[EM] Linked issues, 1-dimensional political spectrum

[Stephane Rouillon]

I agree with Scott,

The image we get of the debates is a 1-dimensional projection
of a multi-dimensional opinions set.

One first consequence is that any set of preference can be totally sincere.
On the contrary of what Mike said in some examples,
there is no need to consider electors as betrayors when they
rank in a reverse order candidates, compared to the reference
order used by a majority. This can easily be explained in real life
with two different debates. The majority of electors decide to order
candidates according to a first debate (War participation in another country for
example):
order: A-B-C-D-E-F. On another minor subject (minor because of the number of
electors that make of it
a priority), the ordered set becomes A-C-D-E-F-B. (It could be about gay weddings
or nuclear plants or whatever)
And so on with as much stakes you want as another electorate subset priority.
Thus any set of ranking can be sincere and plausible.
Only the electoral system forces people to compromise along a weighted reference
axis.

Scott Ritchie a écrit :

> On Fri, 2007-02-23 at 22:49 +0000, Michael Ossipoff wrote:
> > One thing that I forgot to mention when I posted before about this: In the
> > U.S. at least, the issues are strongly linked. People's opinions on them are
> > strongly correlated. That's because an authoritarian is going to be
> > authoritarian on everything, typically.
> >
> > So a 1-dimensional issue-space isn't a bad approximation. With that
> > reasonable approximation, the CW is the SU maximizer.
> >
> > Mike Ossipoff
>
> Of course, the political scientist and voting theorist in me says that
> politics are largely 1-dimensional exactly because of the voting system
> we use.
>
> Oh well,
> Scott Ritchie
>
> ----
> election-methods mailing list - see http://electorama.com/em for list info