coercive penalties.24 The bank is like this because, by being reliable, it makes itself an attractive large-scale business organization. 23

Baier, “Trust,” p. 113. There is a question about the compatibility of coercion and trust and of sanctions and trust. Baier thinks that trust is compatible with minimum levels of each of these phenomena (see p. 120). I do not want to get into this slippery topic here, but my position is that the penalties that banks face if they act badly 24

Trust and Institutions

77

Conversely, the bank is a reliable penalizer. Whenever I do something that violates banking policy, the bank quickly penalizes me. As much as I might not want the bank to reliably penalize me, I do want it to reliably penalize others because not doing so would be bad business; and, it goes without saying, I want my money kept by a good business.25 Unavoidably, I have gone some way in answering the normative question. This has happened because banks are the way they are in virtue of our wanting them to be this way.

3.6 Trust and Institutional Trust One cannot trust a bank though one may rely on it. I expect the following objection: There are two broad categories of trust, personal trust and institutional trust, and you have ignored the latter. Accordingly, you must, of necessity, find it impossible for banks to be trustworthy. But, with a more finegrained approach, no such result need be reached: one may “institutionally” trust a bank. Such an objection does not carry much weight with me. What is this institutional trust after all? It is not simply good enough to be told it exists, rather, I want to be told what it is. That said, let me further explain why I think this objection is no good. My approach to explaining what trust is has been to identify a complex of phenomena that are present whenever trust is present. The point there is that trust itself is not a simple phenomenon, instead, it is a complex of phenomena. What I have shown is that many, if not all, of these phenomena are absent in the relationship between a certain type of institution and an individual. But, if the phenomena are absent, then so is trust. The onus is on the objector to say what complex of phenomena constitutes institutional trust. But, there is a further challenge, one that comes from Benacerraf’s article, “Mathematical Truth.” It is not just good enough to point out some complex and claim it to be institutional trust. Rather, the objector must say why these phenomena constitute a type of trust at all. The personal case is the easy case. All of us have experienced trust, or violations thereof, with other are such as to displace their motives for acting well. That is, the fear of penalty underlies good behavior and this is enough to make it the case that the bank’s behavior directed at me is not that of a trustworthy partner. Moreover, my fear of penalties imposed by the bank is enough to make my good behavior towards the bank good for the wrong reason. Ceteris paribus, my behavior is not trusting behavior either. 25 That I don’t want to be penalized shows no contradiction in my argument, but rather that it is human nature to not want to get penalized.

78

Laurence James

individuals, and we have done so long before any institutions ever existed. Not much has ever been said about institutional trust though, and before I can admit that trust is as fine-grained as the objector would have it, she needs to provide grounds for thinking that some other complex of phenomena can be rightly considered to be in the trust family.26 Moreover, the objector needs to say why what we feel for banks isn’t mere reliance, for then the thesis about there being an important notion of institutional trust would reduce to the idea that institutional trust is equivalent to reliability. If that were to happen, we would fail in our attempts to find a unique notion of institutional trust. Let us suppose the objector to respond that she does have a theory of institutional trust. In “Trust and the Ethics of Health Care Institutions,” Susan Dorr Goold attempts to sketch the difference between interpersonal trust and institutional trust. She begins by asking what the similarities between the two are in the context of health care. She claims that the two types of trust are similar in that in each, “patients (or potential patients) are vulnerable to both individuals and institutions.... Imbalances of power … are unavoidable in both types of relationships. Conditions of risk … are equally important. Uncertainty of outcomes is, likewise, a feature of interactions with both…”.27 Though she has listed a number of things, the most salient is empowerment. Here it looks as if the individual is equally disempowered in both the personal and institutional case.28 Let us follow along and suppose for the sake of argument that there is no difference with respect to empowerment between the interpersonal and the institutional cases. Since empowerment is one of the phenomena that constitute trust this supposition would support the idea that we are onto a robust notion of institutional trust here. Now what of the differences in the case Goold skecthes? Besides the fact that institutional trust has “multiple (and perhaps conflicting) trust objects,” the first difference is that “expectations of competence might be stronger. …[I]nstitutional, and especially institutionalized error, is more frightening and less forgivable. Second, expectations of beneficence, and particularly of advocacy, are weaker, because they must be balanced by institutional fairness.”29 In addition to these two differences, Goold refers to deterrents as 26

In “Mathematical Truth,” Paul Benacerraf raises this idea in connection with theories of mathematical truth. If these theories share nothing in common with our ordinary, empirical notion of truth, why should we consider them to be theories of truth? 27 Goold, “Trust and the Ethics of Health Care Institutions,” p. 30. 28 Note that the personal case here is only quasi-personal since it occurs in the context of having a relationship with a private practice doctor. 29 Goold, “Trust and the Ethics of Health Care Institutions,” p. 30, my brackets.

Trust and Institutions

79

being important components of institutional trust, and she claims that the larger the number of deterrents, “the higher the expectation that trust will be honored.”30 In other words, we hold institutions to a higher standard and are less likely, if at all, to forgive them. We doubt whether they actually care about us and have our best interests at heart (because they have to worry about profit and other ‘objects’ that conflict with worrying about us). Moreover, we desire that there be reliable safeguards (‘deterrents’) against their taking advantage of us. Clearly, it seems to me that on the other dimensions of evaluation, institutions don’t do well. Goold chalks this up to the nature of institutions “diluting” the typical nature of a trusting relationship, that is, she subscribes to the erosion thesis.31 Michael Jellinek, “Erosion of patient Trust in Large Medical Centers,” appeals to similar things.32 It strikes me as odd that both assume that the relationship between institutional and individual trust relationships is one of erosion (as do many others who have thought about it). The No Trust thesis seems to provide a more satisfying explanation of why there appears to be such degradation in trust. People want, perhaps unreflectively, for there to exist a trusting relationship between individuals and institutions. Because one cannot exist, people project a trusting relationship onto the picture. They can’t project a robust trusting relationship because the trust in question is not robust. So they project a diminished relationship. Once one lets go of the Erosion thesis, one is better able to see this. Note, too, that when Goold tries to give a sketch of institutional trust, the differences she highlights are not differences of kind but rather of degree. Institutional trust, on her account, is eroded person trust. That she is unable to give us a distinct theory of trust suggests that at some level she is aware of the problem I am discussing. Goold fails because she does not provide us with either a unique, robust notion of institutional trust, or with an argument to show that institutional trust can realize the complex of phenomena that are associated with trust. What we want from our institutions is reliability. Reliability in protecting our secrets, reliability in avoiding obvious harms, and reliability in accomplishing what they are supposed to accomplish. We don’t want trust when

30

Goold, “Trust and the Ethics of Health Care Institutions,” p. 31. Goold, “Trust and the Ethics of Health Care Institutions,” p. 28. 32 See Jellinek, “Erosion of Patient Trust in Large Medical Centers,” p. 16, for example. 31

80

Laurence James

it comes to institutions.33 If, in fact, this is correct, then it will involve a lot of rethinking about how we deal with and understand institutions. We shouldn’t get our hopes up and be disappointed when an institution fails to show an interest in our welfare. We shouldn’t think, not even for an instant, that institutions will accommodate us in times of need. Nor should we think that an institution can be ennobled by our acts. While trusting a friend is to reveal a great deal to him about how we see him, no such thing can be revealed to an institution for there is no one to reveal it to, and more importantly, no one who could genuinely accept it. What we will expect of institutions is reliability of the highest degree possible, and they should not expect us to forgive their failings. In the absence of trust, my involvement with institutions is for myself (as their involvement is for them), and so I have no reason whatever on which to base forgiveness. I suspect that if not there already, our litigious society is on its way to recognizing the main insight here: that we can have no real trust in institutions.

33

Some of us want trust. But, it is important to note that it is inconsistent for an individual to both want the sort of reliability that institutions have as well as trust. This is because, as I have argued, the two are incompatible on a number of fronts. For those of us wanting both, we need to give up one or the other if we wish to have a consistent set of desires.

Trust and Institutions

81

WORKS CITED Askvik, Steinar and Nelleke Bak, Trust in Public Institutions in South Africa (Burlington, Ashgate Publishing Company: 2005). Baier, Annette, “Trust,” The Tanner Lectures in Human Values (Delivered at Princeton University, March 6-8, 1991) Benacerraf, Paul, “Mathematical Truth,” The Journal of Philosophy 70 (1973): 661-679. Goold, Susan Dorr, “Trust and the Ethics of Health Care Institutions,” The Hastings Center Report 31 (2001): 26-33. Jellinek, Michael, “Erosion of Patient Trust in Large Medical Centers,” The Hastings Center Report 6 (1976): 16-19. Murphy, Jeffrie, “Forgiveness and Resentment,” Midwest Studies in Philosophy 7 (1982): 503515. Thomas, Laurence, “Forgiving the Unforgivable,” in Eva Gerrard and Geoffrey Scarre, eds., Moral Philosophy and the Holocaust (Burlington, Ashgate Publishing Company: 2003), pp. 201-230.