[EM] "Banzhaf Myth"

[Forest W Simmons]

Actually, in Mark Livingston's simulation at

http://www.cs.unc.edu/~livingst/Banzhaf/#results

he did assume that each state voted as a block.

Perhaps you meant that the small population states voted together as a 
block?

I can see where that is a possibility, since most of the small 
population states share (perceived) interests or rural populations.

However, if we are talking about list PR quotas, I'm not sure that you 
can argue that most of the small parties will have some common interest 
like ruralness.

It seems to me that we ignore Voting Power considerations at our own 
peril, since (as Livingston's simulation shows) they can easily swamp 
the fine tuning of PR.

Forest

Warren Smith wrote ...
>
>Forest Simmons resurrected a classic myth that small states have
>less voting power (even under Adams apportionment).
>
>I dispute that.
>
>It is true in the sense that Joe Voter in a small state has less voting 
power
>when filtered into the Fedral House using Banzhaf power at both stages 
of the filter.
>
>
>However, that is all assuming  everybody votes randomly.
>If you view the states as bloc votes, then the small states have
>MORE power than the big ones and in fact will win every such battle.
>
>
>The question is, which model is more correct?
>
>when it comes to big vs small issues, the votes generally
>ARE bloc.
>
>Also, history shows the small states have more power and consequently
>get more federal money.  There is a net outflow of money from
>large states to the feds, but a net inflow of money from the feds
>into small states.   What matters for that is what I might call "pork 
power."
>That is, games are played of the form "give me your vote and I'll
>pay you off with some pork."  The small states have more representatives
>per person, consequently they get more pork per person at the expense
>of the big states.
>
>The proof is in the pudding, not in eggheaded power-definitions.
>
>
>This same error was made in a paper by Hemaspaandra & Hemaspandra on
>apportionment and other issues, viewed by them as computational 
complexity.
>Long story short, I believe their paper was totally wrongheaded.
>
>It in fact made a negative contribution.
>
>Warren D Smith
>http://rangevoting.org
>
>