(Note: this post is largely inspired by two brilliant posts by Sister Y: Living the Epilogue: Social Policy as Palliative Care and The Mathematics of Misery: What Human Behavior Teaches Us about the Value of Life, and represents my attempt to build on them. If you haven’t read these posts yet, I strongly urge you to do so.)
By self-destructive behaviors let us understand a range of things usually thought of as sins and vices, which might are sources of pleasure in the short term might be the source of pleasure, but which in the longer term are likely to interfere with one’s functioning and hasten one’s death. We could all think of many: drug or alcohol abuse, promiscuous unprotected sex, gambling in some cases, even eating fatty food. For the sake of simplicity let’s call engaging in significant self-destructive behaviors live dirty and not engaging in them live clean. In addition to these two lifestyles, we also have the option of be dead.
Another term, which I’ll call a betterness relation is a way of designating how well different ways of living make one’s life go. In many cases (though I will not necessarily commit to all) this will be the same thing as a preference ordering. The simple greater-than symbol “>” will indicate that a given outcome described to the left of the symbol means things going better for someone than that described to the right. And I’ll assume, I hope uncontroversial, that betterness relations are transitive. So to take a trivial example, if you’d enjoy chocolate ice cream to vanilla and some ice cream to none at all, then
(eat chocolate ice cream) > (eat vanilla ice cream) > (eat nothing)
Betternesss relations can be different for different persons, obviously.
Now if we add a couple of simple assumptions about how people should make choices under conditions of risk, we can use a betternesss relation to construct a cardinal utility function among possible outcomes. For example if you could purchase a lottery ticket that gave you a probability of 0.2 of getting to eat chocolate ice cream and a probability of 0.8 of getting nothing, and you saw that lottery ticket as being just as good as getting vanilla ice cream with certainty, we could then assign numbers to the outcomes:
Eat chocolate ice cream = 10, Eat vanilla ice cream = 2, Eat nothing = 0
The numbers are arbitrary, up to a point (that of positive affine transformation, if I’ve got the math jargon right). (5,1,0) would work just as well, as would (36, 12, 6). The point is that a rational individual would try to maximize the sum.
Now in the minds of the happy-clappy, up-with-people optimists who appear to have a death grip over our public discourse and public policy on vices, to the best of my ability to enter into them, everyone has approximately the same utility function over (live dirty, live clean, be dead), and it seems to look something like this:
Live dirty = 100
Live clean = 99
Be dead = 0
So yeah, they might admit that it’s a little more fun to live dirty, but seriously, even a very small probability of that living dirty will result in being dead (in this example, any probability greater than 0.01) would mean that it’s rational to choose live clean over live dirty. (And you’ve all been hearing propaganda for this point of view since grade school at least. “Life is full of so many wonderful things that you will miss out on if you do drugs/get HIV, etc. Don’t you really want to have that wonderful life?”) Since plainly many people live dirty, the conclusion is that they must be irrational. They’re akratic, or myopic, or just plain ill-informed, and so Something Must Be Done in order to Force Them To Be Free. The available vices must be prohibited. People who engage in them must be forced into “rehabilitation.” And so on.
But it’s far from obvious that everyone has that sort of utility function. Some people plainly have a very hard time enjoying live clean. Their real utility functions might be something like this:
Live dirty = 100
Live clean = 50
Be dead = 0
In that case, people’s living dirty is simply a matter of rational risk taking, at least for live dirty behaviors that result in being dead with a probability of less that 0.5.
But of course there are still more profound possibilities. For some people life is really unpleasant. They might have underlying betterness relations of
(live dirty) > (be dead) > (live clean)
And if there are people like that, attempts to take away the live dirty option through some campaign of vice prohibition (even if we assume that such a thing can be successful) are going to result in a boatload of fail, because the response to having some mode of living dirty taken away is going to be either to be to find some other way of living dirty and, if that is not available and they are rational, committing suicide.
And there are still more radical possibilities. Some people may well be procrastinating suicidals. They are utterly miserable, but because of the fear of death, the absence of good means of committing suicide, and so forth, fail to do so. People like this would have a betterness relation
(be dead) > (live dirty) > (live clean)
I discussed in the most recent but one post to this one the possibility that there are people like this who are waiting for death. There are likely to be other people who, even if they can’t muster the will to take their own lives, might not be averse to hastening death. Alcoholism, drug abuse, and promiscuous unprotected sex are all good ways of doing this, of course.
We look out at the world and see rather a lot of alcoholism, drug abuse, gambling, risky sex, and so forth. How are we to understand all this? That it’s all just irrationality that needs to be addressed by do-gooders, somehow? Or does it suggest that there’s a lot of misery out there that fixing people’s rationality will do little to improve?
5 Responses to “Are self-destructive behaviors irrational?”
Sorry, the comment form is closed at this time.
I like this post a lot. I might have an objection, however, to getting from ordinal preferences to cardinal utilities. I doubt that people’s preferences are a reliable guide to what’s best for them, even assuming perfect rationality and prudence. Part of this has to do with I don’t think that the information available on which to develop preferences can ever be complete or neutral — we’re all misinformed in one way or another. Another part is that I’m not convinced that responding to the question “which do you prefer?” with “I prefer them equally” and responding to it with “I don’t know which I prefer” are equivalent. So even before you get to the lottery as a means of going from preferences to utilities, there are problems with preferences themselves.
If cardinal utilities aren’t real or determinable, or even if they are but they’re unknown, then it seems that rather than try to get people to choose among options differently, we should improve the options. If you don’t like people killing themselves, then rather than prohibit it or fix would-be suicides’ rationality, reduce the suffering they experience so they’ll be less likely to kill themselves.
People’s preferences aren’t infallible guides to what would make their lives go well (except perhaps under hypothetical conditions where they have perfect information). That said, I generally think that in most cases, the person who actually has to live a life probably has a shrewder estimate of what will make that life go well or badly than an outside observer. After all, who’s likely to try harder to get things right — the person who will suffer real harm if she gets it wrong, or someone who will suffer only intellectual embarrassment (if that) for getting it wrong? So as a general matter I’m inclined to respect people’s revealed preferences, subject perhaps to specific exceptions if those exceptions are reasonably evidence-based (and if the evidence on which they’re based is uncontaminated by pollyanna-ish biases, etc.)
I most definitely agree that it’s best to try to reduce people’s suffering and if in so doing there are fewer suicides, then well and good.
There’s a difference between someone being able to make a shrewd, or at least a shrewder, estimate of what’s will be for the best for them, and someone acting in accordance with this estimate such that from revealed preference we can infer it from their actions.
I don’t really think this has to owe to akrasia. As I understand people to use the term, akratic behavior is behavior contravening one’s best judgement. I suspect people make judgements much less frequently than they recall in narratives, and much less frequently than is generally assumed by social scientists. In other words, people don’t form shrewd estimates of what’s best for themselves, even with perfect information available, because we just don’t form estimates all that often. And acting on whatever estimate we come up with is another matter.
Sister Y and I had a little back-and-forth about this in the comments to her Mathematics of Misery post. There’s a lot of writing so I’ll just link to the most relevant comments (me me her).
You haven’t included (live clean) > (live dirty) possibilities.
Pleasurable but risky-to-life activities are sometimes so strongly condemned socially that the may no longer be pleasurable, really.
And even if we take away social condemnation, there’s our own “conscience” too! What’s more, obesity from fatty food and cancer from smoking don’t result in quicker death alone, and could be extremely painful all by themselves — especially when euthanasia isn’t available even after terminal cancer.
I think I operate with (be dead) > (live clean) > (live dirty).
The fact that there are so many people willing to live dirty in spite of prohibitions strongly suggests to me that living dirty has a lot of utility for them.
What’s more, obesity from fatty food and cancer from smoking don’t result in quicker death alone, and could be extremely painful all by themselves — especially when euthanasia isn’t available even after terminal cancer.
True, but the debilities of old age are no picnic either (add to which the fact that many elderly people are terribly lonely and pretty much abandoned as “useless” by our society), and what is more, at the end a long life you still face a strong possibility of dying of something prolonged and nasty.
I think I operate with (be dead) > (live clean) > (live dirty).
But you are still with us, clearly. I would like to hear your own account — do you regard yourself as a procrastinating suicidal, or as something else?