Tuesday, November 13, 2012

The Aftermath (Part III): 42nd LD voters for both Cantwell and Overstreet?

In between feeding kids, cleaning house, and raking leaves this weekend, I've been making my Powershell 3.0  prompt zing.  Charts pictured far below are:

  1. McClendon vs. Overstreet 42nd LD (1) only
  2. Cantwell vs. Baumgartner 42nd LD precincts only
  3. Cantwell vs. Baumgartner  All of Whatcom county (e.g. 40th and 42nd inclusive).
  4. McClendon vs. Overstreet for the 100s and 200s 42nd LD precincts

As of the latest count (11.10.2012:  92.5K counted, at least 9.5K  left to count), Republicans Jason Overstreet and Vincent Buys have been creaming their Democrat opponents by 5,540 and 5,192 vote margins respectively in WA 42nd LD:

Jason Overstreet                     Natalie McClendon
31903                                26363

Vincent Buys                         Matthew Krogh
31691                                26499

But curiously, Democratic Senator Maria Cantwell is leading her Republican candidate in those same 120 42nd LD precincts by 3,716 votes. Below is what that looks like in Powershell 3.0. Senator Cantwell even has a higher average per precinct return than candidate Baumgartner for 42nd LD precincts:

$Diff.Cantwell | measure-object -sum -average -min -max

Count    : 120
Average  : 266.325
Sum      : 31959
Maximum  : 641
Minimum  : 17

$Diff.Baumgartner | measure-object -sum -average -min -max

Count    : 120
Average  : 235.358333333333
Sum      : 28243
Maximum  : 561
Minimum  : 11

This might make sense to many of you, since one could assume that there would be a  much higher total vote return for a U.S. Senate over a  WA legislative race especially with a two term U.S. Senator running for re-election. But for the totals for the 120 42nd LD precincts  in question that is just not the case so far :

Votes       Race

58,190 Krogh v. Buys
58,266 McClendon v. Overstreet
60,202 Cantwell v. Baumgartner

So what gives here? Let us assume (with some degree of probability) that the Senate and LD[1]   populations above are stable and inclusive of all voters in the 42nd. (The Senate vote may end up with the greatest volume in Whatcom County second only to the Presidential vote.) If we subtract the 1,936 42nd LD ballots that voted for U.S. Senate but not for the slightly smaller WA LD (1) population from 5,540 that voted for Overstreet over McClendon, does that mean that the remaining 3,604 voters would be the minimum who would have marked their ballots for both the  Democrat Cantwell AND for the Republican Overstreet? Did a minimum 11% of the people who just voted for Jason Overstreet  also vote for Maria Cantwell as well?

31903 - 26363 = 5540  (McClendon - Overstreet)
60202 - 58266 = 1936  (Cantwell/Baumgartner - McLendon/Overstreet)
5540 - 1936 = 3604      (Difference between 42nd votes and those who did not vote for Senate)
3604/31903=0.11         (% minimum vote for both (D)Cantwell and (R)Overstreet?)

Maybe my political logic has outstripped the usefulness of my math.  But these results seem to scream 'hand recount' to me.  Below are some graphs. The left graph is precinct participation. ( I used a 10/31/2012 Voter Database History for precinct totals. ) The right graph is the difference per precinct. Left of the 0 on the X axis is the Republican margin. Right of the 0 on the X axis is the Democrat margin. Notice how much better Cantwell does than McClendon in many precincts in the 42nd. So much better, she didn't even need her overwhelming 40th LD to win. Click on any image to view them all larger in Blogger's photo slider.

McClendon vs. Overstreet 42nd LD (1)

Cantwell vs. Baumgartner 42nd LD precincts only

Cantwell vs. Baumgartner  All of Whatcom county (e.g. 40th and 42nd inclusive)

 McClendon vs. Overstreet for the 100s and 200s 42nd LD precincts

No comments: