My analysis was wrong

Based on the 2009 primary, I concluded that

... the same number of voters like Ms. Longrie, no matter how many people vote; while the same proportion of voters like Mr. Rossbach [emphasis added]

Now that we have general election results, we can see that I was wrong.

	2005	%	2009	%	Change	%
Longrie	2861	53	2348	39	-513	-18
Rossbach	2531	47	3081	51	550	22
Other	21	0	581	10	560	2667
Total	5413		6010	0	597	11

In turns out 513 fewer voters like Ms. Longrie after four years of incumbency featuring bimonthly cablecast council meetings, Mayor's Forums, multiple cable shows, and other appearances.

This loss of support occurred despite an 11% increase in turnout, a pool of 597 more voters.

No one knows who voted for whom, but one way of accounting for the results is that 500 or so voters, when given a conservative alternative, abandoned Ms. Longrie to vote for Mr. Smart, while virtually all of the 500 or so additional voters who turned out voted for Mr. Rossbach. So, it is probably correct to stick with the earlier conclusion that added turnout would not be good for Ms. Longrie, but very good for Mr. Rossbach.

Stephan