Friday, June 05, 2009

Manipulation and Rationality

Supposing you want to calm down a raging bull. Is it "manipulation" to remove the red flag from the bull's sight? You're certainly influencing how the bull will behave. But it seems that you're achieving this precisely by removing a source of manipulation (in some sense), and so enhancing the bull's own self-control.

Or, for a human case, consider how we might move the chocolates out of sight, to where they'll be less tempting. Cases like this seem importantly different from, say, brainwashing ourselves (or being 'manipulated' in any problematic sense). What's the relevant difference? Arguably: rationality-enhancing influences (e.g. that counteract prior biases) are innocuous, whereas problematic manipulation consists in influence that detracts from our rationality and self-control. (What about 'rational akrasia'? Perhaps we should understand 'self-control' here as consisting in authentic rather than deliberate-judgment-driven action.)

If this is right, then it seems we shouldn't consider Sunstein and Thaler's "nudges" to be problematically manipulative (or 'Orwellian'), at least if done right. It's surely true that our heuristics and biases can be exploited in manipulative fashion (cf. advertising). But that doesn't mean we have to just ignore our biases. Better to counteract or accommodate them, i.e. set things up so that our everyday heuristics will more often succeed in 'nudging' us in the right direction (by our own lights).

Of course, the "if done right" proviso is a big one. I haven't said anything here to argue against pragmatic libertarian "slippery slope" concerns. I just don't think there's anything inherently problematic with intentionally influencing choices in the modest ways Sunstein and Thaler describe (e.g. changing from opt-in to opt-out organ donation). Like the bull with the red flag, we are being constantly manipulated by our environment into making senseless decisions. We should welcome 'interference' that serves to mitigate the stupidity caused by our natural biases, enabling us to make more rational decisions instead.


  1. It seems unwise to have an account of manipulation that demands an underlying theory of rationality -- there are lots of cases where no such theory will be available.

    Take the nudge cases. Suppose I have a high discount factor for cigarettes -- if someone does some nudge-style tinkering to make the long-term cost more salient than the short-term pleasure, are they supporting my rationality or undermining it? It doesn't seem at all clear to me.

  2. Is it any clearer to you whether this is a case of manipulation? If not, then that would seem to support my account. Hard or unclear cases regarding the question of manipulation will coincide with hard/unclear questions of rational choice.

    (Of course, just because an answer is non-obvious doesn't mean that there isn't an answer.)

  3. I think this is correct. Obviously so. I wouldn't have anticipated someone calling 'nudges' problematically manipulative, especially since there's going to be a default one way or the other in the first place.

  4. Shouldn't we make such nudges as transparent as possible? I worry about the paternalism implied in the idea of overcoming inherent biases in favour of rationality.

  5. Thom - yet such concerns seem surprisingly common (I linked to an example in the post).

    ADHR - yes, certainly; I believe Sunstein and Thaler also insist on a simple ("one click") opt-out option, even in cases where we decide that people should be "in" (e.g. as organ donors) by default. So I don't see any implied paternalism here.

  6. Richard, it's true, it's not clear to me whether the nudge cigarette case is manipulation or not -- but isn't the job of a theory of X to help us figure out, where our intuitions are clear, whether some particular thing is an instance of X?

    So take, for example, the following claim:

    P is manipulation to the extent that P constitutes an attempt to influence someone's behavior without giving them reasons for behaving as desired.

    I take it that this plausible definition would give us an answer on the nudge sorts of questions...


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.)