Posted by: captainfalcon | January 1, 2013

Puzzle of Failed Compromises

Inspired by reading The Problem of Punishment (which is hard-nosed analytic philosophy, albeit on a pretty intellectually light topic) and for old time’s sake, a post in the analytic style.

Assume a scenario reminiscent of the current negotiations over the fiscal cliff.  (1) Assume two parties, A and B.  (2) Assume three options: A’s preferred option (pA), B’s preferred option (pB) and the status quo (sc).  (3) Assume that pA cannot be enacted without B’s support and pB cannot be enacted without A’s support.  (4)  Finally, assume that pA is better than pB which is better than sc.

Given (1)-(4), these two propositions seem true: (a) B should support pA.  (b) A should support pB if B does not support pA.

Now, if sc eventuates then both A and B have acted contrary to their obligations.  And they would seem to be equally to blame for bringing about sc because each had equal power to avert it.  So, it might seem, they are equally blameworthy.  This is the puzzle of failed compromises; at least in a bilateral situation, parties to the failure seem to be equally to blame for the failure, but we don’t think they are equally the blame.

I think this puzzle can be resolved by correcting a misidentification of what A and B are to blame for.  They are not both to blame for bringing about sc.  They are to blame for bringing about the difference in value between sc and what they could have brought about.  This means that A is to blame for the difference in value between sc and pB, and B is to blame for the difference in value between sc and pA.  B is thus more to blame because the difference in value between sc and pA is greater than the difference in value between sc and pB.

This same approach also entails that if B does not support pA but A supports pB then only B is to blame.  The options A could bring about were pB and sc; the options B could bring about were pA, pB or sc.  A brought about the best option A could and so is blameless.  B could’ve brought about a better option, and so is to blame for the difference in value between pA and pB.

Does this seem right?  Test it against a simplified representation of the fiscal cliff scenario.  Assume three options: the Democrats’ preferred reform (pD) which is preferable to the Republicans’ preferred reform (pR) which is preferable to the fiscal cliff (fc).  If the Republicans are spoilers would you blame them more than the Democrats if the Democrats were also spoilers (making them, together, total spoilers) and we went over the fiscal cliff?  If the Republicans are spoilers and the Democrats accept their proposal would you blame only the Republicans?

I don’t have any clear intuitions here, but that’s because my capacity for formulating other-regarding reactive attitudes — including blame — is underdeveloped because I have almost fully internalized Galen Strawson’s picture of life.



  1. Spoilar alert!

    Anyway, I believe you have stumbled into game theory my friend. You measure utility broadly (I think) as pA>pB>sc, but that preference ordering need not obtain for all parties. If both A and B act to maximize their preferences and sc still obtains (which is certainly a possibility given your set of presumptions, depending on the ordering of decisions/preferences), can either be considered blameworthy for the outcome?

  2. Some of the same elements that often appear in game theory — two players and conflicting preferences — also appear here, but the question that’s being asked is not the standard game-theoretic question of how a preference-maximizer should act but the very different question of how a moral evaluator should assign blame. There’s little overlap in these two inquiries.

    If moral blame is even a thing, I think both A and B are morally to blame for bringing about sc (though to different degrees, as my post argued) if (1) pA and pB are both better than sc (which I stipulated) and (2) A and B each have it in their power to bring about one of pA and pB (which I also stipulated). That’s because each could have brought about a better outcome, but instead brought about a worse outcome. This is so even if both have maximized their own preferences in bringing about the worse outcome. Either one can be morally blameworthy for acting in accordance with one’s preferences or pick your favorite mass murderer is not morally blameworthy (which may be true, but rather explodes the concept of moral blameworthiness).

    An alternative to questioning whether people can be morally blamed for bringing about worse results even though they acted in accordance with their preferences is to question whether there is any sense to be made of the idea that pA is better than pB if each are preferred by different parties; you might think all that can be said is that pA is better for A and pB is better for B. Note, though, that this conclusion is at odds with the view, which I think you favor, that what is valuable is what maximizes preference-satisfaction. Let’s say pA satisfies more preferences overall than pB, which satisfies more preferences overall than sc. (This was a view you were free to help yourself to — all I said was that pA was better than pB which was better than sc; you were free to fill in the theory of value that would make this true.) Wouldn’t you then say that pA is better than pB which is better than sc? (You aren’t forced to this view. You can abandon your equation of better results with results that maximize more preferences and revert to the relativistic view that there is no such thing as an objectively better or worse result; results are better or worse only from a particular person’s point of view. This is, in fact, my position. But, again, you’re doing nothing more than exploding the concept of moral blameworthiness; hardly responsive to a post that assumed the validity of the concept.)

    Finally, wouldn’t you agree that somebody can be blamed — maybe not morally blamed, but at least rationally blamed — for acting to maximize his preferences? Let’s say I would have lived a long and happy life but instead I formed a strong preference — stronger than any of my other preferences — to break my neck (just for the hell of it), which reduced me to a long life full of suffering. Say I spent the rest of my life regretting that choice. Isn’t it fair to say that (a) my regret is appropriate (I shouldn’t be thinking instead: well, I preference-maximized so no regrets!) and (b) I acted irrationally. Which is all to say that because preferences can be irrational, action to maximize preferences can also be irrational.

    I think one of the reasons you might be inclined to push back against this is that it cannot be economically (or game-theoretically) modeled. But that’s fine; there’s plenty of value to knowing what one should do given one’s preferences, so it’s not an existential threat to the enterprise of contemporary economics that it can be guilty of garbage-in/garbage-out.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: