[EM] Re: Electoral College

[Forest Simmons]

On Mon, 28 Jan 2002, Blake Cretney wrote:

> Jurij Toplak wrote:
> 
> >It seems strange to many people that the smallest states have 3 votes in
> >Electoral college although they should not have them according to the "one
> >person - one vote" rule. However, all the federal countries tend to give
> >small states some more power than they should have. There is a good reason
> >for that.
> >Consider federal country of Yugoslavia (the new one, FRY). It is a
> >federation of only two parts - one is large (about 90% of the population -
> >Serbia) and one is small (less than 10% of the population - Montenegro). Do
> >you think that there should be one-person one-vote rule when they are
> >selecting president? Then the Montenegro will never get their man in the
> >office. Also in the parliament Serbs will always outvote them. To give both
> >parts equal power is unfair, too. Then serbs would be unhappy. So something
> >in the middle is needed. And this is what Electoral college is doing, too.
> >
> I'm not sure I entirely disagree with you.  On the other hand, what's to 
> stop me from arguing that any arbitrary 10% of the population is 
> hopelessly outnumbered, and really should be over-represented, just to 
> be fair.  Maybe when there is some major ethnic/religious conflict the 
> division isn't arbitrary.  But the current set up of US states seems 
> like a historical accident.  Certain parts of the country formed large 
> colonies, others formed multiple small colonies.  The ones that formed 
> multiple small colonies are now over-represented.
> 
> I could look at the top 1/10th of California, and claim that this is 
> only a small fraction of the US, unlikely to have much influence unless 
> their power is made disproportionately high for their population.  But 
> wait, under the electoral college, this region is not only small, but is 
> under-represented based on their population (at least according to 
> conventional wisdom).
> 
> So, as I say, I actually think it might be desirable sometimes to give 
> an embattled minority more representation than their numbers deserve. 
>  But I wonder what criteria you would suggest to determine which small 
> sections of a country deserve over-representation, and which may be 
> justly under-represented.
> 
> ---
> Blake Cretney
> 
> 

Rather than give some groups super representation by fiat it seems that
using PR methods that tend to amalgamate the support of (otherwise
diffuse) natural interest groups (to use a phrase of Lani Guinier's) would
be preferable when sharp divisions along geographical boundaries are not
discernible. 

Some forms of PR do this better than do other forms.  PAV (Proportional
Approval Voting) certainly does this better than closed list PR, and seems
to do so better than STV;  PAV was specifically designed to reward
overlapping interest groups with extra chances of representation.

You might belong to a group too small to merit representation by itself,
but if some of the interests of your group overlap those of various other
groups (different interests, perhaps different groups) your vote may help
elect several (for you, partial) representatives.

Open list PR methods seem to be attempts in this direction. 


Forest