[EM] STV's rejection: it's "not a defect, it's a feature!"

[David Catchpole]

On Thu, 16 Mar 2000, Bart Ingles wrote:

> This sounds like the block voting method used in most U.S. city
> council-type elections.  I can see how the vote would be split by three
> or more factions contesting for all of the seats (because voters are
> limited to 8 choices), and I can see how running more than 8 candidates
> could cause an individual group to split its vote, but how would
> standing only 5 candidates cause the vote to be split?

There are then 3 "dangly bits" on each vote which contribute to the
election of other candidates. If votes for "dangly bits" edge opposition
candidates over the edge..., and that's the worst case scenario-
more likely the votes head for minor-candidate limbo and
contribute to no-one's election... Also, the "dangly bits" from voters
for an "independent" are not going to uniformly go your
way. With 8 candidates, all your votes are all your votes, plus as soon as a
voter picks you to be the dangly bits to her vote, all the dangly bits
will be your dangly bits...

Running 5 candidates as opposed to 8 candidates is not quite
"splitting" in the same sense as a single-winner FPP system, but it does
have the same effect- it reduces the numbers of winners for a side. Our
current State representative was edged out in the local government 
election of 1986 by the above effects. It seems strange to say that the 
block vote was the difference between Councillor Hollis and Speaker
Hollis. The one member of the team that year who got elected got in
thanks to the donkey (top-to-bottom) vote, has turned feral, and is now
quite definitely no-one's friend (he's a present councillor and is one of
the 10 "independents"). He's had the tactic of getting into everyone
else's "how-to-vote" directions. That's great for an individual, but more
difficult for a team that wants to make up a majority of the next council.

> Incidentally, where the block vote method has been replaced recently in
> various U.S. jurisdictions, the only approaches that have been
> successfully enacted are single-seat districts and cumulative voting. 
> As an example, in San Francisco a few years back there were two
> referenda on the ballot, one to adopt STV (which lost), and the other to
> switch to single-seat districts using plurality/runoff (which won).  A
> later attempt, pushed by the CVD and its local subsidiaries, to adopt
> IRV for those districts never made it onto the ballot.  I wonder what
> would have happened if CVD had pushed for cumulative voting instead of
> STV in the first place.
> 
> In the U.S., STV is used New York City school board elections, where it
> seems fairly unpopular, and in Cambridge city and county council
> elections.  All recent attempts to adopt it elsewhere have failed.  But
> there are probably dozens of school boards and town councils who have
> adopted cumulative voting over the past 10 years, apparently with no
> major complaints despite the method's known limitations.
> 
> 

--------------------------------------------------------------------
"Never ascribe to conspiracy what can be put down to stupidity"- DJ No MC