[EM] San Francisco pairwise tallies

[Bart Ingles]

Justin Sampson wrote:
> On Thu, 11 Nov 2004, Bart Ingles wrote:
>>From the Chronicle article I saw, it looked as though _none_ of the 
>>four races requiring a second round ended up with the eventual winner 
>>recieving a majority of votes.  Did this turn out to be the case?
> 
> 
> Yes, I made the same observation based on the pairwise tallies. On the
> other hand, Steven Hill of the Center for Voting and Democracy points out
> that "all winners were elected with many more votes than in previous races
> for Supervisor", presumably because of the typically low turnout at
> December run-off elections and the abnormally high turnout of this
> election:

It looks as though the district 5 winner ended up with 33.7% of the 
total ballots.  The "effective turnout" for the final round of this 
election was about 67% of the overall turnout.
http://www.sfgov.org/site/uploadedfiles/election/RCV5.htm

For district 7, the winner ended up with 39.7% of total ballots, with a 
"surviving" turnout of around 70% of the original turnout.
http://www.sfgov.org/site/uploadedfiles/election/RCV7.htm

 From the links above, it seems obvious that the vast majority of 
exhausted ballots in these races were due to the 3-choice limit, and not 
from people voluntarily undervoting.  So the low winning totals in these 
races are not really analogous to a low December turnout-- obviously 
we'll never know how many of the voters with exhausted ballots would 
have preferred the runner-up enough to vote in a December runoff (or how 
many with non-exhausted ballots would have stayed home).

Not that I place any great value on having a majority winner-- actually 
none, unless there is an outright first-choice majority evident on the 
ballots-- but this does does discount the idea of "guaranteeing a 
majority winner" as a selling point.

Bart