Tuesday, October 13, 2009

A Solution to Moral Luck?

Tomorrow I teach Nagel's 'Moral Luck' essay. I wonder if the solution to resultant, and perhaps circumstantial luck, is easily solved by the concept of moral risk. Winning the lottery is lucky, but it is not pure luck. Merely finding a winning lottery ticket in your coat pocket is pure luck. Playing the lottery and winning is something else. It is a risk one takes--deliberately accepting the cost of a few dollars for the low possibility of many thousands. It strikes me as perfectly reasonable to say that one deserves whichever outcome, even though that result is out of one's control. The outcomes may be widely divergent (a $2 sunk cost or $50,000 on the Pick 4), and yet equally deserved. Similarly, if I choose to drive over the speed-limit, I am taking a moral risk--and deserve whichever outcome, however divergent (getting to work on time vs. vehicular manslaughter). Not only is this a solution to resultant moral luck, but I believe that it's a fair exposition of our intuitions on the matter. Right?


  1. Dear Michael,
    There is some ambiguity in your suggestion as between one deserves whichever outcome because one is more blameworthy if a harmful result occurs and one deserves whichever outcome even though one is no more blameworthy because of the occurrence of harm. If it is the latter that you meant, have a look at my 'Moral Luck and Liability Lotteries' (http://ssrn.com/abstract=1478216). If it is the former, a similar suggestion is made by Tony Honore in Responsibility and Fault and elaborated on by Gardner (John Gardner, 'Obligations and Outcomes in the Law of Torts' in Peter Cane and John Gardner (eds), Relating to Responsibility: Essays for Tony Honoré on his Eightieth Birthday (Hart Publishing, Oxford 2001)). Gardner goes as far as ascribing this view to Kant. I believe that this view is mistaken too and hope to refute it sometime soon, among other reasons, because it does not cope very well with cases where the outcome may be either good or bad.

  2. You and I have discussed this, but I thought I'd reiterate:

    I think you're making a mistake to link responsibility to the tort law conception of breach of duty. Legal duties are determined by the legal system. If your goal is to identify that point when you've 'taken a risk' rather than acting 'normally,' then you're caught in an infinite regress because risks are defined by their a posteriori tendency to lead to blameworthy harms or praiseworthy benefits.

    To my mind, it is preferable to adopt a non-blame conception of responsibility as attribution and then worry about moral and legal praise and blame only as the products of moral and legal institutions. So all the results of my actions are attributable to me, though we can only determine whether I've taken a negligent risk after the fact in comparing my actions (and perhaps motivations) to an artificial set of standards. This obviously raises a different set of questions around the apportionment of attribution or the anxiety of influence, but it's better to focus on that issue than try to handle moral luck alone and with that issue unresolved.

  3. Guy, thank you for the suggestion. As to the disjunction you pose, I take the former: one deserves whichever outcome, not BECAUSE one is more blameworthy--but as a RESULT, one is more blameworthy.

    In Nagel's article, he addresses moral risk, but in the following way: recognizing the results are not totally under my control, the responsible rational agent assesses risk-willingless and probabilities, and takes a 'rational' action--but this is all done BEFORE the action. I am arguing that hold off on assigning blame or praise until AFTER the consequences.

    JAM, I'm not opposed to your suggestion. But after the results have been tallied, when we are ready to classify some results as blameworthy or not, isn't a popular criterion(the way Nagel suggest) just going to be those results which are things you've done vs. those results which are things that just happen? Nagel gets confused in deciding how moral luck conflates the two. My suggestion is that, assuming moral risk, we count as 'things you've done' and therefore as 'things you're blameworthy for' those results which follow under the domain of moral risk. So, on my model, if you drink then drive, then hit a kid, that's something you've done, pace Nagel's model.

  4. But pace Williams, isn't the lorry driver responsible for having hit the kid even when he acted impeccably? Certainly, he has a different relationship to the act than a spectator does. *That's* the moral luck that worries me.