In a diary that TPTB at dKos kindly promoted to the front page, yesterday I reported on the first full day of the recount. At that point, Dino Rossi (R-smarmy wingnut in sheep's clothing) had added 25 votes to his razor-thin margin over Christine Gregoire (D-triangulating wimp). As I discussed earlier in an extended series of diaries, the original result of the November 2 election gave Rossi 261 more votes than Gregoire, out of nearly 2.9 million ballots cast in Washington.
By Tuesday night, with 36 of 39 counties in the fold, the recount had added a total of 55 votes to Rossi's margin. In other words, if the three remaining counties reported the same results as before, Rossi's overall margin would be 316 votes. However, one of the late counties is King, by far the largest in the state and strongly Democratic, so the outcome remains uncertain.
Let's assume that Kitsap adds another 20 votes to Rossi's margin. I think that probably overstates its effect, but I'd rather set up a tough case than an easy one. Our question, then, is this:
Can the revised count in King County increase Gregoire's within-county margin by more than 341 votes?
According to RonK, who has access to far more detailed information than I do, the counties using optical-scan technology to tabulate votes have generally increased their total number of valid ballots by about 0.1% to 0.2%. In King County, where the initial count came to 898,238 counted votes, that would mean somewhere between (rounding off) 900 and 1800 extra votes. Applying the previously-seen percentages won by the candidates in King County (Gregoire 56.25%, Rossi 39.05%, Bennett 2.10%, blank/write-in 2.60%) to these values, we obtain the following counts of new votes:
adding 900 total votes: Gregoire 506, Rossi 351, Bennett 19
adding 1800 total votes: Gregoire 1012, Rossi 703, Bennett 38
The marginal differences in these two situations, respectively, +155 and +309. So even at the upper end of the range of expected additional valid ballots, this model shows Gregoire coming up short.
But then there are the 700 or so "extra" ballots that Republican state chair Chris Vance has been bleating about. I think those are ballots sent to the county's Canvassing Board for adjudication after the counting-room observers failed to agree on the intent of the voter. How many of those 700 ballots might be found valid by the Canvassing Board? Would they distribute among the candidates as previously-tallied votes have? Will the potential 900-1800 votes discussed above distribute that way? Will there actually be that large a number of added valid ballots? How many more lawsuits will be filed in the next few days (and by whom)?
My speculation here is based on unfounded assumptions and estimations. But, in the end, it hardly matters. However today's results come out, you can take it to the bank that whichever candidate is trailing when Secretary of State Sam Reed certifies the results on December 2 will request a manual recount. I'm sure they're both raising funds to cover the $700,000-plus deposit that will be required in order to trigger that step.
Stay tuned! I'll probably write still more diaries about the second recount. And, eventually, I plan a more comprehensive look at Washington's patterns of voting for governor, senator, and president.