The WA Secretary of State has just released, instructions to county auditors on conducting the gubernatorial recount. I've posted a copy of the document here
All valid write-in votes for Governor must be individually tallied. Optical scan counties may need to out-stack overvotes, undervotes, and write-ins to find all the valid write-ins to tally. RCW 29A.60.021(3).These instructions include watching for and examining undervotes (e.g. hanging and pregnant chads, partially marked optical scan circles) and situations where the voter (or someone) wrote in the name of a candidate who is on the ballot without also punching/filling-in the appropriate circle on the ballot. Applicable state laws are cited, but some observers tell me that these instructions are beyond what is required during a machine recount. The opportunity for foul play, or at the very least widespread confusion and controversy seems likely. Posted by Stefan Sharkansky at November 19, 2004 11:40 AM | Email ThisIn optical scan counties, if the ballot counter cannot out stack undervotes for the office of Governor, then a manual check for undervotes must be performed.
Any write-in votes for Christine Gregoire, Dino Rossi, or Ruth Bennett must be added to your final results for those candidates. Please remember how the following situations are counted:
- A write-in vote is valid if the voter included the candidate’s name, office, and party. For optical scan counties, the voter indicates the office by writing in the name in the spot for write-ins under listing of Governor candidates. RCW 29A.60.021
- As Christine Gregoire, Dino Rossi, and Ruth Bennett were declared candidates, exact spelling of the candidate’s name and including the office and party would not be required for the write-in to be valid. RCW 29A.60.021
- In an optical scan county, the oval or arrow does not need to be marked for the write-in to be valid.
We needed uniform statewide guidelines, and we apparently got them from Sam Reed's office. They will actually benefit Rossi, by making the playing field level.
King County already does everything that Sam Reed's guidelines require. Other counties do not, and in the past, simply fed the ballots back through the machine with no further ado.
In 2000, Cantwell and Bonker each picked up about 0.05% extra votes (from their original total) in the King County recount. Gorton and Reed each picked up about 0.03% extra votes.
In 1998, there was a Dem primary recount for King County Prosecuting Attorney, in which the total Dem primary votes were increased by slightly over 0.1%.
This procedure mainly picks up improperly marked poll ballots, since the absentees and provisionals have all been manually inspected before putting them through the machine in the original count.
In King County, there may be a greater increase for Democrats, since the poll ballots are several points higher for the Dems than the overall ballots. Also, Dem voters improperly mark their ballots at a slightly higher frequency than Rep voters.
I would estimate that Gregoire is likely to have a net pickup of 100 to 200 votes from this procedure, which King County would have done anyway.
Since other counties now have to do a thorough procedure, hopefully Rossi will have some net gains in some of those counties to offset this.
By the way, your link to the guidelines document does not currently work. So I am assuming that these guidelines require the other counties to do as thorough of a job as King County would have done anyway.
Failure to have uniform statewide guidelines would be an equal protection violation, exactly like what seven justices condemned in Bush v. Gore (five stopped the Florida recount, the other two would have ordered a new recount with uniform statewide guidelines).
Posted by: Richard Pope on November 19, 2004 12:34 PMThanks for the background and for letting me know about the broken link. It's now fixed.
Posted by: Stefan Sharkansky on November 19, 2004 12:39 PMIn optical scan counties, all overvotes and undervotes for Governor will have to be manually examined, as I read these rules. Gregoire has the lead in optical scan counties, since King is optical scan.
However, there is no corresponding requirement in punch card counties to manually examine overvotes and undervotes. Nor is there a prohibition against doing so either. The rules are silent for punch card counties.
The punch card counties are Asotin, Benton, Franklin, Stevens, Lincoln, Okanogan, Clark, Lewis, Pacific, Thurston, Mason, Clallam, Island and Whatcom. On the whole -- probably a 15% to 20% edge for Rossi.
Based on the Florida experience, a manual examination of punch card ballots can result in a 0.1% or more increase in votes for all candidates. In Florida, the punch card ballots had already been run through the machine three times when this 0.1% increase was achieved: (1) the regular election night machine tally, (2) a machine tally the next day (without further ado) as required under Florida law, (3)another machine tally to sort out the undervotes and overvotes when the Democrats requested a hand count.
Running the ballots through the machine a second time will cause some of the hanging chads to fall out, and give Rossi some help in the punch card counties.
However, unless Sam Reed clarifies his rules to require manual examination of all overvotes and undervotes in punch card counties, I think the folks supporting Rossi may very well need to cough up $150,000 or so to manually recount all the punch card counties. This money can be a lot less if Thurston, Whatcom and Pacific Counties are omitted.
We may have to do this even if we win the recount. The Democrats intend to pay for a hand recount of King County. If that changed the lead to Gregoire's favor, the remainder of the state would have to be recounted. But if Gregoire's lead was more than 150 votes as a result of the Dems requested recount, then the remainder of the state would only be counted by machine, and not by hand.
Since this would result in a nonuniform recount (by hand where the Dems requested it, by machine elsewhere), it is almost certainly an equal protection violation under Bush v. Gore. However, it is what the law says: RCW 29A.64.050 (partial recount changing results requires a complete recount, but will be done pursuant to RCW 29A.64.021 -- by hand only if less than 150 vote difference).
If a requested partial recount changes the results of the apparent winner, then the remainder of the state has to be recounted as well.
However, it would appear that the remainder recount would be by hand or machine based on whether the margin (after the partial recount)was less than 150 votes, and not based upon whether the partial recount was by hand or machine.
I complained to Sam Reed today about the equal protection violation due to treating optical scan ballots differently than punch card ballots.
Nick Handy, Director of Elections, responded to me. Unfortunately, Sam Reed is not willing to change his position on the matter. Punch card ballots will not be manually inspected for overvotes and undervotes, while optical scan ballots will be.
Posted by: Richard Pope on November 19, 2004 07:39 PMIf so, what a can of worms!!!!
Posted by: leaddog2 on November 20, 2004 09:03 AM