[EM] Reply re: CVD. The Heritage Criterion.

[MIKE OSSIPOFF]

Alexander Small asked:

Do people think it's worthwhile to nonetheless pay the $15 to become a
member of CVD and try to persuade them to change? This is akin to when the
NAACP director bought stock in the TV networks so he could go to
shareholder meetings and lobby for more diverse casts on prime time shows.
Or, do people think that joining is simply pledging support to a flawed
election method, with little hope for changing their position?

I reply:

The latter. I was a member of CVD, from the time when it first
formed, under a different name, till I became disillusioned around
'94.

I can assure you that being a member of CVD doesn't buy you any
influence on their policies. CVD isn't a democratic organization.
The "leaders" accept your money, and do exactly what they want to.

I once suggested to Richie that he change the name of CVD to
CVA, "Center for Voting & Autocracy".

A mathematician once suggested to me that "MPV", (CVD's earlier
name for IRV) should be changed to "NPV", "Nonmonotonic Preferential
Voting".

In '93, the undemocratic nature of CVD became more obvious than ever,
and that's when I stopped sending them my annual re-subscription.

It's amusing that Don (Davison, aka New Democracy) said that
opponents of IRV are in a timewarp. The IRVies are promoting a
19th century method, for the 21st century. Not that IRV was any
good in the 19th century either, but at that time the subject hadn't
been studied as much. There's no excuse now.

I should add now that it's best to ignore Don. That IRVie twit
hasn't changed in the years that he's been on EM, and nothing that
you say to him will help him any.

We've talked about criteria that IRV fails, and which Approval passes:
FBC, Monotonicity, Participation, Consistency, IIAC (if we use my
IIAC, or Regularity, which Markus accepts for IIAC--I haven't seen
Arrow's own IIAC definition).

I recently posted an example in which someone caused his last-ranked
candidate to win, because he showed up to vote in IRV. Nothing
like that can happen with Approval. Have you been on the list long
enough to hear the discussion about Monotonicity, Participation,
& Consistency?

Let me add another criterion, the Heritage Criterion:

If, by a certain set of ballots, a certain candidate wins, among
a certain set of candidates, then, by that same set of ballots,
that same candidate should win among any subset of the candidates of
which he is a member (when the candidates not in that subset are
deleted from the ballots, and the ballots recounted).

[end of definition]

That's equivalent to my IIAC, but it's more convincing, it seems
to me. Again, then, Approval passes that criterion and IRV fails it.

That's the actual-votes version of Heritage.

Mike Ossipoff





Alex Small



_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com