[EM] [Fwd: Possibly naive question on polarizing candidates]
Bart Ingles
bartman at netgate.net
Fri Oct 5 22:32:21 PDT 2001
Alexander Small wrote:
> Is there a quantitative measure for how polarizing a candidate is?
If there is, it would probably have to use more than the rankings shown
below. Consider the following two examples, both of which fit the
rankings profile you provide with your question:
> 9% B>C>A
> 51% A>B>C
> 40% C>B>A
Example 1:
Voter Rating or utility level
100 50 0
---------------------------
9% B C A
51% A B C
40% C B A
Example 2:
Voter Rating or utility level
100 50 0
---------------------------
9% B C A
51% A B C
40% C B A
In example 1, B is too popular to be very polarizing. He might in
example 2 though, since B is highly favored by approximately half the
electorate, and despised by the other half. The same is true for A, but
the voting constituencies are reversed.
I'm not sure how you'd quantify this, but a candidate strongly favored
by half the voters, and strongly despised by the other half, might be
one example of polarization. Or rather than half, he should be strongly
favored by a number of voters nearly equal to that of his nearest
competitor, with little or no overlap.
A candidate strongly despised by nearly all voters wouldn't be
polarizing, since the voters would be united in their dislike of the
candidate.
Welcome to the list, BTW.
Bart
> A is the Condorcet winner, but A is also quite polarizing, in that almost
> everybody either loves or hates A (likewise for C). There is universal
> agreement, however, that B is a reasonable choice (no last place votes).
> Perhaps A is really liberal, C is really conservative, and B is moderate.
>
> I know the list focuses on technical aspects of voting rather than
> ideologies. However, are there any technical criteria for quantifying
> just how polarizing a candidate is? It would be interesting to evaluate
> how various election methods work in the presence of polarizing
> candidates.
More information about the Election-Methods
mailing list