[EM] Unranked IRV versus Approval - divergent winners exist!

Martin Harper mcnh2 at cam.ac.uk
Sun Apr 1 17:55:43 PDT 2001

[some snippage: apologies if I've mis-snipped, but I wanted to keep some vague
attempt at focus]

Tom Ruen wrote:

> Plurality among 3 is like a game of chicken between toy cars and real cars,
> except the toy cars can afford to crash while the big ones often manage to
> get wrecked too!

> Runoffs (instant or not) take away that game until the toy candidates grow
> into viable options. Then the game of chicken returns

So only big cars will crash? ;-) Seems like your game of chicken is actually
going to be worse under IRV...

> and everyone has a chance to vote as they like, whether compromising or not,
> and finally get
> the strongest of the biggest two coalitions.

Well: the strongest of two of the coalitions, selected by the weirdness of the
runoff process, which may or may not be the biggest.

> It is a power game and king of the hill wins.

Umm - this would be true in all election methods?

> Many people into politics love the top-runoff game, the excitement!

So advocate Random Ballot. I didn't think the point of elections was to give
multiple orgasms to politics enthusiasts....

> Plurality and runoffs both support 2 dominant parties.

Dictatorship supports one, and Condorcet and Approval support as many as you
need. Two is better than one, but surely lots would be better than two?

> [it's a game of compromise]

But not a good game of compromise. If the best minimum wage is $4, the two
dominant parties will stick on $3 and $5 - they won't compromise any more
because they'll lose turnout, and set themselves up for defeat via spoiler
candidates. Over the course of time the minimum wage will flip-flop over the
correct point, without ever getting to it.

> We can say that better methods exist than runoffs, but I'm not prepared to
> say that runoff really produce bad winners...

But why should we use any method except the best?

> Runoffs produce winners through
> coalition power, and coalitions are made from people who care to participate
> in the process more.

"participate in the process" == "make deals in smoke filled rooms"?

I'm old-fashioned, but surely the people who should decide which issues are
important, and which should be compromised upon, should be left up to the

> the process strengthens all the parties as they strive to stand
> for things and be inclusive enough to win.

And why exactly would this not happen in any other method?

More information about the Election-Methods mailing list