[PEN-L:28052] Re: On median voters and minority rights

[Gil Skillman <gskillman@xxxxxxxxxxxxxxxxx>]

Where I wrote

>This is way off the mark. I'm not referring to any "model" of voting,
silly, neoclassical, or otherwise, other than to note that if a majority
rule choice obtains, then the preferences of the voter with "median"
preferences among the available options will generally be satisfied. This
is obvious in the case of a vote between two choices (e.g., Pepsi vs. Coke,
the case we were talking about), since in any majority vote the median
voter *must* be on the winning side. Of course, with more extensive sets of
alternatives a majority choice may not exist in the first place.


> This reminds me of a former colleague of mine who objected to my use of
> the phrase "the theory of comparative advantage." It's not a theory, he
> would say (and probably still does), since it's true by definition (or
> something like that). When this kind of rhetoric comes up, it's time to
> cut the cards (and count them, too), just as when someone says "it's
> self-evident that..."
>
> Actually, the median voter rule _is_ a model or part of a model of
> majority voting, in that it assumes that the spectrum of voters'
> preferences is single-peaked (since otherwise the equilibrium can be unstable).

Rather, the "equilibrium" might not exist in the first place, as opposed to
being "unstable".  I already addressed this possibility in my final
statement above (which Jim elides from his reply.)  But *if* a
majority-rule "equilibrium" existed--that is, an unambiguous choice by the
majority, with no cycles of vote-switching-- that choice would correspond
to the wishes of some median voter. (On this point see Jim's own comment
below about the practical interpretation of the meaning of the phrase
"median voter," which is the sense in which I originally used the
term.)  Such a choice always exists (if we include the possibility of
indifference) in the two-alternative case of "Coke vs. Pepsi" we were
talking about to begin with (that's why Arrow posits the existence of at
least *3* alternatives in his theorem).  If people are willing to make a
choice between Pepsi and Coke, then preferences are as "single-peaked" as
they need to be.

>  Unlike voting under actually-existing capitalism, it's assumed that each
> voter has equal power,  so that the "median voter" is exactly half-way
> down the list of voters -- rather than corresponding to half-way down a
> list where the preferences of some (rich) individuals are, in effect,
> counted many times. Perhaps the median-voter model would apply better
> with socialism than with capitalism.

Right, that's the hypothetical case we were talking about to begin with, in
a democratic socialist regime.  We weren't talking about the
pseudo-democracy that arises under capitalism, so I don't know why Jim
brings this up here.

> However, in line with the orthodox economist's gut-level and _a priori_
> assumption of methodological individualism, it assumes that the only
> interaction between the voters is in the voting itself, so that the act of
> voters _talking_ to each other in order to convince each other is
> forgotten. That is, in the model, I am not allowed to convince Joanna that
> her preferences are wrong, while she can't change her preferences. This
> fits with the timelessness of the standard model: Her preferences at time
> t may not be "satisfied" even though they may be satisfied at time t+1,
> when the vote actually takes place.

And since I wasn't referring to any "model" in the first place, I wasn't
ruling out this possibility, nor was I assuming a static or one-shot world
in which people can't talk before voting.  Again, Jim is imposing this
"model" interpretation--I wasn't.


> Further, the identity of the median voter would likely change, as issues
> and preference distributions change over time, so that Gil becomes the
> median voter in time t+2.

Perhaps, and nothing I said ruled out this possibility, since I never
suggested a world in which the identity of the median voter was frozen.

>  (Horrors!) For example, if anyone decides not to vote or non-voters
> suddenly start voting, that also would also change the nature of the
> median voter.

Great.  Same comment as above.


>  This point implies that since the entire set of voters determines who
> the median voter is, saying "the median voter decides" is pretty much the
> same as saying "the majority decides," adding little or no content.

Right. I agree. It's *Jim* who's been making this big, and now evidently
pointless by his own estimation, deal about the meaning of the phrase
"median voter."  I originally meant it in the sense he now
suggests:  suppose the majority insisted on Pepsi when you favored
Coke--would that be an acceptable instance of eliminating "fake variety"?

> Joanna _per se_ is forgotten, so she no longer seems like some sort of
> dictatorial arbiter of taste. This point is reinforced if preferences are
> distributed in a single-peaked way, since there would be a large number of
> people with preferences similar to the current median voter's.

Right again.  The only reason I mentioned Joanna at all in the first place
is because she voiced the opinion that there was no difference between Coke
and Pepsi, which claim Jim contradicted.  The issue was, would a minority
(say Jim) find it acceptable if their choice was limited because "a large
number of people with preferences similar to the current median voter's"
saw no difference where the minority saw a big difference?

> So the answer is: sure, I'm in favor of majority rule as the basic
> political principle, compared to the alternatives, especially since
> majorities are in favor of giving themselves minority rights.

I don't think that's what was at issue.  The issue was whether one could
expect democratic voting in a hypothetical socialist world to distinguish
"true variety" from "fake variety," and my point, which should now be
evident, is that this was problematic--in part because what might
constitute "fake variety" for one group might constitute "true variety" for
another.

> > This is obvious in the case of a vote between two choices (e.g., Pepsi
> vs. Coke, the case we were talking about), since in any majority vote the
> median voter *must* be on the winning side.  Of course, with more
> extensive sets of alternatives a majority choice may not exist in the
> first place.<
>
> Frankly, I can't see why anyone would have to vote on Pepsi vs. Coke. Most
> people -- the majority -- would probably want more variety than the
> actually-existing "free market" provides.  Most people seem prefer to have
> a variety of choices available and would thus likely to be willing to
> accept the cost of a little bit of fake variety. Perhaps Joanna thinks
> that fake variety is a big problem (though I doubt it), but most people on
> the left think of poverty, imperialism, exploitation, sexism, racism,
> environmental destruction, cruelty to animals, etc. as much more important
> problems.

No doubt, but now by Jim's own representation we've gone entirely beyond
Joanna's post (which introduced the "Pepsi vs. Coke" choice), and my
response, that prompted this exchange in the first place.  Below is the
exchange that led to this discussion with Jim, and as you can see it makes
no reference one way or other to the problems of poverty, imperialism,
exploitation, sexism, racism, etc. etc.  Jim has in essence changed the
entire point of our discussion.


> At 01:22 PM 07/10/2002 -0700, you wrote:
> > Enormous amounts of resources are spent to market products that are
> > essentially identical.  Maybe you don't get phone calls from the phone
> > companies ....
>
> Say whaaaaat!!! You don't think there's a difference between Pepsi and
> Coke? .....not to mention the difference between Microsoft and Microsoft?
>
> I'm shocked! Shocked!!!

>The problem with this line of argument is that for every seemingly
trivial example of product differentiation one can come up with, you can
also point to product differences that seem trivial to the average
consumer/would-be voter and yet are critical to some.  Some people find no
difference between oysters and mussels; others are deeply allergic only to
mussels.  Aspirin is fine for most people but gives some bleeding
ulcers.  The sugar substitute aspartame is a boon to some--particularly
diabetics--but causes migraine-like headaches in others. Personally, I've
found only one shaving cream that doesn't make me sneeze.  Etc., etc.  In
light of such cases it is perhaps less obvious what products count as
"essentially identical,"  and maybe even less obvious that all decisions on
this score should be made collectively.

Gil

> I wrote:>>In any event, democracy does not mean simply "majority rule." It
> also involves "minority rights." People value rights for themselves as
> individuals, so they are willing to grant rights to others... In fact,
> they might decide to allow markets under certain circumstances.<<
>
>
> >Sure, but to the extent that "minority rights" are guaranteed, there is
> correspondingly less assurance that what you've called "fake  variety"
> will be excluded, since there is no way to ensure that "minorities"
> wouldn't exercise their rights by demanding alternatives that the
> majority would consider trivially different. To use one of your examples
> [which, in this context, was somewhat flippant], suppose a minority
> insists that, for them, Dodges are strictly better than Plymouths; what then?<
>
> Since people have different preferences on most items of individual
> consumption, I doubt that the majority would decide that only one kind of
> car would be produced. Why should they, unless there's a clear and present
> benefit to doing so? If there economies of scale that impose sameness,
> that affects markets just as it affects democracy. But unlike in markets,
> democracy doesn't have other reasons why Dodges and Plymouths should be
> produced by the same company (Daimler-Chrysler) and therefore end up being
> the same beneath their tinny skins. In real-world markets, a company with
> the right brand reputation, marketing & distribution & repair facilities,
> connections in the financial sector (and a good credit rating), and
> political pull is likely to survive the battle of competition,
> concentrating and centralizing its capital, so that two almost-identical
> cars would be produced. (In case anyone has forgotten, Chrysler was able
> to survive to later become part of Daimler because of its political
> influence.) A centralized firm can afford to diversify its portfolio of
> cars to gain market power, whereas a more decentralized system produces
> variety naturally (not as part of a strategy). Now it's quite possible
> that a democracy might make a mistake and would limit the number of models
> available (perhaps in imitation of Henry Ford). But when economies of
> marketing, distribution, finance, and political pull are less important,
> there are fewer constraints than seen in real-world markets.


Right.  And all of these considerations reinforce my point that it's not
obvious how democratic choice (interpreted as including minority rights)
will limit "fake variety" while promoting "real variety."

> BTW, referring back to my list of more important issues, I doubt that a
> democratic movement concerned with the on-going enviromental mess would be
> more concerned with limiting the role of autos in society _period_ than
> with whether to abolish Plymouths and/or Dodges.


> I had written:>>Also, my point was (and is) that _all_ societies put
> limits on what kind of products are available to consumers. The question
> is how this decision should be made: should it be made by a
> self-perpetuating state bureaucracy? by a class of hereditary nobles? by a
> self-perpetuating class of rich people? or should it be made democratically?<<
>
> Gil:> As stated in this general and abstract form, of course the latter.<
>
> then we agree, since I was meaning to be talking about such things on a
> high level of abstraction.
>
> > But things get more difficult when the meaning of "democracy" is tested
> in concrete cases.  For example, suppose that, under the  socialist
> regime of your choice, a super-majority votes to return to a capitalist
> system. Would you regard this as an adequate warrant for re-establishing
> capitalism?<
>
> Talk about high level of abstraction! But let's try an experiment -- i.e.,
> set up socialism -- and see how it works out.
>
> If a super-majority really decides in favor of capitalism rather than
> democracy, it's got the right to do so. To have some minority save
> socialism via dictatorship or manipulative tactics would undermine
> democracy -- and thus socialism -- too. Clearly, the system is in trouble.
> I think it would be better to look at the objective conditions that put
> this abstract socialism between the devil and the deep blue sea in the
> first place.
>
> However, the point is, again, that democracy is more than voting: people
> would be discussing this possible transition for a long time before any
> kind of constitutional convention occurred. I could try to convince others
> to stick with socialism, while they could try to convince me. (Heck, they
> might succeed!) Further, of course, a well-operating socialist system
> would encourage people to change their preferences and expectations, so
> there would never be a super-majority of voters in favor of abolishing it.
>
> We should note in closing that capitalism would never allow a vote to
> decide whether or not to abolish _it_.

We could note it, I guess, but this observation, as with most of what Jim
has introduced above, was not at issue in the first place.

Let's move on.  This discussion seems to have gone beyond pointless.

Gil