Friday, November 18, 2011

2011 Whatcom County General Election Charts

Do you believe in what you see?...Wastin my timein the waitin lineDo you believe in what you see?
(from "Do You Believe" by Zero7)

Preliminary Election Charts.Standard disclaimer applies: Accuracy of this data and these charts cannot be guaranteed. You should crunch your own numbers and not present my work as your own. Send me any corrections or comments you may have.  Data  is here. I am missing 11/10 count such that 11/10 and 11/12 roll up into one large sample.  I have not bothered with the last counts of a few hundred voters. Clicking on my screenshot open office charts enlarges them in blogger.

Some concepts appropriate for the 'science' for "[registered] voter participation": "Fall off" describes the decreasing voter participation rate. The theory, as statistically demonstrated below, is that early voters complete greater percentages of their ballot than later voters. "Ballot fatigue" describes the tendency of voters to complete less and less of their individual ballot. "Ballot fatigue" is such an important issue that states, like WA, mandate the ballot order in such a fashion so that state based initiatives and offices receive 'priority' or an earlier position on the ballot.  Both phenomena also help us here in this election to [possibly] infer that the ballots received and counted later were actually mailed or 'drop boxed' later.

Participation Rates: Left hand are cumulative. Right side are per count.
Below we see that voter participation (per count) for Larson/Kremen was statistically similar for Ericksen/Louws.  Despite the fact that Crawford/Maginnis was arguably Whatcom County's the most defined left/right county race (HE a land use consultant for developers/SHE a regulatory scientist with ECY), the race suffered a collapse in voter participation rate. It is true that port commissioner, auditor, and Brenner vs. Black races had  lower participation rates. But these races started out at lower participation rates as well.

Vote counts come in different sized samples in a 'Vote By Mail' election. After the sample that counted 57,044 (11/12) ballots, the vote swung decidedly against the Democratic backed candidates. Maginnis lost her race on the 11/14 count. She went in up 601 and came out down 623. A 1222 vote swing in a count of less than 5396 votes for that race! She never recovered.  As I see it, Crawford increased his percentage of the vote in over 100 of 120 precincts in that count. And yet, the participation rate in the vote steadily declined with every count, suggesting that these were truly 'late, conservative' voters.

This is the existing precinct vote totals for '(Crawford -Maginnis on 11/12)/(Crawford-Maginnis on 11/12). (e.g. a ratio   of increase/decrease per precinct).  For the 11/14 count (5396 total votes for the Maginnis/Crawford race), Crawford appears to have increased his percentage of the vote in 100 out of 120 precincts for that vote count. This statistic is fairly contrived. 

No comments: