Sunday, December 30, 2012

Weight Discrimination in Drug Rationing

Suppose we have a severe shortage of a drug that is prescribed on a "grams per kilogram" basis, i.e. a 300 pound patient needs three times as much of the drug as a 100 pound patient.  Should weight then play a role in allocation decisions, such that (all else equal) lighter patients will have priority over heavier patients, or would it be more fair to simply allocate by lottery until the drug runs out?

It seems to me pretty clear that we ought to prefer to help more rather than fewer people, all else equal, even if it means that all those helped have a characteristic in common (low weight).  But I expect that a lot of people would disagree with this, and automatically regard it as "discriminatory" and hence "unfair".  Can such objections be rationally defended?

Such objections remind me of Harris' confused claim that the QALY metric is inherently "ageist" and "ableist".  In both cases, the objector moves too swiftly from unequal outcomes to the procedural conclusion that members of the less-helped group must have been unjustly neglected.  But this is simply fallacious.  It's true that it would be unjustly discriminatory to count the interests of some people as less important than those of others, on the grounds of some morally irrelevant feature like age or weight.  But unequal outcomes are not always the result of counting some people's interests for less.  When utilitarians advocate for these unequal outcomes, they do it precisely because they are counting all people's interests equally, and there just happen to be contingent practical reasons why some people can be more easily helped than others.  A morally irrelevant feature (i.e., of no intrinsic moral significance) may be highly practically relevant (i.e., of great instrumental significance) to the securing of morally worthwhile ends.

In the age case, the appearance of "ageism" results from the practical fact that younger people tend to have greater life expectancy, and so a persisting health benefit will tend to help them for longer than a superficially similar "benefit" to an older person.  The duration of a benefit, while less salient than the momentary impact of a benefit, is highly relevant to how great a benefit it is.  The accusation of "ageism" thus results from a failure to look beyond the immediate impact of a benefit.  When we see that a certain intervention actually offers a greater benefit to a younger person than it does to an older one, then there can no longer be any reasonable objection to it.  There is nothing objectionably "discriminatory" about prioritizing more-beneficial health interventions over less-beneficial ones.

The weight case is slightly different.  Here the issue is not that heavier people get less benefit from treatment, but just that they require more of the scarce resource in order to receive treatment.  This means that, if we want to help as many people as we can, without regard for their morally irrelevant characteristics, then we will naturally end up prioritizing lighter patients.  This is not because we count their interests for more.  It is just that they are easier to help -- we can help more of them for the same amount of resources.  And given a choice between helping more people or fewer, we should surely prefer to help more.

To think otherwise is, in fact, the objectionably discriminatory position.  If one wants to avoid the appearance of differential outcomes, then one must actually take morally irrelevant features into account -- to treat a single 300-pound person as morally equal to three 100-pound people, and hence to treat the former person's interests as three times as weighty as the interests of each of the lighter people.

This is not what the superficial egalitarians intend, of course -- they do not think of their preference for superficial equality as "discriminatory" in this way -- but it is the actual moral upshot of their approach to the issue.  It is deeply morally misguided.  We should not prioritize superficial equality over the more fundamental equality embodied in the utilitarian ideal of giving equal consideration to the interests of all, regardless of their superficial characteristics.


  1. Does it matter whether those who require more resources are morally responsible for their condition or not?

    Here's a real-world case to draw this out. Black people in the U.S. have been unjustly denied economic and educational resources. As a result, suppose (as I suspect is true) that if you take two equally naturally-talented and motivated people, one of whom is black and the other of whom is white, and admit them to the same college, on average, the black person will require more resources to achieve the same educational outcome. (Because the college will have to compensate for the terrible pre-secondary education the black kid received, etc.)

    Assuming college resources are limited, your argument seems to entail the alarming conclusion that we ought to have reverse affirmative-action, admitting more white kids who are cheaper to educate. But surely this can't be right?

  2. (I suppose you might say that the black kid will gain more overall benefit from the intervention, but let's assume that away for purposes of argument, and suppose that anyone who doesn't get admitted to college will have a life that goes very badly.)

  3. Paul - I don't think it matters whether they're "morally responsible for their condition or not". Rather, the question seems to be whether there are further downstream consequences to consider. One could easily accept everything I've said here, and yet argue for affirmation action on grounds particular to that case, e.g. the role-model argument, worries about the effects on society of systemic racial inequalities, etc.

    I'm not sure I follow your argument that this would lead to reverse affirmative action. College admissions committees don't have access to any tests for "natural talent" (DNA scans?). They presumably look at evidence of realized talent, e.g. test scores. Are you imagining a case where even blacks who achieve good test scores require disproportionate investment to succeed in college? I'm not sure why that would be, but sure, in that case, the fact that a college could educate more people if they only admitted people who happened to be white would suffice to show that such a policy isn't intrinsically unjust. Again, the question would instead be whether there are downstream consequences to consider (which, for race in America, there almost certainly are).

    1. Fair point wrt test performance. So lets make the case a little more artificial to chop off some of these features. Actually, lets go back to the original case from your post.

      Suppose the 300 pound people weigh that much because some
      malicious agent A imposed this weight on them (A force-fed them, say, or secretly put tons of butter in their food.)

      My intuition is that it matters, for the allocation decision, that those who require more of the scarce resource to get the same result are not morally responsible for their predicament.

      (If you disagree that this gives those who weigh 300 pounds some extra claim to the medicine, would it matter if A is either the person doing the distributing or one of the 100-pound people who would benefit from it?)

    2. It's better to help innocent agents than malicious ones, so A's actions might well disqualify him from receiving the medicine. But we should still prefer that three innocent people get the medicine rather than just one innocent person. This remains true even if A is doing the distributing. To help fewer people would just be to compound the wrongness of his earlier act -- two wrongs don't make a right, and all that.

      Could you clarify the sense in which you think moral responsibility "matters" here? That is, do you think that the innocence of a 300-pound person means that they should count for three times as much as an innocent 100-pound person? That would be odd. Or is it instead that you agree with my basic account, but would add on top of it that 300-pound people who are responsible for their weight should be even less likely to receive the medicine? (I guess I don't have any objection to the force-fed people having priority over others of equal weight who are responsible for their predicament. You might want to punish or disincentivize getting into conditions where people are more of a burden on scarce public resources. I just don't see how you could defensibly prioritize one innocent person over two or three equally innocent people.)

    3. I confess that I'm having trouble teasing out the exact idea that drives my divergent intuitions (which may be evidence that they should be rejected), but two ideas seem tempting.

      First, those who have greater resource needs, through no fault of their own, have some claim on common resources to satisfy them. And this is, I guess, just because one of the ways we ought to treat one another is with solicitude toward the unlucky ills of life. Maybe this is a virtue ethics sort of idea: it seems callous, miserly, a bit nasty to just tell someone "because you need more resources to reach the same level of well-being, we're going to give them to all these other people."

      The other is that there's a sense in which the victims of injustice can fairly claim a disproportionate share of goods just in virtue of their being victimized. This is the idea that drives things like reparations---the community as a whole (including innocent members of it) owe something to those who it has treated unjustly.

  4. Timmermann (see "The Individualist Lottery" in Analysis) would seemingly argue for procedural fairness in distribution by creating a lottery in which the heavy man receives a 25% chance of getting the cure and the other three a 75% chance.

    There is some intuitive appeal here, but I have always thought that this takes the concept of fairness too far. Why is it fair for three people to die so that one may live given that none have a greater claim to life than anyone else? Why would a flip of a coin or roll of a dice justify an obviously less good state of affairs coming to be rather than a better state of affairs? I have always thought of this as a version of self indulgence; the roll of the dice allows us to psychologically avoid blame for producing bad outcomes and thus produce the illusion of moral innocence in allowing worse states of affairs. Fairness is certainly important in many cases but not when choosing one innocent life over three.

    1. Indeed! I've never understood the appeal of weighted lotteries (except as a kind of psychological self-indulgence, as you note). I've previously discussed them a little bit here.

  5. Weight discrimination and its moral implication seem posed to create issues in the near future with numerous universal health care program on the horizon. the economist recently had a special report on obesity. Obesity is a leading risk factor of heart disease and other chronic illnesses. the fat among us are both less productive (use more sick days) and heavier medical burden than a skin person. Check out The Economist (Dec 15th-21th) for more information.

  6. This seems to me to sidestep the issue of discrimination entirely.

    Consider a scenario where weight is irrelevant and I am considering two possible methods of distributing a scarce drug:

    1) random lottery
    2) random lottery among those belonging to my favorite race, and if there is any left over a random lottery for my second most favorite race, and so on.

    Assuming that all lives are equally valuable, is there any difference between these two? The same number of lives are saved.

    I'm going to guess that you believe that there is a difference, which means there is some harm caused by #2 in addition to the harm to the lives of those who need treatment and can't get it -- perhaps it harms even those of my least favorite race who do not need treatment? But you seem to assume that no similar harm from discrimination exists in your scenario, that you do no harm to those who are heavy and don't need treatment by informing them that society takes care of them last? I find it difficult to imagine harms in my scenario #2 which would clearly not exist with your distribution method in your scenario.

    If discrimination causes some additional harm, then the question becomes whether that harm is greater than the harm of not saving some additional lives by treating the lightest, and while that's not a slam dunk for random distribution it's a bit more ambiguous than your argument suggests.

    1. Hi Eric, interesting objection. I think that what's objectionable about your #2 is that it involves a distinctive kind of disrespect (rather than harm). The discriminatory behaviour in this case seems to stem from a pernicious failure to regard everyone's interests as equally important -- how else to explain the differential treatment?

      But my scenario is very different from that. There's a moral reason for differential treatment (namely that this way we can help more people), and so the differential treatment need not constitute any kind of favouritism or disrespect.

    2. A couple of reactions...

      But what is the negative consequence in my scenario? Am I just an awful disrespectful purpose for favoring some over others for no reason, but there is otherwise no difference in the consequences if I go with #2? Or is there some consideration that should be given to the fact that some will feel disrespected as a result of #2? If the latter, does it make any difference that some heavy people may not be persuaded by your arguments on why they should not feel disrespected in your scenario?

      I am also wondering whether there is a limit to this. Would you make the same decision in a case where someone who weighed 200 lbs required only 1% more medicine than someone who weighed 100 lbs. It seems you should still pass out medicine from lightest to heaviest; you might save a few more lives that way. Is your position that if putting people in a priority order will result in even one just more life saved, it is a moral imperative, but if doing so does not save one more life in total then it is a moral imperative not to? What of the case where there is a 10% chance that you would save one additional life by prioritizing people by weight and a 90% chance that no additional lives would be saved?

    3. Depending on the details (whether the procedure is publicized, etc.) there may not be any negative consequences in your scenario. But if the patients do find out and so some "feel disrespected", I wonder whether it's significant that their feelings in this case (but arguably not in mine) would be genuinely warranted?

      If even unjustified feelings need to be taken into account, then we have worries about having to cater to mistaken morals.

      "What of the case where there is a 10% chance that you would save one additional life by prioritizing people by weight and a 90% chance that no additional lives would be saved?"

      Then it's like my case: there's a good moral reason to prioritize by weight and hence it isn't objectionably discriminatory. (Though policy-wise it might be more trouble than it's worth, especially dealing with the political backlash from everyone with mistaken morals.)


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.