Friday, August 29, 2008

Selfish Irrationality

Via Chris Dillow, I discover a finance professor disputing the "strict classic economists [who] have argued that charity and other selfless acts do not make economic sense." Helping other people can be rational, he insists. Why? Apparently, because it will give you a warm glow. (He quotes evidence that "individuals report significantly greater happiness if they make charitable donations or give gifts to others rather than spending on themselves".)

The implication seems to be that if it weren't for the warm glow -- if we imagine someone who really wouldn't get any personal benefit or welfare boost from helping others -- then, upon seeing others in need, it would be "irrational" for this person to help them. See the neighbour's kid drowning? Fortunately, most of us would get enough of a warm glow to make saving their life rationally permissible. But for the others, they're apparently not thinking straight if they let such a distraction trump their personal desire to watch TV.

Actually, it gets worse. Most of us unthinkingly comply with the obligation not to murder. But if you one day find yourself in a position where a surreptitious poisoning could safely advance your interests (broadly construed: we're not just talking about money, as the prof. says, but 'warm glows' too), and you wouldn't feel any guilt about it, why, it must be downright "irrational" to let dispassionate moral considerations get in your way.

I take it nobody actually believes this. But it's strange, then, how many give lip service to the theory that implies it. Presumably they just haven't thought through the implications; but I have trouble seeing even a prima facie motivation for identifying rationality with self-interest. Why would you ever think that the only relevant considerations, in deciding how to act, are the consequences for oneself? If an action will affect other people too, it seems plain that this, too, might merit - or even require - consideration. But the above view implies that not only is such a blinkered assessment rationally permissible (already a dubious claim), but it's outright impermissible to consider anyone or anything besides oneself. How could a view this loopy be so widespread?


  1. I agree, it's a silly and erroneous model of human motives and preferences. I think that economists sometimes adopt it as a hack to simplify their models, e.g. the mathematics and empirics of welfare functions are much more tractable if people's preferences don't refer to anything outside themselves.

    This is also used as a way to smuggle in moral values at times. When economists ignore both positive and negative other-concerning preferences, they sometimes see themselves as enabling an ethic of treating each person's welfare as equally important. Otherwise, we might imagine a world where every member of an majority ethnic group prefers that other members of their group have their preferences satisfied, but prefers that 'enemy' group members have their preferences thwarted. Under those circumstances crude modeling might suggest that oppression of hated minorities was welfare-maximizing. Rather than make extended excursions into normative ethics, economists find it easier to just assume the problem away in their models.

  2. Actually I agree with the professor -- if we didn't feel good after doing altruistic acts our society would probably not consider them morally obligatory -- I don't see this as any different to Hume's idea of morality as based on sentiment.

    And yes, economically charity often doesn't make sense, but that's because there are non-economic considerations for most people, even die-hard economists.

  3. But note the difference between "not widely believed/considered to be morally obligatory" and "is in fact irrational". The prof. began with the question, "is it rational to give to charity?", so his later remarks about "economic considerations" (assuming he wasn't changing the subject) suggest that he doesn't appreciate that other considerations are relevant to rationality.

  4. Maybe I can suggest a way in which this view might not be so strange. Economists tend to have two notions of rationality, they have a formal (roughly game theoretic) notion of rationality and the normal informal notion of rationality. Its an open question as to how much these notions overlap but one of the things that economists do is to try and capture as much of what we informally consider to be rational with their formal notion. This is why the case of charitable acts is a puzzle, it is clearly (informally) rational but is tough to explain adequately how it is rational in a game theoretic sense. That is unless you start talking about a "warm glow". So saying that charitable acts are "irrational" is not to say that the action is stupid or should not be done, rather it is to make a statement about what actions we can explain with the formal tools at our disposal.

    And why is the formal notion of rationality approximately self-interest? Well, that's turned out to have the best predictive power, and there aren't really any competing notions from which we can build theories and predictions.

    Now I don't think that all economists are clear about this, some are very confused and conflate the two notions, but I think the talk of the finance professor can be rationalised.

  5. I would somewhat disagree with what Harjit said and develop it. The formal notion that economists support can itself be split in two. There is the more general version of the game theoretic (or decision theoretic) version: where one tries to maximize expected value. Then there is the specific version where value is the agent's welfare. The first of these is a good formalization of rationality. It might not capture everything, but it cuts the world at the joints and does service to our ordinary usage. It tells one how to get more of what one values. The additional assumption that the value involved in rationality consists solely of the agent's welfare is not a sensible part of a definition of rational. To make this kind of claim, we should say 'Charity is not rational for a purely self-interested agent' or something like that. When they smuggle self-interest into rationality, economists are making a very misleading and damaging mistake, and one whose repetition should be embarrassing to their profession.


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.