MPSA? Cool!
A few days ago my friend, Terrence Watson, was told that our paper, entitled "Conventional Stability: Spontaneous Order v. Collective Resolution" was accepted at the Midwest Political Science Association Conference to be held in Chicago in April.
I'm pretty excited, since this is about the most "serious" conference I've ever been accepted at. The two of us plan to rent a car and make a nice road trip out of it. That's going to be fun, I bet. I hope to meet at least a few Political Science Journal editors at the conference and convince at least one of them that the paper is worthy of publication in a scholarly journal. Since it hasn't been updated for a really long time, the paper needs some work before it'll be in the right shape for that. But it will be.
Terrence and I argue that norms or conventions that emerge spontaneously (as a consequence of human action but not design) are more likely to be stable than any order "imposed" by collective methods. The most obvious collective method is legislation, or some other governmental edict or diktat. Conventions, reflecting this spontaneity, should be given some kind of prima facie respect, even if we don't like them much. Their stability, and the way they "capture" information that can't be had by any one of us or group of us, are reasons for respecting them. So, too, is a secondary argument we offer in the paper--we think that conventions not only capture additional information and help maintain stability, we also believe that they do something remarkable. Namely, they reflect underlying utility patterns.
This second bit may be more controversial, but the case runs roughly like this. For anything to emerge out of the interactions of various individuals into an overarching pattern, it will have to meet with (at least) the acquiescence of a large-enough set of individuals over the domain in question. While acquiescence is not agreement (and therefore is lower with respect to normative importance), it still recognizes an important element in our interactions. This is given the alternatives actually available we choose to go along with it, rather than try to disrupt or offer an alternative norm/convention. We think this means that there is at least a good reason to suspect that norms/conventions operate by countenancing at least a preferred outcome to those on the table by some relevant group of us, over some significant foreseeable period of time.
This is not to say that alternative equilibria (understood synonymously with norms/conventions) are not preferred, or better than the given norm/convention. Game theory is good at telling us why this or that pattern of behaviour is in equilibrium, but it can't tell us why this equilibrium, as opposed to the countless (infinite?) alternative possible equilibria arose. It can only tell us that some given equilibrium is stable on the grounds that each of the actions are "best responses" to what the others are doing (and this is reciprocal: my action is a best response to yours, and yours is a best response to mine). Some set of those alternative equilibria may be better at maximizing utility, recognizing our preferred moral norms, ensuring liberty, or whatever else may be our most ideal outcome, but, and this is humbling, we can't know that it would be an equilibrium!
We just don't have that kind of information.
What's more, we can't know that each stage toward our most ideal outcome is itself a stable equilibrium. We might do the impossible and see that some alternative arrangement is, in fact, an equilibrium (suppose we knew this, even though I, and I'm betting Terrence agrees, doubt that this is an epistemic possibility), but to move from the status quo to that equilibrium will require intermediate steps. Any one of those steps may be out of equilibrium, exposing it to rational free riders, or to some other process that may turn it all upside down.
This is the neat thing about Ken Binmore's work in political philosophy. He recognizes this limitation in our system-building propensities. He likes Rawls, for instance, but he doesn't abstract away from the process that gets us to the ideal outcome we want after our venture behind the veil of ignorance and then the 'working out' of those conclusions with our intuitions (reflective equilibrium). We can't say, "this is our ideal world, now let's go get it!" Instead, we have to be aware of the limitations both in our information with respect to the outcome (whether or not it will be a stable equilibrium) as well as the information with respect to the processes (whether or not each move toward the ideal is in equilibrium).
Think of it like a game of moral Risk. You see Australia there before you, and you could take it. You think you have enough armies to just barely get it. But you have to calculate the response of the other players. Supposing you could get it, that will motivate some other players, if they are within range, to immediately snatch some portion of Australia away from you. The goal can only be attained with sufficient accounting for the (likely) responses of other players, coupled with the information about the probabilities contained in the role of the dice. We can't just drive for Australia regardless of process, because the outcome may be much worse, even if things would be peachier if we had control of Australia.
The tricky bit is that the actual game of morals (to steal from David Gauthier) includes different players with different conceptions and goals. Similarly, different people may have different tolerances for certain outcomes, some are more risky than others, and still others are willing to pay much more for one outcome rather than another. Conventions, I think we'll end up arguing, do quite well at balancing all of these different spinning plates simultaneously, adjusting here and there as the situations warrant.
If all of this sounds pretty conservative to you, that's because it is. But it's conservative in the Hayekian sense. And the more Hayekian, the better.
I'm pretty excited, since this is about the most "serious" conference I've ever been accepted at. The two of us plan to rent a car and make a nice road trip out of it. That's going to be fun, I bet. I hope to meet at least a few Political Science Journal editors at the conference and convince at least one of them that the paper is worthy of publication in a scholarly journal. Since it hasn't been updated for a really long time, the paper needs some work before it'll be in the right shape for that. But it will be.
Terrence and I argue that norms or conventions that emerge spontaneously (as a consequence of human action but not design) are more likely to be stable than any order "imposed" by collective methods. The most obvious collective method is legislation, or some other governmental edict or diktat. Conventions, reflecting this spontaneity, should be given some kind of prima facie respect, even if we don't like them much. Their stability, and the way they "capture" information that can't be had by any one of us or group of us, are reasons for respecting them. So, too, is a secondary argument we offer in the paper--we think that conventions not only capture additional information and help maintain stability, we also believe that they do something remarkable. Namely, they reflect underlying utility patterns.
This second bit may be more controversial, but the case runs roughly like this. For anything to emerge out of the interactions of various individuals into an overarching pattern, it will have to meet with (at least) the acquiescence of a large-enough set of individuals over the domain in question. While acquiescence is not agreement (and therefore is lower with respect to normative importance), it still recognizes an important element in our interactions. This is given the alternatives actually available we choose to go along with it, rather than try to disrupt or offer an alternative norm/convention. We think this means that there is at least a good reason to suspect that norms/conventions operate by countenancing at least a preferred outcome to those on the table by some relevant group of us, over some significant foreseeable period of time.
This is not to say that alternative equilibria (understood synonymously with norms/conventions) are not preferred, or better than the given norm/convention. Game theory is good at telling us why this or that pattern of behaviour is in equilibrium, but it can't tell us why this equilibrium, as opposed to the countless (infinite?) alternative possible equilibria arose. It can only tell us that some given equilibrium is stable on the grounds that each of the actions are "best responses" to what the others are doing (and this is reciprocal: my action is a best response to yours, and yours is a best response to mine). Some set of those alternative equilibria may be better at maximizing utility, recognizing our preferred moral norms, ensuring liberty, or whatever else may be our most ideal outcome, but, and this is humbling, we can't know that it would be an equilibrium!
We just don't have that kind of information.
What's more, we can't know that each stage toward our most ideal outcome is itself a stable equilibrium. We might do the impossible and see that some alternative arrangement is, in fact, an equilibrium (suppose we knew this, even though I, and I'm betting Terrence agrees, doubt that this is an epistemic possibility), but to move from the status quo to that equilibrium will require intermediate steps. Any one of those steps may be out of equilibrium, exposing it to rational free riders, or to some other process that may turn it all upside down.
This is the neat thing about Ken Binmore's work in political philosophy. He recognizes this limitation in our system-building propensities. He likes Rawls, for instance, but he doesn't abstract away from the process that gets us to the ideal outcome we want after our venture behind the veil of ignorance and then the 'working out' of those conclusions with our intuitions (reflective equilibrium). We can't say, "this is our ideal world, now let's go get it!" Instead, we have to be aware of the limitations both in our information with respect to the outcome (whether or not it will be a stable equilibrium) as well as the information with respect to the processes (whether or not each move toward the ideal is in equilibrium).
Think of it like a game of moral Risk. You see Australia there before you, and you could take it. You think you have enough armies to just barely get it. But you have to calculate the response of the other players. Supposing you could get it, that will motivate some other players, if they are within range, to immediately snatch some portion of Australia away from you. The goal can only be attained with sufficient accounting for the (likely) responses of other players, coupled with the information about the probabilities contained in the role of the dice. We can't just drive for Australia regardless of process, because the outcome may be much worse, even if things would be peachier if we had control of Australia.
The tricky bit is that the actual game of morals (to steal from David Gauthier) includes different players with different conceptions and goals. Similarly, different people may have different tolerances for certain outcomes, some are more risky than others, and still others are willing to pay much more for one outcome rather than another. Conventions, I think we'll end up arguing, do quite well at balancing all of these different spinning plates simultaneously, adjusting here and there as the situations warrant.
If all of this sounds pretty conservative to you, that's because it is. But it's conservative in the Hayekian sense. And the more Hayekian, the better.
posted by P. M. Jaworski at 10:12 p.m.
0 Comments:
Post a Comment
<< Home