[PEN-L:28115] RE: The market as a preference aggregator?

["Devine, James" <jdevine@xxxxxxx>]

Title: RE: [PEN-L:27932] The market as a preference aggregator?

Gil wrote: > I agree with Justin that it's a bit of a stretch to think of "the market"  as a mechanism for aggregating individual preferences into a "social  preference ordering."  It's more appropriate to think of the market as a mechanism of social *choice*, i.e. as something that selects *particular* outcomes given particular initial conditions (including individual preferences), rather than something that yields a "social preference ranking" based on individual preferences.  The difference, plainly put, is that social preference orderings have to be combined with social constraint sets--the set of what's socially feasible at any given historical moment--in order to yield actual outcomes. To treat the market as a social preference ordering, for example, you'd have to read "social preference" from a given pairwise comparison as "society prefers the allocation that constitutes a market equilibrium over the one that doesn't, and if they both constitute market equilibria, society is indifferent."  Not clear that that makes any sense.<

maybe it doesn't make sense, but this is the way much or most of the _laissez-faire_ school sees it: only the market, in their view, can be the judge of the worth of any item or activity.

>
> But supposing that the market is understood as a social preference
> aggregator, then I agree with Gar that Arrow's theorem would
> apply.  I
> understand Arrow's impossibility theorem a little
> differently, though--as I
> read it, it states that there is no coherent (i.e., complete and
> transitive) social preference ordering over choice sets with
> at least three
> alternatives that simultaneously satisfies
>
> (U)  Universal domain:  any possible array of (coherent) individual
> preference orderings is permissible;
>
> (P)  Pareto principle:  if all individuals (weakly) prefer any some
> allocation A over some other allocation B, then the social
> ordering will
> also reflect this (weak) preference;
>
> (I)  Independence of "irrelevant" alternatives:  the social
> ranking of any
> two feasible allocations does not depend on what other
> allocations are
> included in the choice set;
>
> (N)  Nondictatorship:  the social ordering will not simply
> reflect the
> preferences of any single individual.
>
> The most obvious way in which the market mechanism,
> understood in the above
> sense, would fail as a social welfare function is with
> respect to (U) ,
> since it will not generally be true that for any two
> allocations, at least
> one will constitute a market equilibrium (however that might be
> defined--perfectly competitive or otherwise, e.g.).  Fulfillment of
> condition (I) is also problematic given the possibility of
> income effects
> on individual choices.
>
> On the whole, it's probably more accurate and less confusing
> to say that
> markets don't "aggregate" preferences, but rather translate
> given arrays of
> individual preferences into social outcomes.
>
> Gil
>
>
>
>
>
>
> >>>people should realize that Arrow's theory is a critique of
> _all_ collective
> >>>decision-making mechanisms, not just democracy. It also
> applies to markets.
> >>>Can you think of a method of collective choice that isn't
> subject to the
> >>>theorem?
> >>
> >>Um, how so? The theorem says you can't have: nondictatorship,
> >>independence of irrelevant alternatives, independence of
> order of choice,
> >>and, dammit, one other thing I can't recall, all together.
> It's a theorem
> >>in voting theory. These things are irrelevant to markets.
> Of course a
> >>market with dictatorship (a monopoly) is distorted, but the
> independence
> >>conditions simply don't matter to formulating a market model.
> >
> >
> >
> >
> > >    * Universality. The voting method should provide a
> complete ranking
> > of all alternatives from any set of individual preference ballots.
> > >    * Monotonicity criterion. If one set of preference ballots
> > preference ballots would lead to an an overall ranking of
> alternative X
> > above alternative Y and if some preference ballots are
> changed in such a
> > way that the only alternative that has a higher ranking on
> any preference
> > ballots is X, then the method should still rank X above Y.
> > >    * Criterion of independence of irrelevant
> alternatives. If one set
> > of preference ballots would lead to an an overall ranking
> of alternative
> > X above alternative Y and if some preference ballots are
> changed without
> > changing the relative rank of X and Y, then the method
> should still rank
> > X above Y.
> > >    * Citizen Sovereignty. Every possible ranking of
> alternatives can be
> > achieved from some set of individual preference ballots.
> > >    * Non-dictatorship. There should not be one specific
> voter whose
> > preference ballot is always adopted
> >
> >Note that "voting" in this context is simply a means of aggreating
> >individual preference into social choice. A "market" in
> which everyone has
> >the same number of dollars would be a democratic vote by
> this criteria -
> >one in which multiple choices are made by multiple voters.
> >
> >Note also that uneven distribution of money , short of
> severe monopoly
> >does not resove the paradox, any more than uneven
> distribution of voting
> >powers (where some people were allowed to vote multiple times) would
> >resolve the paradox in any other election - short of dictatorship.
> >
>