Autonomy, Trust, and Respect

Abstract

This article seeks to explore and analyze the relationship between autonomy and trust, and to show how these findings could be relevant to medical ethics. First, I will argue that the way in which so-called “relational autonomy theories” tie the notions of autonomy and trust together is not entirely satisfying Then, I will introduce the so-called Encapsulated Interest Account as developed by Russell Hardin. This will bring out the importance of the reasons for trust. What good reasons do we have for trusting someone? I will criticize Hardin’s business model as insufficiently robust, especially in the context of health care, and then turn to another source of trust, namely, love. It may seem that trust-through-love is much better suited for the vulnerability that is often involved in health care, but I will also show that it has its own deficiencies. Good health care should therefore pay attention to both models of trust, and I will offer some tentative remarks on how to do this.

Trust is much easier to maintain than it is to get started and is never hard to destroy.

(Annette Baier 1986, 242)

I. INTRODUCTION

This article seeks to explore and analyze the relationship between autonomy and trust, and to show how these findings could be relevant to medical ethics. In the first part, I will present a well-known and influential attempt to bring together the notions of autonomy and trust––an attempt that has clear consequences for the practice of good health care. I will not present any arguments either to support or discredit this theory of relational autonomy, but I will show that there is an interesting dialectic between the demands of trust and autonomy that renders its practical recommendations much less straightforward. Next, I will turn to the topic of trust itself, and I will set out the rough contours of this notion with the aid of the work of Annette Baier and Trudy Govier. I will then introduce a more precise yet broader framework for thinking about trust: the so-called Encapsulated Interest Account as developed by Russell Hardin. A brief discussion of this account will bring out the importance of what I will call the “reasons for trust.” What good reasons do we have for trusting someone? Can trust ever be justified or not? I will criticize Hardin’s business model as insufficiently robust, especially in the context of health care, and then turn to another source of trust, namely, love. I will claim that although there are no ultimate reasons for love, the matter of “being loved” is nevertheless epistemologically accessible. As such, we may have good evidential reasons for trust.

It seems that trust-through-love is much better suited for the vulnerability that is often involved in health care, and for the asymmetry in both knowledge and power that exists within this context. Nevertheless, I will also show that it has its own deficiencies.

Good health care should therefore pay attention to both models of trust (the business model and the model of trust-through-love). My modest suggestion as to how we can do justice to this second strand, that is, how we can integrate trust-though-love within health care, is that the principle of respect for autonomy should be interpreted in a way that incorporates respect for existing relationships of trust-through-love.

II. THE DIALECTIC BETWEEN AUTONOMY AND TRUST AND ITS RELEVANCE TO MEDICAL ETHICS

There is an interesting dialectic between the concepts of autonomy and trust. As is well-known, autonomy-based moral theories––especially Kantianism––were criticized for their overemphasis on this particularly narrow, overbearing, and “masculine” value, and it was argued that the focus should shift to particular (instead of universal) human relationships, to the role of emotions (instead of reason), and to the importance of trust (Baier, 1986; Held, 1990). As a result of this criticism, such a trust-oriented ethics was sometimes presented as an alternative for its autonomy-centered rivals.

However, the appeal of autonomy also offered a strong appeal to the project of emancipation that was essential to feminist ethics (Meyers, 1989). Therefore, in the wake of these forceful criticisms efforts could be discerned that would integrate trust and autonomy. Most importantly, autonomy was no longer conceived as an ideal of independence, of persons as isolated “mushrooms” without any ties to others (Held, 1990). The Aufhebung of the conflict between these two notions consisted in what was called “relational autonomy” (Mackenzie, 2008). Social relationships are causally necessary for and even constitutive of individual autonomy (Meyers, 1989; Govier, 1993).

Now, importantly, relationships of trust play a fundamental role in this process. Catriona Mackenzie, for example, refers to Axel Honneth and Joel Anderson’s work on social recognition as quintessential for a healthy relationship-to-self, and thereby for autonomy (Honneth, 1995; Anderson and Honneth, 2005; Mackenzie, 2008). One of the prerequisite components for autonomy is self-trust: people need to have a sufficient level of self-confidence in their own desires and emotions. They need to understand themselves as beings with needs and desires that they can expect others to take into account. The dynamics of this process is complex. Object-relations theory teaches that babies and small infants need to be able to rely on the care of their parents in order to develop “basic self-trust,” they need to be able to trust others so that they come to see themselves as beings who can express these needs and emotions without fear of rejection and expect that––within boundaries, of course––these needs and emotions will indeed be taken into account. They also learn to trust themselves by not having all their wishes immediately granted because that is how they come to see themselves as separate individuals who can––within boundaries, of course––do without others. The important lesson, however, is that this independence is gained through dependence, that is, through decent caring relationships.

So, according to the relational autonomy thesis, trust and autonomy are related via self-trust: in order to develop self-trust we need to be able to trust others, and self-trust is a necessary constituent of autonomy. One of the most important lessons of this literature is that self-trust develops from a very early age; we develop it by being able to trust (some) other people, our primary caregivers. We develop it because our vulnerability, in terms of our not being self-sufficient, is recognized as such, and “taken care of” by other people responding appropriately to our needs and emotions. This, of course, does not mean that, once we have developed “basic self-trust,” nothing can go wrong. Violations of one’s physical integrity (e.g., rape) are clear examples of when a rupture of trust with regard to other people—when our legitimate expectations are brutally violated––often translates back into a breakdown of self-trust and basic self-confidence (Govier, 1993). ¹

Now, the complex relationship between autonomy, trust-in-others, and self-trust is relevant to medical ethics. According to Mackenzie, for example, we can second-guess a person’s expressed medical preferences or her informed consent on the basis of the evidence that we have for that person having an insufficient level of self-trust. Self-trust, says Mackenzie, is necessary for a person to possess autonomy-as-normative-authority (Mackenzie, 2008). To illustrate this, she discusses the case of Mrs. H., a woman who, due to her illness (aggressive bone cancer), was abandoned by her dominant and dominating husband because he found her condition “burdensome and embarrassing.” Previously, she had always deferred to her husband’s decisions, and now––although her prognosis is very promising––she refuses further treatment. Mackenzie, however, urges that we should not take this wish at face value. It is obvious that Mrs. H. presently feels worthless and that her lack of self-respect is due to her husband and his behavior (i.e., his abandoning her). That is why she now wants to die. But this is not an autonomous decision. It is a decision that came out of a breakdown of autonomy, and in such cases it is important to restore this level of self-respect instead of yielding to the expressed preferences of the patient.

So clearly, the notion of relational autonomy has implications for the field of medical ethics. By reconceptualizing the notion of autonomy, that is, by revealing its social character, the principle of respect for autonomy is also modified, for it means attaining to this underlying social dynamic, rather than ignoring it. However, before we move on to the topic of trust and its problems, it is important not to forget––in light of the complex dynamic between trust-in-others and trust-in-oneself––a far more mundane and unspectacular point. Surely, in order to develop self-trust, one also has to receive trust from others. ² Only then can one come to know oneself as a trustworthy person. Parents, for example, allow their teenagers to go out on a Saturday night, or let them use their cars, and this is how teenagers should prove that they are trustworthy individuals. And, by doing so, they not only prove their trustworthiness to their parents but also to themselves.

Of course, there are important differences between these two processes. The dynamics described in object-relations theory, for example, happens, from the perspective of the infant, without her being explicitly aware of it (Baier, 1986, 240–1). This is not the case for the second process. The teenagers in the example are perfectly aware of the fact that their parents trust them, and what they have to do to deserve that trust. Also, on the face of it, it is obvious that the latter process is about trustworthiness whereas the first is about trust or self-trust. However, it is difficult entirely to separate trust from trustworthiness. Babies and infants “learn” (although unaware of this learning) that their parents are “there for them,” that they are able to rely on their care. What they learn is essentially that their parents are trustworthy. They also “learn,” of course, that they can do without the care of their parents, and that they will not perish if their desires are not immediately fulfilled. They learn in that respect that they can also take care of themselves, as independent beings. As such, they come to understand themselves as trustworthy in the sense that they can rely on their own judgment and competence in order to get what they want. It is this latter process that is both continued and, at the same time, also transformed when children grow up and come to see themselves as trustworthy in the eyes of others, that is, as individuals on which other people can rely. The fact that it is about trustworthiness for others marks a clear difference then. But, of course, if there is any lesson to be learned from the Hegelian importance of social recognition, it is that there is an “inside” to this process as well. In order to see myself as trustworthy, I have to be (re)assured that I am trustworthy to others. Others bestowing trust upon me teaches me a thing about myself: that I am––apparently––a trustworthy being.

Now, putting emphasis on this second (and secondary) process has some implications for medical ethics as well. The story of Mrs. H. is a case in point. Mackenzie’s approach has been charged with accusations of paternalism and perfectionism (Christman, 2004), for it allows us to go against Mrs. H.’s contemporaneous wishes and preferences in order to assure that she is “genuinely autonomous.” This parallels Nancy Fraser’s criticism of Honneth that a psychological account of justice is bound to be a case of “blaming the victim:” sufferers of injustice will be denied the authority of autonomy, because they are said to “suffer” from impaired autonomy (Fraser and Honneth, 2003). Simply put: Mrs. H., once at the mercy of her dominant husband, is now––due to a benevolent attempt to address this injustice––not taken seriously by her health care professionals as well.

We can interpret these accusations of paternalism and perfectionism as a continuation of the dialectics between autonomy and trust. If so, then we should not necessarily reject the notion of relational autonomy altogether, on account that it is unduly paternalist or perfectionist. Instead, we could emphasize that the relational account should also incorporate the “self-trust-through-being-trusted” dynamic in order to do justice to its relational aspect. The upshot of this, however, is that we should acknowledge that there is no easy way of dealing with Mrs. H.’s impaired autonomy, for in denying her to have her dying wish respected we are in danger of further eroding her sense of self-respect. She needs to be able to see herself as someone who is capable of making her own decisions, and for that she needs to be the object of trust. No doubt, she also needs to trust the medical team that is responsible for her care, but in order to restore self-trust, she needs to be able to see herself as someone who can also be trusted in taking care of herself. In order to develop and maintain self-trust (as a necessary condition for autonomy), we should be able to trust others as well as receive trust from them. If trust has this dual aspect, if self-trust is developed and maintained by being able to trust others and by being trusted by others, then the dialectical relationship between autonomy and trust is further complicated and these complications would render the relational autonomist’s practical recommendations for health care ethics far less straightforward. But it would not invalidate the theory itself.

III. THE CONTOURS OF TRUST

Until now, in presenting and complicating the relationship between trust, self-trust, and autonomy, I have taken the notion of “trust” for granted. I pretended that we all know what trust is, and what it means to trust someone. However, if we further want to examine the relationship between autonomy and trust, it might be helpful to examine the concept of trust more clearly. Let us therefore focus on what is relatively uncontested.

First, trust is something that holds (or fails) between people. Although we say things like, “I trust this bridge to hold me,” or “my trusted steed,” what we mean to convey by such phrases is that we expect this bridge not to collapse under our weight, and this horse to take us wherever we want. Trust necessarily involves expectations, but when mere expectations turn out to be unwarranted we tend to be disappointed whereas, when trust is violated, we feel disregarded or betrayed (Baier, 1986, 235). We respond with reactive attitudes (resentment or gratitude) and hold people accountable for such violations of trust.

Secondly, trust involves positive expectations: we expect the trusted party to take our interests into account. Trust involves that “one believes that [the trusted party] is likely to act kindly and benevolently toward one, that she is unlikely to harm one, that she is well-disposed toward one” (Govier, 1993, 104) And indeed, phrases like “I trusted you to steal my bike, or hurt my feelings,” just sounds odd (and again, the neutral “to expect” seems to capture our intentions much better).

A third feature is that trust involves an assessment of the trusted party’s competence. We believe that she is able to do what we expect her to do. Reversely, we do not trust a person whom we know is incompetent in that field. We trust our banker in different domains than our hairdresser or close friend. And it is perfectly possible that we do not even trust a close friend with our financial affairs, no matter how benevolently disposed he might be towards us.

Fourthly, although the expectations that are involved in trust may seem very specific, trust is always, to some degree open-ended (Govier, 1993, 104). When I say things like, “I trusted you to lock the door,” this seems to be about very specific expectations (i.e., about you locking the door), but this is circumscribed by a much more diffuse and vague expectation that is related to the trusted party being benevolently disposed toward oneself. Even if I would have been very specific, and would have asked you “Please, lock the door when you go out!” I would have expected you not to lock the door if you knew that I was still inside and had no way of getting out.

Baier, in this regard, points to the discretionary power of the trusted party (Baier, 1986, 240). In determining her action, we expect the trusted party to use this power both with regard to her competence (e.g., when we rely on her professional judgment it is indeed important that she decides because we acknowledge our own lack of competence in these matters), as to her ability to track our interests in new or unexpected situations (what would be good for us in one case would not be necessarily good in another).

Fifthly, trust is intimately related to vulnerability, and in at least two ways. First, we render ourselves vulnerable through trust as it puts in place a certain risk that was not there before. When we trust a specific someone, that someone holds a power over us that others (usually) do not have. Thus, Baier defines trust as an “accepted vulnerability to another’s possible but not expected ill will” (Baier, 1986, 235). And indeed, if I ask a stranger on the train to watch my belongings while I go to the bathroom, not only can she fail me by not paying attention to my belongings at all, but she is also aware of my vulnerability and this makes her perhaps the best candidate for actually stealing my bags. Second, vulnerability is often also involved as a background condition for trust: we only need to trust others to the extent that we are vulnerable prior to, and therefore without, the trusting relationship. I need the stranger in the train to watch my bags because I (urgently) have to go to the bathroom and cannot take them with me. Trust is a way of dealing with such vulnerability. So, paradoxically, trust is a response to a vulnerability that involves making us even more vulnerable to the specific (trusted) party that needs to protect us from being so vulnerable.

IV. TRUST: THE ENCAPSULATED INTEREST ACCOUNT

The above-mentioned “rough features” might help to bring out the contours of trust, but they do not settle the question of what trust essentially is. Russell Hardin gives the following very concise definition of trust:

To say we trust you means we believe that you have the right intentions toward us and that you are competent to do what we trust you to do. (Hardin, 2006, 17)

Hardin believes that all tenable accounts of trust agree on this definition, but disagree on what counts as “the right intentions” towards the one who trusts. Let us focus first on the common ground.

Trust, according to Hardin, is essentially a belief and not a voluntary attitude or disposition. ³ This seems strange because we think of ourselves as freely engaging in acts of giving or withholding trust, that is, we tend to think of trust as an attitude we “take on,” whereas a belief seems far less under our control. But this is mistaken. Of course, attitudes and dispositions are involved but must be located on the side of those who are the object of trust, not on the side of those who do the trusting. When we trust someone, we believe that she has a favorable attitude towards us, that she will take our interests into account, but this is indeed a belief on our part (and a belief that may turn out to be incorrect, for that matter).

More specifically, the belief is a belief about that person’s trustworthiness. If we believe someone is trustworthy, we trust her (that is, the belief is constitutive of trust). This, again, might strike us as wildly implausible, because we may deem many people trustworthy and yet choose not to trust them in any specific sense. But, on second thought, it does seem that we trust these other people, although we choose not to make use of their trustworthiness. We do say things such as, “It is not that I don’t trust them, it is just that I don’t believe that we should burden them with this information.” We simply have certain expectations with regard to how others behave in relation to our interests. Moreover, it does seem that, as a belief, trust indeed shows a sense of recalcitrance. We cannot simply choose to trust others if we believe them untrustworthy. I cannot choose, for example, to simply trust my fellow train passengers (whom I previously distrusted to the extent that I needed my neighbor to keep an eye on my belongings). On the other hand, it does seem that we sometimes take a leap of faith. We sometimes do choose to trust someone even when we know (believe) that she is untrustworthy. But, come to think about it, it is perhaps more correct to say that we still do not trust this person, although our interests (or even our fate) may depend on her. We sometimes do not trust people on whom we nevertheless have to rely. So trust is essentially a belief about another’s trustworthiness.

Hardin distinguishes between three models of trust: one based on the encapsulation of interests, one on morality, and one on dispositions or character. He does not say much about the latter two, but simply favors the first––the Encapsulated Interest Account––mainly because it has the strongest explanatory force. On this view, to believe that someone is trustworthy (i.e., to trust someone) means that you believe that she has encapsulated your interests, that is, that your interests are somehow her interests, and that she has encapsulated your interests qua your interests (Hardin, 2006, 19). For example, although it is true that I “trust” other drivers to keep to their side of the road, it is obvious that these other drivers only care about their own interests. Yet, I can expect them to remain where they are just because they care about their own well-being, and not necessarily about mine. This, according to Hardin, is not a matter of trust. What happens in this case is that our interests merely coincide: they are not genuinely encapsulated by the other party (Hardin, 2006, 20).

Hardin then goes on to distinguish three “mechanisms” on the basis of which one’s interests can become encapsulated by another person: (a) because we have an ongoing relationship with the other that she wants to maintain, (b) because the other loves me or considers me her friend, or (c) because the other values her general reputation. Hardin disqualifies love and friendship because, according to him, this cannot explain trust in institutions. But (a) and (b), he says, go a long way in explaining many ordinary relationships of trust.

Most importantly, this model is highly applicable to market transactions. For example, it is clear that most companies want happy customers, not just because they want them to be happy as such, but because happy customers also tend to continue their business with them. Therefore, the businessman’s (or businesswoman’s) interests are tied to the interests of the customer, and because he is expected to profit from their ongoing relationship, he can be the proper object of trust. A concern for reputation even further extends the Encapsulated Interest Account, for it is now in the interest of the businessman to be trustworthy to his customers because this will also be beneficial with regard to his dealings with others. Not taking into account your interests will damage his reputation and, as a foreseeable consequence, his relationships with others will also suffer from this fact. For example, on this model of trust, if I leave my children at the daycare center, I can assume that any complaints on my part will have negative consequences for its reputation. Other parents might lose their trust as well, and the daycare center would be in danger of losing their clients. Therefore, this mechanism allows me to trust the daycare center and its employees, because I know that they care about my caring about good daycare. They have encapsulated my interests.

The question of trustworthiness is therefore about whether or not we have good reasons to believe that there is indeed an encapsulation of interests. Indeed, the idea that trust is essentially a belief has the advantage that we can easily explain that it can be unwarranted, for we might trust in the absence of good reasons. Now, although this business model––based on repeated interaction and reputation––may indeed warrant such a belief, we should notice that it still delivers a rather superficial and shaky basis for trust. First, the line between encapsulated and merely coinciding interests (e.g., the driver on the other side of the road) is very thin. Companies do not really care about my interests qua my interests, but only to the extent that such ‘caring’ might contribute to their own (financial) well-being. My interests are ultimately instrumental to their own. Secondly, and more importantly, the trust that is the upshot of such relationships is not very robust, since the trustworthiness only holds as long as the relationship promotes the trusted party’s self-interest. However, we often want to be able to trust someone even if this would imply a considerable cost to the trusted party. If that person would betray our trust whenever it would be beneficial for her to defect, our trust would be very limited and provisional. Also, if the risks are distributed unevenly over the parties, then the strength of the encapsulation is weakened. The businessman might stand to gain a little (e.g., one customer), while I may stand to lose a lot (e.g., my children). Market relationships are often characterized by such inequalities, and the fact that they are often regulated by contracts should make us aware of the fact that we do not really trust these “selfish” mechanisms after all. Encapsulation-through-self-interest might go a long way in explaining some kinds of trust, but it surely does not cover them all. (On an intuitive level, the example of the daycare center already suggests that it is the wrong model in some contexts.)

V. LOVE AS A BASIS FOR TRUST

Let us therefore turn to the mechanism of trust that Hardin puts aside: love. This does not fundamentally deviate from the Encapsulated Interest Account, because love is a ground for encapsulation. See for instance Baier:

Why should [the child] have confidence that parents are dependable custodians of such goods [such as nutrition, shelter, clothing, health, education, etc.]? Presumably because many of them are also goods to the parent, through their being goods to the child, especially if the parent loves the child. They will be common goods, so that for the trusted to harm them would be self-harm as well as harm to the child. (Baier, 1986, 243)

The interests of parent and child are, at least from the perspective of the parent, not just encapsulated but fundamentally intertwined. If the parent loves the child, not to take her interests into account would be to disregard her own well-being. This, however, should not be misunderstood.

Harry Frankfurt is particularly clear on this point. It is not that the parent cares about her child because she cares about herself. Her love is certainly not instrumental in this regard. Love as exemplified by the love between parents and children is essentially disinterested: it is focused on the well-being of the beloved object (the child) and not on the well-being of the loving parent (Frankfurt, 2004). This makes sense, given the fact that many of these loving parents willingly make considerable sacrifices in order to take care of their children. Also, if parents would fundamentally be concerned about their own well-being, then they would try and give up their love for unhappy children, since they would “drag them down with them.” But they usually do not do this.

It is this feature of love––its disinterested character and the corresponding willingness for self-sacrifice on part of the lover––that makes it a profound source for trust. We can trust such loving people even if a concern for our interests would go against their own. Or, put differently, we know that our interests are intertwined to such an extent that they virtually coincide. Yet, whereas Hardin uses the term “coinciding interests” to capture a purely selfish concern for taking into account another’s interests, a Frankfurtian analysis reveals a deeper sense of “coinciding” that is fundamentally disinterested. If you love someone, then the interests of your beloved also become your interests, but not in an instrumental way; her interests come first and, because of that, love will often require you to go against your purely selfish interests. This feature gives trust-through-love a depth that is lacking on the business/market model.

But if trust is a belief in the other’s trustworthiness, in terms of having encapsulated your interests by virtue of loving you, how do you know––that is, have good reason to believe––that you are loved? If we may disregard Cher’s heartfelt advice that the answer to that question is in a person’s kiss, ⁴ this seems a very difficult question to answer. Perhaps we should even agree with Cher that the evidence can only be circumstantial. The main problem is that so-called Reason Views on love have fallen out of fashion. If there were something about us that would give others a reason for loving us, then it would be a lot easier to feel justified in trusting these other people. To the extent that they should love us, we would have a reason to rely on them. But such a reason-based view of love seems both preposterous and scandalous. We do not love someone to the extent that she possesses certain qualities (beauty, wit, intelligence, kindness, etc.). Of course, we do believe that our loved ones have these qualities, but the relation seems to be reverse: it is because we love them that they have them (or so we believe) and not the other way around. On the Reason View, our love would only be deserved to the extent that it tracks these positive features, but this would imply that we would have more reason to love our neighbor’s wife or children, should they indeed be more “qualified” in this regard. And that is absurd. All love would be provisional until the opportunity would come to “trade up” (Wolf, 2002). And that is quite appalling.

To the extent that there are no reasons for love, all love-based reasons for trust seem to be epistemically unstable as well. If there is ultimately no reason for loving us, then it seems that we have no reason for believing that we are the object of love as well. But this is a non-sequitur. We sometimes just know that we are loved, and therefore know that we can trust that person. Children know this about their parents, and we know it about the ones with whom we choose to share our lives. Epistemologically then, love is still accessible. Although there may be no good reasons to love someone, and although trust-based-on-love may be quite groundless in that sense, someone may nevertheless have good reasons, that is, evidence for believing that she is being loved. And that is what trust requires.

In fact, the no-reason view professed by Frankfurt gives us an indication of what love requires of the lover, so that, if she does not hold up to these requirements, she does not really love us. First of all, since the well-being of the lover and the beloved are so intertwined, one’s own setbacks should somehow be mirrored in those who claim to love you. If they thrive and do particularly well when you are utterly miserable, then there is reason to doubt their love for you. Frankfurt, given his will-based account of love, locates the test, so to speak, in a willingness to act on behalf of the lover for the sake of the beloved. In other words: if the lover does not do anything to alleviate the shroud of misery or depression, then his love is bound to be untrue. However, although it is true that love involves more than “rooting for the best” (Velleman, 1999), it may not require some explicit act of will either. What seems vital in any case is that there should be an appropriate emotional response to the beloved’s happiness or despair. If that does not affect the lover in the appropriate way, then there is reason to believe that there is not the required “encapsulation of interests” or “common good.”

So, there are ways of knowing that one’s interests are sufficiently encapsulated so that trust is warranted. What is very difficult, however, is to make people love you so that you might trust them. This is also why it is difficult, at least from a love-based perspective, to install or create trust from the ground up. There is no basic reason that can justify your trust in others because there is no reason that others should love you. ⁵ There can only be evidence for the fact that particular others do, in fact, love you (or not). The only way, then, that trust-based-on-love can be grounded in reasons is on the basis of evidence. The trusted party has to show that she cares for the trustee.

VI. AUTONOMY AND TRUST IN MEDICAL ETHICS

What lessons can we draw from this analysis of trust? How is this relevant to medical ethics? First, we need to note that, on first sight, trust-based-on-love seems very appropriate or even necessary in the context of health care. People that need to rely on health care are often very vulnerable, and so they need to be able to trust others. The stakes are often very high as their health or even their lives are at risk. Yet, the relationship between patients and health care professionals is very asymmetrical, both in terms of competence and power. Also, and more importantly, the “costs” to the patient in case of a breach of trust are considerably higher than the ones faced at the other end of the relationship. This, as in the example of the daycare center, makes it doubtful that trust should be based on such superficial grounds as mutual beneficence or reputation. The business model might be valid, and may even explain or justify some kinds of trust in health care, but it clearly has its limits and, in some contexts, does not seem particularly suitable.

Trust is easily destroyed, and this goes for trust in health care professionals as well. Sometimes we are confronted with blatant cases of abuse and neglect: health care professionals who are incompetent, indifferent, or viciously selfish, and who are therefore not worthy of trust at all. Trust on such occasions makes place for distrust and suspicion (O’Neill, 2002). We then start to wonder: What justifies our belief in the trustworthiness of our caretakers? Why should we assume that they are generally well-disposed towards our interests? Why should they indeed take our interests into account?

In light of such a breakdown of trust, it is an asset of the business model that it can help to install or improve trust where it is wanting. One can try to tie the interests of the professionals to the interests of the clients and this gives us an instrument to increase or create trust. But, as noted before, the kind of trust it could yield might well be defective.

Should we then opt for love? Could this model offer the appropriate basis for trust in health care? This seems doubtful for various reasons. For instance, although it might explain and justify the depth and sturdiness of the trust that is called for under conditions of increased vulnerability, it seems a stretch to say that health care professionals are trustworthy to the extent that they love (or care for) their patients. And even if this is so, it seems absurd to say that they should love us because this is not something we can either claim or demand. Or we might object that all this talk about love just begs the question. The demand for trustworthiness is merely replaced by a demand for love, but we essentially just want these professionals to take our interests into account. And finally, even if we would be able to put all these doubts behind us, we might still agree with Hardin that the model is not very adequate with regard to institutions (do they love us?).

And there are still more serious problems that haunt the love-based approach, for although it makes the trusting relationship non-instrumental, the intertwinement of interests might also be problematic. We want our health care professionals, at least to some degree, to be impartial; we do not want their medical competence (their knowledge and know-how) to be clouded or compromised by their love for us. They need to be able to keep their distance; the same distance that potentially allows for them to betray our trust.

So things are complicated when it comes to trust in health care. Neither model of trust is particularly satisfying. Or perhaps it is better to say that the one succeeds on the point at which the other fails (and vice versa): the business/market model tallies well with the competence requirement that is involved in trust, whereas the element of love seems better capable of securing the “general benevolent disposition” on the part of the trusted. So what should we conclude?

Let me end by offering at least one suggestion that might be helpful, a suggestion that ties these findings about trust to the dialectical relationship between autonomy and trust that I mentioned in the first section. If our vulnerability and the asymmetry in the health care relationship make the business model unsuitable, and if it is inherently difficult to create or install trust on the love-based model, then we should shift our attention to existing relationships of trust. That would at least be a way of combining both models and take into account their complementary nature.

What I mean is this. When someone relies on health care, this often involves the interests of many other people, most notably the ones who care about and love this person. My suggestion is that when such love is involved, the entailing trust should also be taken seriously. It is only natural and, if the analysis above is correct, also perfectly justified that people in a state of vulnerability should look for those who are well-disposed toward their fate. Of course, they also have to rely on the competence of health care professionals, and they might put trust in them as well, but trust-based-on-love might offer the depth that is so desperately called for. Respect for autonomy is nowadays often translated in the practice of informed consent. Now, it is clear that patients seldom decide in complete isolation and that there is often a strong intertwinement of interests: the interests of the patients are tied to the interests of those who love her, and also tied to the ones she loves (and these might be the same people). Therefore, to show respect for autonomy means that this social context and this intertwinement of interests should be respected as well. I have claimed elsewhere that proxy decision-making, in the case of incapacitated patients, could be justified as a way of respect for autonomy (Nys, 2013). I do not need to make this claim here. All I want to say is that whenever a patient trusts another person, this relationship of trust should be taken seriously. This, of course, does not imply that medical professionals or institutions should simply do whatever patients say, or yield to whatever decision is reached on the basis of such trust in friends and lovers. Much to the contrary, if trust is essentially a belief, then it can be misguided. What it means is that the evidence should be considered and that it needs to be part of the dialogue that is involved in respecting the autonomy of the patient. So, unfortunately, my analysis of trust and autonomy in health care does not offer any clear-cut answers or neat protocols. However, it opens up a different way of taking trust seriously and relating it to the principle of respect for autonomy.