Kious10/12/18
“Impermissibility, Consent, Coercion”
Rocky Mountain Ethics Congress
August 6-9, 2009
~
I. Background
It’s often suggested that consent is morally transformative.[1] Often, it can be permissible for us to do something to another person when he has consented, though we could not permissibly do it otherwise. If Tammy and John are both adults, then in the general run of things, Tammy may have sex with (or perform surgery on) John only if he consents. John’s consent transforms, as it were, the moral standing of Tammy’s act, making it permissible when it wouldn’t otherwise be.
Accounting for this transformative significance is a philosophical problem, and it isn’t easy to solve. Today I’ll say a little about some accounts others have given, and a little about why someone might be dissatisfied with them. Then I’ll propose a rather different sort of account—really, an error theory—according to which there’s no need to assume that consent has transformative significance at all.
The most familiar theory of consent is loosely Kantian. On that theory, the main thing, morally, is treating other rational creatures only in ways that respect their rational autonomy, and doing that involves treating them in ways they could rationally will. Accordingly, consent makes doing something to another person permissible when and because it is the result of a virtuous deliberative process: one that is rational and based on adequate information, and which, moreover, is attributable to the agent—it must not result from the control of others or from the effects of ‘external’ forces, so that the deliberations are the agent’s own.
This way of explaining things makes good intuitive sense. It seems to explain why the consent of children and the mentally disabled, or my consent when I’ve been brainwashed, isn’t morally transformative: children and the mentally disabled are not rational, autonomous creatures, and the consent I give when I’m brainwashed, if it results from the brainwashing, isn’t mine. This account can also explain the significance of coercion, in two interrelated ways. One suggestion is that coercion undermines an agent’s deliberative faculties, so that if he is coerced, he does not deliberate rationally. Another suggestion is that coercion makes it the case that, even if the agent deliberates well, his deliberation is not his own.
I think that there are a variety of problems with the Kantian view as it is usually presented, but the most straightforward of them pertains to the account of coercion mentioned just above. One problem is that, though it is probably often true that, if A mugs B, B’s deliberation about whether to hand over his wallet is defective, it surely is not always so. But when it isn’t, it still doesn’t make A’s taking B’s wallet permissible. One might just say that this shows that the problem of coercion is a problem with attributability. But those who take this line have a hard time distinguishing mugging someone from other cases where my consent is given under duress: if you’re a surgeon, and you tell me that you can prevent me from bleeding to death, after my nasty car crash, but will do so only if I agree to later pay you $10, it looks as though I’ve consented in a transformative way to your taking $10 from me, even though I faced the threat of death, otherwise.[2] Why should the decision be attributable to me when I need surgery, but not when you’re mugging me?
Worries like these have led many to eschew the Kantian account; at least, there have been many attempts to account for the significance of coercion in a non-Kantian fashion. Robert Nozick claimed that one person coerces another if and only if, via the execution of a plan of action she proposes to him, she would make all of his options worse than they would be at the statistical baseline.[3] This gets a number of cases right: if you take my wallet at gunpoint, either of the options you’ve presented me with make me worse off. And if Tammy tells John that she’ll fire him from his job unless he has sex with her, it’s plausible that she’s coerced him, that his having sex with her was not consensual, that it was not permissible for her to have sex with him, and that both of the options she gave him made him worse off—assuming he did not want to have sex with her anyway.
Nozick’s proposal runs aground on other examples, though. To borrow one from TM Scanlon:[4] John’s telling Tammy that he’ll quit his job unless she gives him a raise doesn’t usually amount to impermissible coercion, and if she gives him the raise it’s reasonable to suppose she did so consensually, and that it is permissible for him to take the raise, even though either of her options makes her worse off.
On Scanlon’s own view, consent and coercion should be understood at least partially in moral terms. He claims that “a threat renders a person’s consent invalid just when it is a threat the agent is [not morally] entitled to make.”[5] On Scanlon’s view, a threat is one an agent is not morally entitled to make just in case the threatened action is one that the agent is not morally entitled to perform. But there look to be counterexamples to this proposal: if you’re a terrorist and you give me several million dollars because otherwise I’ll tell the CIA about your nefarious plans, then arguably I’ve coerced you, and your action is non-consensual. It also seems that my threat is impermissible. But I’m surely entitled to inform the CIA—you have no claim against me that I not. Or consider another example:[6] John really loves Tammy’s old couch—they were once roommates and he has fond memories of all the time he spent on it. Tammy wants to get rid of the couch, but even more than that, she wants to have sex with John, so she tells him that she’ll throw the couch away unless he sleeps with her. To many, it seems as though Tammy has impermissibly coerced John, even though she’s entitled to throw the couch away.
Scanlon’s proposal may also face something of a bootstrapping problem. Presumably, his account of entitlement will hinge upon his account of moral permissibility, generally; but he also thinks that the principles determining moral permissibility are those that could not be reasonably rejected as the basis for informed, unforced general agreement.[7] If unforced general agreement is the same thing as non-coerced general agreement, then his account of coercion in terms of entitlement does not get us very far.
Now, I don’t think any of the problems above—which have in any case been sketched in a rather slapdash way—are fatal for the accounts against which they’ve been raised. But they lead me to suspect that we’ve been thinking about consent, coercion, and related phenomena in the wrong way all along. I despair of the possibility of providing an analysis of consent, or of related notions like coercion and autonomy, that helps us make sense of the moral terrain. I suspect that we could do better if we simply quit ourselves of the assumption that consent is morally transformative.
If this is right, we should be able to meet two challenges. First, we should be able to provide an account of when interactions with others are morally impermissible that does not make essential reference to consent, coercion, or related notions—as a view like Scanlon’s or Kant’s does—but which gets our intuitions about a broad swathe of cases right. Second, we should be able to explain the phenomena that have led us to say things like ‘consent is morally transformative’ and ‘coercion vitiates consent,’ even if we cannot explain what consent and coercion are.
II. Impermissibility
In the remainder of this paper, I will try to accomplish these tasks. I’ll start by describing, in outline, an account of moral impermissibility with which I’ve been tinkering lately. I can’t do much to motivate the account here, other than to say that I arrived at it because I thought that Scanlon’s contractualism is plausible, and wanted to see what it would look like if you detached it from anything that resembles a notion of consent.
The account of moral impermissibility has two parts. The first part is a theory of interests. I take it that persons can have interests—by which I just mean that certain things, happenings, circumstances, ways of being treated, and the like can be good for persons. I also take it that persons’ interests can be diverse. This means both that any given person can have interests in a variety of things, not all of which are necessarily reducible to an interest in some ultimate good like pleasure or happiness, and that the interests of different persons are not always the same. I further suppose that the diversity of persons’ interests is a result both of variation in their conative attitudes (what they want, care about, believe is important, and so on), as well as of facts about them and their circumstances apart from their attitudes. So, for example, I think it is in my interest to go mountain biking while I’m here in Boulder. Part of what makes this the case is that I like mountain biking. It also helps that I am fit and healthy. There are some for whom mountain biking would not be in their interest, and this may be because they do not like mountain biking, or because their circumstances are such that it wouldn’t be good for them, independently of what they want. If Milt has a broken wrist, say, then mountain biking probably isn’t in his interest, even if he likes it.
Finally, I assume that persons’ interests can be more and less important, and that comparisons of importance can be made both between the interests of a single person and between the interests of two or more different people. I have an interest in doing good philosophical work, and an interest in mountain biking. It would not be implausible to say that my interest in doing good philosophical work is more important than my interest in mountain biking. Likewise, though I’m willing to concede that my wife has an interest in watching “American Idol,” and that this interest is somewhat important, I think it is probably less important than my interest in mountain biking, and is certainly less important than my interest in doing good philosophical work. Of course, my wife may disagree about this.
A qualification is in order: though I assume that persons’ interests can be of greater or lesser relative importance, this is not to assume that all the interests of all persons can be ranked against each other. The most I want to assume is that there is a partial ordering of interests, such that we can rank some interests against each other, but not others. Likewise, I do not suppose that it will always be clear what the relative importance of two interests are, or that we can always reach agreement about this.
Now, if persons’ interests have the properties mentioned, I think the following principle, if properly understood, does a pretty good job of predicting when actions are morally permissible and impermissible:
Pair-Wise Comparison Principle (PCP):an action is impermissible if and only if the net expected harm resulting to some person is greater than the net expected benefit to be achieved for others, where this comparison is made in a pair-wise fashion.
This principle needs some interpretation. What it says is that we should perform an action if and only if the net expected benefit of the action outweighs the net expected costs. The expected benefit or harm of an action is proportionate to the relative importance of the interests that are likely to be promoted or set back, and to the likelihood that this will occur. The net expected benefit or harm of an action is the difference between the expected benefit of the action under consideration and the other most beneficial course of action open. Finally, to suggest that the comparison of benefits and burdens is to be performed in a pair-wise fashion is to say that we are to ask, for each person affected by the action, whether the net expected harm or benefit to her is outweighed by the net expected benefit or harm to someone else. If there is some person for whom the net expected harm is not outweighed by the net expected benefit to someone else, then the action is impermissible. If there is someone for whom the net expected benefit is not outweighed by the net expected harm to someone else, then the action is permissible, and, in fact, required.
Why think PCP is true? Well, suppose A wants to murder B, just for fun. This would be impermissible. PCP can explain this, in that A’s interest in petty amusement is, plausibly, less important than B’s interest in living. Or suppose that I am outdoors and my toes are at risk of frostbite, and I see that you have an enormous stack of firewood—enough to heat your house every night for a year.[8] I think I may steal some of your wood and make a fire to warm myself, if that is the only way I have of saving my toes. PCP predicts this because, presumably, my interest in saving my toes is going to be more important than any interests you’re likely to have in maintaining your stock of firewood. Note, though, that if there was some other way for me to keep warm—I have some wools socks and could easily put them on—it wouldn’t be permissible for me to steal your firewood. PCP predicts that, too, because in that case the net expected benefit to me of my action is slight. Or, for another case, suppose that I have promised to help you move your desk at 3:00 this afternoon, but while I am on my way to meet you I see that someone has been in a car accident and needs emergency medical attention. If I stop to aid her, I will be too late to help you. If her need is serious, I am competent to provide the care, there is no one else around who is also competent, and in breaking my promise to you I am not likely to expose you to very serious harm, then I think I ought to stop and help, my promise notwithstanding. PCP clearly gives this result; and would add that it will tend not to be true if the victim’s need is less, my competence to aid her is less, there are others equally well positioned to aid her, or the harm to you of breaking my promise is greater.
III. Consent
There are, of course, a lot of reasons to worry about PCP. But, largely for want of time, I’m going to skip over those worries and see whether the principle can ground a satisfying theory of consent. I think it can. I’ll call that theory the evidentiary account. It says:
To a first approximation,[9] a person’s consent to being treated in some way is morally significant just insofar as it is evidence that it is morally permissible to treat her that way, given her interests; consent will not be morally important when, for one reason or another, it cannot play this evidentiary role.
If PCP is true, A’s consenting to B’s doing something to him is usually (not always) morally important just because of B’s epistemic limitations: A may be aware of reasons for and against B doing what she plan to do, of which B is not. In particular, A may have a better idea of how it is B’s actions are likely to affect A’s interests, and of the importance of those interests, than B does. This will be because A’s interests, their importance, and how they are likely to be affected by B’s actions, depend in large part on facts about A’s psychological and non-psychological circumstances of which, often, A will have better knowledge than B. Consequently, B’s asking A whether B may do something that is likely to affect A is often a good way to determine whether the balance of reasons favors B’s doing it.
How does the evidentiary account handle the cases I mentioned earlier? To start, it makes good sense of some of the paradigm cases where consent seems to be morally important. It is typically impermissible for a physician to perform non-emergency surgery on her patient without his consent; on the epistemic account, this is because having surgery might not be in the patient’s interests, given the range of interests a person could have; asking him may be the best way to find out.[10] Likewise, it is almost always impermissible for Tammy to have sex with John without John’s consent; this is because whether sex will be in John’s interest depends largely on whether he wants to have it. In particular, it’s true that if he doesn’t want to, it will be very bad for him. And in the usual run of cases, it’s never true that Tammy’s reasons for having sex with John are good enough to outweigh the badness of unwanted sex for John.
The evidentiary account accordingly explains the appearance that consent is morally transformative as a mere appearance. It’s not the John’s consent makes Tammy’s having sex with him permissible. Rather, it was (in the case where he consents) permissible all along; his consent merely shows that this is the case. And when he does not consent—especially, when he refuses—it shows what was already true, that Tammy’s having sex with him is impermissible.
The evidentiary account also makes sense of many cases where consent looks unnecessary, including cases of so-called ‘hypothetical consent.’ The reason it is permissible for me to push you out of the way of the bus without asking is that it’s very unlikely that you have good reason for standing in front of the bus, and very likely that you have good reason to want to be out of its way (it’s also very unlikely that I’ll do you greater harm by pushing you than the bus would; it wouldn’t be permissible for me to push you out of the way of the bus with my semi-truck).