January 24, 2005
King County vote count off by more than 3,700 votes

I've finished the massive project of trying to reconcile the King County voter list to the vote count as accurately as humanly possible. My conclusion: their numbers are off by at least 3,748 votes, or roughly 30 times former Attorney General Gregoire's alleged "victory" margin.

The previously reported discrepancy of 2,150 (348 unverified provisionals + 1,800 unexplained more ballots than voters) represents only the net difference between ballots and voters. While many precincts counted more ballots than voters, a lot of precincts counted more voters than ballots. When you add up all the voterless ballots and all the ballotless voters, the true number of improperly counted votes reaches at least 3,700 and is presumably somewhat higher.

These are the issues I had to deal with given the state of the data that King County delivered.
1) The latest voter file (Jan. 7) had been updated to reflect changes to the registration rolls since the election. In order to get precinct information that was correct as of election day, I used precinct information from the Nov. 1 file. The 557 magical mystery voters who were in the Jan. 7, but not the Nov. 1 were counted in the precincts recorded in the Jan. 7 file.
2) I recovered the records of 253 voters who were listed in the Dec. 29 file, but had been dropped by Jan. 7.
3) I removed the 636 duplicate records that the County included in the Jan. 7 file.
4) For whatever reason, the 566 magical "Larry Phillips" ballots that were counted by the canvassing board on the last day of the manual recount were not added to the precinct totals in the Manual Recount Canvass file. They were listed as a separate line item called "Canvass Additions". In order to include these ballots in precinct totals, I obtained from the County the list of the voters in question and added their votes to the precinct counts based on their precinct assignments as of Nov. 1.
5) I found in my earlier analysis of ballots vs. voters by precinct that a lot of the discrepancies could be explained by polling places. Ballots that should have been counted in one precinct were counted in a different precinct that shared the same polling place. This is still a problem, but more innocent than some of the other explanations for ballotless voters and voterless ballots. In order to factor out this less serious error, I totaled ballots and voters by polling place. The County failed to satisfy my request for a clean list of precincts and polling places, so I took this PDF file of polling places, converted to text and manually corrected for variations in spelling.
6) I noticed that within polling places, polling ballots appeared to be misattributed to provisional voters and vice versa. Another problem that if not excusable also does not imply there are illegitamate ballots being counted. I rolled up provisional and polling ballots into a single category per polling place.
7) Finally, the Manual Recount Precinct Canvass, unlike the canvasses for the two earlier counts, does not make separate counts of absentee, precinct and polling ballots. I used the counts of each type of ballot per precinct from the machine recount, but the precinct total from the manual recount, realizing that the totals would still be off by 1 or 2 in some precincts. (In addition to the 566 "Larry Phillips" ballots, there were also 266 ballots that disappeared and 325 ballots that magically appeared between the machine and manual recounts, for a net change of 59 brand new ballots)

After all of that, I found the following total discrepancies between ballots and voters:

Polling Place
Absentee
Polling/Provisional
Voterless Ballots
1,267
1,975
Ballotless Voters
(485)
(341)
For a total of 4,068 ballots that either undercount or overcount the number of identified voters. 320 of the voterless ballots can be explained by the unregistered military and Address Confidentiality Program voters who don't appear in the voter rolls. So we're left with 3,748 miscounted ballots.

I still consider this number a lower bound on the number of miscounted ballots, because (a) it doesn't take into account the unexplained net increase in ballots in the manual recount, and (b) the roll-up of absentee ballots by polling place was a convenience for these calculations, although I can't think of any plausible reason why absentee ballots should ever be mixed up between precincts, even within the same polling place (When I summed absentees by precinct and not by polling place, I found a total of 1,980 voterless ballots and 1,198 ballotless voters), and (c) there are likely to unnoticed discrepancies of missing ballots and missing voters within precincts, even if they're masked by the net total for the precinct.

The following tables show the hall of shame of polling places with the largest discrepancies between ballots and voters. The numbers in the Absentee, Polling, Provisional columns represent the number of extra ballots/missing voters [in black] or missing ballots/extra voters [in (red)] of the various types of ballots. The "Total" column is not the total of these three ballot types, but is the difference between the total precinct count for the manual recount and total number of identified precinct voters. [as per item (7) above]

The "imputed lead" column estimates how much the miscounted ballots might have added to (subtracted from) Gregoire's lead, based on how the other ballots at that polling place broke to the various candidates. It's true that the net affect of the missing ballots was to reduce Gregoire's lead, but they didn't appear to hurt her nearly as much as the extra ballots seemed to help her.

Top 5 polling places with more ballots than voters
Polling Place City Absentee Provisional

Polling

Total Imputed
Lead
King Co Adm Bldg,
500 4 AV
Seattle 1 64 1 65 27.6
Bell Tower,
2215 1 AV
Seattle 26 (5) 21 42 17.7
Denny Terrace,
100 MELROSE AV E
Seattle 3 (2) 39 41 29.8
The Josephinum,
1902 2 AV
Seattle 3 (5) 39 40 15.8
Rainier Free Meth Ch,
5900 RAINIER AV S
Seattle 1 5 34 40 21.8
Also notable is Ashton Hall, Sea Pac Univ, 611 W DRAVIS ST, Seattle, which wins the prize for counting the next largest number of absentee ballots for which there is no known voter -- 25.

In all fairness, I'll suggest the possibility that some of these polling places may be disproportionately associated with the unregistered military voters and the address confidentiality voters, whose ballots are assigned to a precinct, but whose voters are not identified even by number per precinct. But that can only explain so much. 95 out of about 600 polling places show a net of at least 10 more ballots than voters.

But as fair as I can be, I can't come up with a good explanation for precincts that have more voters than ballots.

Top 5 polling places with more voters than ballots
Polling Place
City
Absentee Provisional

Polling

Total Imputed
Lead
Clark Elementary,
500 2 AV SE
Issaquah
(10) (16) 12 (14) 1.5
Ballard House,
2445 NW 57 ST
Seattle
(7) (4) (2) (11) 6.3
Martin Sortun Elementary,
12711 SE 248 ST
Kent
(5) (7) 1 (10) 1.1
Ravenna Eckstein Comm Ctr,
6535 RAVENNA AV NE
Seattle
(10) (2) 4 (9) 5.7
Bennett School, 17900 NE 16 ST
Bellevue
(5) 0 (2) (8) 0.7
The bottom line? The imputed lead for all of the mismatches between ballots and identified voters was 649. That calculation is based on the fact that the unidentified voters includes the unregistered military and ACP voters, who can't be easily subtracted from this number. But in all probability, the number of extra votes that Gregoire picked up from the various extra ballots and missing ballots is hundreds more than her margin of "victory".

Posted by Stefan Sharkansky at January 24, 2005 12:27 AM | Email This

Comments
1. What is the bottom line? When you add up the imputed lead figures for all 600 or so polling place locations for King County, what is it? Considering that it was around 112 net benefit for Gregoire, just for the top five voterless ballot precincts, it would seem that the county as a whole would have a net advantage for Gregoire of several times the 129 vote victory margin.

Posted by: Richard Pope on January 24, 2005 12:46 AM
2. Richard,

I answered your question by adding a final paragraph to the post. The one that begins "The bottom line?"

Posted by: Stefan Sharkansky on January 24, 2005 12:57 AM
3. Great job Stefan. I've worked at both dailies as a reporter and I am stunned they haven't worked as hard as you at the statistical analysis. They have teams of librarians trained to sift through data so the reporters have get a clear, lucid picture. But it's not happening for some reason.
thanks so much.

Posted by: haze on January 24, 2005 01:08 AM
4. Stefan,

No comment as such, just a thank you for all the hard work. it is clear and to the point.

Posted by: TomasM on January 24, 2005 01:25 AM
5. Stefan,

Outstanding analysis! You are truly a marvel and a valuable asset to the people of WA.

Jeez man, don't you ever sleep?

God Bless! And take some vitamins or something. Keep that mind healthy!

Posted by: Clint on January 24, 2005 05:05 AM
6. The cleanest most accurate election ever? Really?

Seems Rossi clearly has a compelling case. Also it looks like SecState Reed might have some explaining to do.


Wow....Good job Stefan -Spin

Posted by: SpinDaddy on January 24, 2005 05:10 AM
7. ...And if you ever attend a dinner thrown by gregoire supporters, DON'T EAT THE BORSCHT!

Thanks for all your work, and God Bless you and your loved-ones!

Posted by: Clint on January 24, 2005 05:11 AM
8. So, with nothing better to do that actually contributes to society, SS posts hid self- proclaimed "definitive analysis" of the HIS reconciliation of the King County voter list to the vote count.

The "bottom line?" In an election with exceptionally high turn out, in a county with almost a million votes cast, with the luxury of time and biased assumptions . . . SS concludes that the error rate for King County is less than ½ of one percent.

But wait . . . SS isn’t "100% accurate" as the rabid Repubs on Sound Politics demand. NO . . . his "definitive analysis" is only as accurate as humanly possible.

So now there are lies, damn lies and the "definitive analysis" of SS.

Posted by: baXter on January 24, 2005 05:13 AM
9. baxter,

Wait a bit for your morning meds to kick in, and then go troll someplace else.

I'm sure Equine Rectum or one of the other pro-gregoire blogs would appreciate your wisdom much more than the folks here.

And God Bless you too!

Posted by: Clint on January 24, 2005 05:23 AM
10. I love it when the 'sunlight' of FACTS disells the 'foggy' or should I say 'fuzzy math' that has been thrown as us, as the TRUTH!

BaX, get a grip. Do your own research. If you're honest, I think you will be closer to the SHARK than to the drivel that has been published.

Hat Tip & a Wink to SHARK, again!!

Posted by: Arky on January 24, 2005 05:27 AM
11. Stefan,

You (rightfully) assumed no coordinated fraud when you came up with the number 649. It seems to me the easiest way to explain more ballots than voters is that some ballots were "lost." The human bias in a Democratic leaning precinct would be to loose a few Rossi ballots. In the least, the bias doesn't exist the other way.

The big point is, where are the "lost" ballots? Are they still stuck in machines? I mean, if ballots were found 9 different times, how can the public know those findings where not cherry picked out of other missing "piles."

Posted by: VaCSProf on January 24, 2005 05:27 AM
12. I believe that the Seattle Times suddenly breaking the story about the felons who voted is meant as an olive branch to pacify all of us who are very upset with the obvious unexplained discrepancies of the election.

Ultimately this will be spun by Democrats, particularly Goldstein at HorsesAss.org as "so what, there were mistakes made, even felons voted, but let's continue with the healing and concentrate our efforts on fully realizing the crossing of the aisle in a truly bipartisan election reform, blah, blah, blah, blah"

And of course, all the way, Democrats will fight meaningful reform such as providing voter identification at the polls before receiving a ballot, etc.

That's why we are not going to let this go and why I applaud Stefan. If this one gets away, then the Democrats won't learn that what rational people really want is not the gaurantee of outcome, but the gaurantee of opportunity.

King County screwed up bad and Gregoire only won because of these screw ups.

Gregoire is not our Governor.

Posted by: Jeff B. on January 24, 2005 05:36 AM
13. I can't wait to hear SCHRAMBO squeal about THIS!!
Well done, Stefan!

Posted by: Bad Bob on January 24, 2005 05:59 AM
14. JeffB: I agree with your sentiments. I would only take slight exception to your saying "...Gregoire only won because of these screw ups". I think we're better served sticking with the claim that there's no way of knowing WHO won.

BaXter: Does it occur to you that you and your side have done NOTHING to shine further light on the disparities in the election results? You simply snipe and throw insults from the sidelines while Stefan advances the ball. You resort to minimizing any given PIECE that he puts forward, but the CUMULATIVE effect of his work has been to reveal huge problems in the election results. If your side is so confident in the results, why not cooperate and SHOW US how reliable the results were? We all know the answer: you hope to delay any findings until after a reasonble point for disqualifying the election. You want to win by default---which tells me you don't really believe you won in the first round.

Posted by: jeanneB on January 24, 2005 06:35 AM
15. Great work Shark...i can see the dem spin machine trying to explain this one away..the cleanest election has just become dirty once again:)

PR

Posted by: PR on January 24, 2005 07:12 AM
16. According to Rebecca Cook's AP article this morning, the Ds are going to argue that the only way the results could be invalidated is to prove that enough legally questionable votes went to Gregoire to change the results. The counter-argument is that if this is true, then the law on challenging elections is pointless and could never be applied in any situation, because there is no way of knowing for certain how the votes in question were cast.

She says that even before that point comes up, the Ds will argue that the the court has no jurisdiction to set aside the election.

In any case, it looks to me like the kind of careful analysis Stefan has done is essential to understanding the size and scope of the fiasco. Thank you Stefan.

Posted by: Boonie on January 24, 2005 07:34 AM
17. jeanneB:

"BaXter: Does it occur to you that you and your side have done NOTHING to shine further light on the disparities in the election results?"

Cockroaches prefer the dark.

Posted by: ScottM on January 24, 2005 07:38 AM
18. It's really semantics. The fact is that Gregoire is in office. We may not know who won, but through the quirk of our laws that allow us to stop counting and declare a winner, with any contest occuring after the fact, justice and truth are suspended while we are governed by the winner.

Gregoire is only currently in office because of the semantics of our system and not because she is the deserved or proven winner.

The fact is that had King County been more honest about its flawed practices, or had Secretary of State Reed withheld certification of the election based on all of the flaws, Gregoire would not be in office and we could let the courts determine what went wrong and who won, and then install a governor.

It is wrong to put the cart before the horse.

Posted by: jeanneB on January 24, 2005 07:39 AM
19. With regards the above post, I DID NOT WRITE THAT.
Someone else posted it in my name.

Posted by: jeanneB on January 24, 2005 07:58 AM
20. I just finished reading your post entitled KING COUNTY VOTE COUNT OFF BY MORE THAN 3,700 VOTES. Like yourself, I have conducted my own independent analysis over the last few weeks. I agree 100% that the data provided by KC is a total mess! Assuming for the moment that there are no errors in the data, the manner of its presentation in terms of apples and oranges is enough to thwart any casual analysis. Your efforts at trying to reconstruct The Truth are Hurculean! With regards to issue number 7, which involves the level of detail available in the manual recount data set, I can't tell which data set you used. Stefan, did you use the manual recount data set available on the KC web site which summarizes ballot counts by precinct or did you use the manual recount data set provided internally by KC elections to the parties which breaks down ballot counts by precinct as well as by source, namely, an absentee bucket or a polls & add-ons bucket?

Posted by: bobha on January 24, 2005 07:59 AM
21. Amazing. The Crook County elections department, with all its computers and paid staff, is incapable of coming up with any sort of analysis, virtually everyone in State office says it just can't be done, it'd take too long, cost too much, yet one man with a computer consistently knocks out one after another.
It's overwhelmingly obvious they're hiding something, and the longer this plays out, the more we'll learn, thanks to people like the Shark.
Thanks for your diligence, Stefan. I hope to see you doing more speaking engagements, you were great in Olympia!

Posted by: Bruce Long on January 24, 2005 08:29 AM
22. Impressive analysis, but the bottom line is you don't have the one set of data you need to defend your conclusions. You don't have a complete list of voters who were credited with voting in Nov. 2004 along with the precincts where their credit was assigned.

You've done impressive work attempting to reconstruct this list using data from publicly released voter lists and other sources. And, you acknowledge some possible sources of error in your list, but you don't do much to quantify that error. Even if your list is 99.5% accurate, it would still contain nearly 4500 errors -- and it's quite possible that your error rate is higher than that. Here are three possible error sources you need to consider:

1. While you picked up voters from the Dec. 29 file who were dropped in the Jan. 7 version, I don't see any attempt to pick up voters who were registered on Nov. 2 and dropped before Dec. 29. I estimate that KingCo dropped nearly 60,000 voters from its rolls over the last year. (They added approximately 110K new voters, yet, their total voter registration only increased by 52K.) That works out to an average of nearly 5000 voters per month. From Nov. 2 to Dec. 29, it's possible that nearly 10000 eligible voters were deleted from KingCo's records. I don't know how many of them voted in the election, but neither do you. However, this one source of error is potentially as large (or larger) than the number of mismatched voter/ballots you report.

2. Your assumption that precinct assignments in the Nov. 2 file are correct is suspect. It's quite possible that some of these assignments were in error and corrected in later lists. In fact, this is practically a certainty, since one of the reasons we have provisional ballots is to deal with errors in precinct assignments.

3. Your attempt to allocate provisional, absentee, and poll ballots in the final recount is pure guesswork, as you admitted in your post:

I used the counts of each type of ballot per precinct from the machine recount, but the precinct total from the manual recount, realizing that the totals would still be off by 1 or 2 in some precincts.

Uncertainties of +/- 1 or 2 voters in some precincts could easily add up to total errors in the hundreds when you consider than KingCo has over 2600 precincts.


To borrow a phrase, it would seem that the margin of error in your analysis is at least as large as your conclusions.

Now, before I retreat to the safety of my asbestos suit, let me save the other SP posters some time by being clear about my argument. I am not saying that KingCo has no problem with unaccounted for voters or ballots. I simply don't know -- and neither does Stefan. The data presented so far does not support that conclusion. The data you need will be found in the poll records of the individual precincts and in the absentee ballot lists maintained by the county. Attempting to quantify the error with precise numbers like 3748 is pure BS at this point.

Posted by: scottd on January 24, 2005 08:48 AM
23. Those are all good points, Scott.

And it's King County Elections job to provide all of this information to the public and do the analysis, if they wish to dispute mine.

You should be proud of yourself for being such an enthusiastic shill for so miserably underperforming a government agency.

Posted by: Stefan Sharkansky on January 24, 2005 08:54 AM
24. Stefan-
You've gone above and beyond!!! There will always be synicals in the crowd. It would be a crying shame if all of your hard work went unnoticed by the GOP. I sincerely hope that either they corrolate the numbers with you or have a team of "qualified" people finding these mistakes and numbers that are so far off.

I am very new to this whole thing and have been astonished at how poorly run elections are and how politicians get away with so much double talk, hypocrisy and flat out lies. It seems so easy to twist laws around to fit your own needs and wants, it seems, anymore.

In my line of work-medicine-there is this little thing called... scientific evidence!!! You base judgement and diagnosis around what has been proven-you don't lie or twist things around to benefit you-PEOPLE DIE!! We've all talked about how these mistakes would never be done in a lot of different occupations- So how is it that we continue to allow politicians to get away with it??

I want to see the day-hopefully sooner rather than later-that gregoire, cantwell, berendt and other democrats get ousted from their positions.

I am just sorry that I have lost faith in the system.

Posted by: mary on January 24, 2005 09:31 AM
25. The party was on the boil.

The Appointed One had saved them. He had worked on the computers for hours. When they had finally seen him smile, their anxiety had dissipated. In triumph, he had them each press the red Print button.

Each time the creatures without souls pressed the button the printer whirred and the results sprang out instantly. 3748 more ballots than voters. Press it again. More voters than ballots. Press again. The totals changed. Press again. Press again. Press again. It was magic! No total ever came out the same. It was a software problem.

Oh, the merriment! It was the software! It was the software! They danced and played and made sly comments. They partied into the night.... At the height of festivities a knock was heard at the door. The Appointed One (Demon Resident) shouted out: “who’s there?”

“I am The Counter” the voice replied.

“The counter? Well do come in. We’ve been counting the results ourselves” shouted DR with a wink to the hoard. The door opened and a young man with blond curly hair entered. “Do you have a name, mister counter”, DR said with derision. The young man replied “My friends call me Zeke”.

The young man went over to the voting computer and looked at the equipment. DR shouted “We’ve identified the problem. It’s the software. Look!” His soulless finger pressed the red button. Nothing happened. He pressed again. Still nothing. In growing rage, DR slammed the button with his fist. The printer whirred and the results spilled out: 3748 more ballots than voters. The same as before.

“Doesn’t appear to be a software problem” the young man announced.

DR rammed his spiny finger into the Red button again. Nothing! But when he tried to withdraw his finger, it was stuck.! He tried furiously to free himself. No relief. It was as if the Glue of Ages had fastened his finger permanently to the red orb. Then, without warning, the Red button started to glow. First dull red, then as it increased in temperature, bright red and finally, white hot. The creature’s arm became a funeral pyre. Ezekiel watched as the demon joined his ancestors.

The others never heard his screams. They only heard the sound of drums.


Posted by: Ken Muller on January 24, 2005 09:43 AM
26. scottd:

"Impressive analysis, but the bottom line is you don't have the one set of data you need to defend your conclusions. You don't have a complete list of voters who were credited with voting in Nov. 2004 along with the precincts where their credit was assigned."

I think that is the whole point - neither does King County!!! As soon as KCEC can produce the complete list of voters who were credited with voting along with the precincts where they were given credit, we will ALL know the TRUTH.

Perhaps...that is why they have been unable to produce this list?

Posted by: Larry on January 24, 2005 09:51 AM
27. Stefan,
As a former auditor and a current CPA, I'm impressed with your analysis. It is thorough and not unlike what I used to do as an auditor of financial statements and it actually makes sense unlike King County's explanations or lack thereof. I would suggest that the " voting reform board" appointed by Sam Reed should include less of the state's county workers and auditors that don't seem to have a handle on the system and more business people including some CPAs who are financial statement auditors with experience in internal control analysis. Oh yeah, and maybe they should be independent as auditors are required to be.

Posted by: Nancy on January 24, 2005 10:07 AM
28. "You should be proud of yourself for being such an enthusiastic shill for so miserably underperforming a government agency."

I dont see how he is a "shill" by pointing out in very specific detail problems in your analysis. Why not address the specifics of his concerns rather engage in ad hominem attack.

Posted by: ralph on January 24, 2005 10:36 AM
29. "Why not address the specifics of his concerns rather engage in ad hominem attack."

Because, shill "ralph", the real issue is not any potential imperfections in my analysis, but the fact that I'm doing this analysis because King County failed to. If King County staff did a proper job, I wouldn't have to do this.

Posted by: Stefan Sharkansky on January 24, 2005 10:40 AM
30. There is an interesting article in today's New Republic on Seattle Governor's race. As I am sure you know, it is a liberal, Bush basher publication. I am a Texas/Louisiana transplant, and enjoy your blog and it's information. Nice change from the sturm and drag of grey Seattle.

Posted by: sue gill rose on January 24, 2005 10:58 AM
31. "I don't belong to an organized political party. I'm a Democrat." - Will Rogers

Stefan Sharkansky wrote (without the comments from yours truly):

Those are all good points, Scott.

True, although all it really proves is that KC should be a bit more forthcoming with the actual data. Still, I presume (hope) that they are working on it and will present the actual numbers and information to the court in sworn statements and testimony.

And it's King County Elections job to provide all of this information to the public and do the analysis, if they wish to dispute mine.

Well, it would be nice if were readily available, and there is no doubt that if KC had it's act together, it would be. Still, I think we've already established that this is not the case, so we seem to be beating a very dead horse here. Since you base your numbers on KC's rather poor data, your conclusions only show once again that the data is of poor quality. Garbage in, garbage (not Gospel) out.

You should be proud of yourself for being such an enthusiastic shill for so miserably underperforming a government agency.

Cheap shot. I read his post, and it doesn't say that KC did a good job. In fact, it says "I am not saying that KingCo has no problem with unaccounted for voters or ballots." Personally, I'd go so far as to say they did a lousy job.

An argument could be made that if they did their job properly, counting (and accounting for) all the legal ballots the first time, we would have Governor Gregoire with a clear victory, instead of having to go through this mess.

Still, we're playing in your sandbox, and you're welcome to make cheap shots if you wish. I find myself writing a lot of them, but usually remove them before clicking on the "post" button.

Posted by: John Barelli on January 24, 2005 11:09 AM
32. Stephan,

Thank you for all your hard work and determination with this analysis. I agree that, in the absence of any serious counteranalysis offered up by King County, yours must stand as the best shot at the truth in these murky waters. And I also agree that those who take issue with the precision of your final estimate of discrepancy do not understand how statistical analysis works. You have ended up with an exact number; however, it is well understood by the context of your analysis that this number represents the midpoint of a scattergram of possible numbers that may represent the true discrepancy. Therefore, though it may not ultimately prevail as the true number, it is manifestly not "pure BS."

By virtue of your scrupulous work, you have gotten as close to the truth as can so far be gotten. It is a fascinating and impressive work.

You do yourself and your work little credit by using words like "shill" in describing those who disagree with you, however. A shill is someone who has been paid or who has a stake in defending or promoting another party; you have no knowledge of whether this is the case with your detractors. Without this knowledge, Ralph is correct that the use of the term constitutes an ad hominem attack. Some posters may simply take issue with parts of your analysis. I can appreciate many of their individual points, but as I said above, I think these arguments are irrelevant in the absence of meaningful counteranalysis.

Thanks again for your stellar work, Stefan.

Posted by: JB on January 24, 2005 11:12 AM
33. ralph, take another look at scottd's post. There is no "specific detail" in it whatsoever, just the flailings of an anonymous sycophant.

Posted by: Bruce Long on January 24, 2005 11:17 AM
34. jeanneB,

Sorry for that earlier post. I had meant to put my name in there and your name at the top of the message, but it was early and I was rushing out the door.

It was my post. Sorry for any confusion.

I still stand by my points though in light of everyone elses's comments; the bottom line is that there was more than enough reasonable doubt from Stefan all along the way for the results to have been certified by Sam Reed. It was totally irresponsible to put Gregoire in office given that KC can't even produce a meaningful voter database for a meaningful date.

This why we have the concept of "certification." There is no way that Gregoire should be in office now while the analysis of everything that went wrong is still in progress.

I have no faith whatsoever in our elections system. It's a sham. Anyone involved in King County elections with a conscience should resign. But wait, they are almost all Democrats....


Posted by: Jeff B. on January 24, 2005 11:33 AM
35. Stefan:

Let me echo the many voices who have said thank you for all your hard work. I know that I would never have had the will nor capability to ferret out all this information, and probably not the tenacity to continue to push forward. I am very glad that there is SOMEONE out there with the skills and drive to make a difference.

Thanks!

Posted by: Darrell on January 24, 2005 11:35 AM
36. John Barelli:

"An argument could be made that if they did their job properly, counting (and accounting for) all the legal ballots the first time, we would have Governor Gregoire with a clear victory, instead of having to go through this mess."

I've made changes in CAPS:

"An argument could be made that if they did their job properly, counting (and accounting for) all the legal ballots the first time, WHILE DISQUALIFYING ALL ILLEGAL BALLOTS, we would have Governor ROSSI with a clear victory, instead of having to go through this mess."

I'm assuming, John, that you mean:
1. Counting all legal ballots ONLY ONCE;
2. Not counting ballots from the deceased, people who are not residents of WA state, people who are not properly registered, and not allowing provisionals into the process before they are checked for legitimacy.

Otherwise, you're not just chillin', you're shillin' too.

Posted by: Larry on January 24, 2005 11:38 AM
37. Wow, that is a ton of work, Stefan, and done in a way that seems to me very reasonable.

Edit: In pgh. 7, sentence beginning "It's true that the net affect" should read "net effect".

Posted by: j.a. on January 24, 2005 11:41 AM
38. Stephan,

I understand you calculation to arrive at 649 imputed vote gain for CG. This shows that elections officials botched the election in KC.

Yet, according to the RCW governing how to handle the errors caused by the obvious misconduct we have seen in KC, the difference will be quite a bit more. According to the RCW the vote (all the votes) from precincts like the ones you showed in your article would be thrown out. I do not have the numbers, but I bet it is a lot more than 649 net gain for CG.

Maybe you could give the vote break down for the top ten precincts with the two types of errors you discussed (total of 20 precincts or polling places). This would give a better view of how the court would look at these problems.

Posted by: Barry on January 24, 2005 11:51 AM
39. Stefan:

The closer you come to the truth, the more some people will be upset with the truth. That's just good old human nature. Truth is unsettling to minds that want a certain outcome. And you know the old expression: "Don't confuse me with the facts".

Thanks for all of your hard work to get at the truth of the election We all owe you alot.

Posted by: Susu on January 24, 2005 12:03 PM
40. Larry wrote:

Otherwise, you're not just chillin', you're shillin' too.

From the Democrat's position. When I say "An argument could be made that if they did their job properly, counting (and accounting for) all the legal ballots the first time, we would have Governor Gregoire with a clear victory, instead of having to go through this mess." I mean exactly what I say. Please note, I did not say that it was a certainty. Just that an argument could be made. You are welcome to disagree with that argument, but the fact is that due to KC's incompetence, we will probably never know.

There is no doubt that the errors and just plain incompetence of KC's election commission has caused the final tally of this election to be in (serious) doubt.

It is also far from certain that if all the legal votes were counted the first time (do I really have to spell out that this means not counting illegal votes?) that Mr. Rossi would have won.

Now, for everyone's enlightenment, the definition of the word "shill," taken from Webster's dictionary:

Main Entry: shill
Function: noun
Etymology: perhaps short for shillaber, of unknown origin
1 : one who acts as a decoy (as for a pitchman or gambler); also : one who makes a sales pitch

Please note that the word "Democrat" does not appear in the definition, nor does it indicate "someone that disagrees with me politically." You are welcome to write to Mirriam-Webster about these deficiencies in their definition.

Posted by: John Barelli on January 24, 2005 12:03 PM
41. Exact? To the nth degree? How is that possible, in the face of flawed data? We cannot “know” the facts; therefore we let the side most closely affiliated with the errors win without a contest?

Perhaps in Hollywood. Hopefully not in the real world.

Pre-computer, much of scientific inquiry was carried along by slide rules, which are inexact by definition. Yet decisions were made by best guess estimates.

The best and the brightest sent men to the moon, and yet the failure of hardware store items nearly snuffed out the lives on Apollo 13.

At its very core, the argument against the re-vote is a combination of obfuscation and pointing to errors in precision. Yet we all make decisions based upon incomplete information. It’s what decision makers do.

The things that put Gregoire in office and keep her there and delay justice are precisely the opposite of what these same people do in their personal lives. Anyone who says different is either lying or has a personal life that is in shambles.

Posted by: scott158 on January 24, 2005 12:06 PM
42. John Barelli:

I know you mean exactly what you say. I just think that you say nothing of worth.

An argument could be made that the sky is blue, or it isn't, or it just looks that way. An argument could be made for any position and any opinion ever stated - but that doesn't mean it will be a GOOD or LOGICAL argument.

"do I really have to spell out that this means not counting illegal votes?"

You personally, John? No. But for once I'd like to hear Gregoire or Logan or Sims say it. Just once.

Posted by: Larry on January 24, 2005 12:12 PM
43. Stefan said:
Those are all good points, Scott.

And it's King County Elections job to provide all of this information to the public and do the analysis, if they wish to dispute mine.

[ad hom deleted]

Well, no, it isn't King County's job to generate data to your specification -- and it's not their job to refute your faulty analysis that was based on inadequate data.

Their job requirements are laid out by statute and regulation which may be met by organizing data in ways that are not convenient to the way you want to analyze it. For example, there is no legal requirement for them to provide you with a list of every voter who cast a ballot in the last election. That information is encoded in the various poll books, precinct reconciliation forms, and absentee ballot lists. They are not required to collate that data to your satisfaction just because you ask.

They are required to provide public records that already exist and they have satisfied that requirement by delivering lists of currently registered voters -- with the caveat that that data is constantly being updated. The only problem is, that's not the data you need for your analysis.

On the other hand, you are responsible for your own analysis and the conclusions you draw from it. In this case, you've taken data that you know is inadequate for your purpose -- and from this you've made assertions that simply cannot be reliably derived from the data you have. Then you try to pass off responsibility for your own analysis by claiming it's KingCo's job to prove you wrong.

Maybe you think KingCo should provide you with data in a different form. They apparently don't think so. That may be a mistake on their part, or it could be that they are busy producing records for people who actually have legal standing in this case. None of that changes the fact that you don't have the data you need, so the statements you make are just speculation. And yet, you present them as fact.

Posted by: scottd on January 24, 2005 12:23 PM
44. Scottd,

You wrote "Their job requirements are laid out by statute and regulation which may be met by organizing data in ways that are not convenient to the way you want to analyze it. For example, there is no legal requirement for them to provide you with a list of every voter who cast a ballot in the last election. That information is encoded in the various poll books, precinct reconciliation forms, and absentee ballot lists. They are not required to collate that data to your satisfaction just because you ask."

I'm new at this, so I may be off here, but your argument has a flawed basis. Using your statement above you say that Stefan should have gone and used the items above to do his work.

Your suggestion to use "various poll books, precinct reconciliation forms, and absentee ballot lists." would undermine and flaw any statistical analysis performed on this fiasco because the validity of all those items are what is in question here. The only way to produce any results with certainty is one of 2 ways. Either produce the list of voters that voted so proper analysis can be done or revote with proper rules in place, (oh, and strictly adhering to them too.)


Cid

Posted by: CidinDupont on January 24, 2005 12:43 PM
45. I would well imagine that KC has done exactly the same analysis as our friend Stephan has, and hence we have seen some desperate, albiet futile, attempts to actually reconcile the numbers, all the while hoping (if not praying) that they will be taken for their word and that no dignified representative of the MSM (yes, that is an intended oxymoron) would attempt to do the same. What they didn't count on, and now probably wish they had, is that someone like the SHARK would crunch the numbers and leave KC feeling a tad bit exposed, naked and embarassed.

Stephan, this is just the kind of work that will force KC, and other self-respecting counties to generate an accurate reconcilliation of the votes. I'd well imagine that KC is now wishing that they hadn't released so many conflicting voter records between Nov '04 and Jan '05.

Nice work, your commitment to getting to the the bottom of this issue is commendable and inspiring.

Posted by: theBerean on January 24, 2005 12:45 PM
46. All your numbers and painstaking research cannot hide the fact that you were OK with a purported win by Rossi of 42 votes but you cannot abide a win by Gregoire of 129 votes. This in my book still equals hypocrisy and dishonesty on your part and a philosophy of winning at any cost---even the good of the state.

Posted by: headless lucy on January 24, 2005 12:48 PM
47. Actually scottd according to RCW 29a.08.125 which is effective until Jan 1, 2006 "The computer file (maintained by the auditor) must include, but not be limited to, each voter's last name, first name middle initial, date of birth, residence or address, gender, date of registration, applicable taxing district and precinct codes, and the LAST DATE THE INDIVIDUAL VOTED." So, these public records already exist and should be simple to provide to the public. Press print after witholding the birthdates, easy.

Posted by: ChuckJ on January 24, 2005 12:52 PM
48. headless lucy,

Sorry to have to point it out to you, but no one know how messed up KC was until after the third count.

We will never know what Rossi or the WSRP would have done with this info if they got it eariler. Since he is only asking for a revote and not to reverse the results (which the facts would support), one could guess that he would still ask for a revote.

This blows a hole in your statement: hypocrisy and dishonesty on your part and a philosophy of winning at any cost---even the good of the state.

But it seems the truth doesn't matter.

Posted by: ubuck on January 24, 2005 12:59 PM
49. All your numbers and painstaking research cannot hide the fact that you were OK with a purported win by Rossi of 42 votes but you cannot abide a win by Gregoire of 129 votes. This in my book still equals hypocrisy and dishonesty on your part and a philosophy of winning at any cost---even the good of the state.

I'll plead not hypocrisy, but ignorance. I wasn't aware of the extent of the inadequacies in the election system at the end of the machine recount. Indeed, the voter lists weren't made available until after the end of the manual recount.

Of course if the Democrats had raised the same issues immediately after the machine recount that I'm raising now, they would have persuaded me that the elections system was a mess and probably would have persuaded most of the rest of the public as well. But for some reason they didn't do that, did they? hmmm.

Posted by: Stefan Sharkansky on January 24, 2005 12:59 PM
50. CidInDupont said:
I'm new at this, so I may be off here, but your argument has a flawed basis. Using your statement above you say that Stefan should have gone and used the items above to do his work.

Sorry if I wasn't clear, but that wasn't my point.

I can't fault Stefan for not using data that he doesn't have access to. I'm just saying that the data he does have isn't sufficient for his analysis. Without better data, he can't support statements like the headline of his post ("King County Vote Count Off by More Than 3700 Votes"). My suggestion would be to hold off on such statements until he can back them up with hard data.

Stefan's point seems to be that it's OK to draw bogus conclusions because he hasn't been provided with the data he wants. Rather than acknowledge the flaws in his analysis, he'll simply call names on the people who point them out.

Posted by: scottd on January 24, 2005 01:01 PM
51. ChuckJ: The RCW you cited says that the auditor must maintain a list of registered voters -- and indeed they do. Stefan has obtained copies of this list at various times. The problem I pointed out is that this list will not include voters that have been deleted from the list since the last election.

Posted by: scottd on January 24, 2005 01:07 PM
52. headless lucy

You complain that Stephan did not seem upset at the time of the 42 vote Rossi win, but then became upset when Gregoire 'won.'

This is analogous to a person being not upset about some rumors of persecution of the Jews in Nazi Germany, but fighting for justice (Nazi hunting) after the truth began to come out - in part by their own work after some initial evidence was made known.

Your monicur fits.

Posted by: Jericho on January 24, 2005 01:08 PM
53. Scottd,

O.K. thanks for the clarification on the point you were trying to make. However, I still don't agree with what you wrote in your reply above "Without better data, he can't support statements like the headline of his post ("King County Vote Count Off by More Than 3700 Votes"). My suggestion would be to hold off on such statements until he can back them up with hard data."

Reason being, if you don't blow the horn, then no one will come looking. I think everyone on this blog agrees that the facts haven't come out yet, but if Stefan hadn't blown the horn, we would've all been fooled then...wouldn't have we?

Thanks Stefan for 'blowing the horn'.

Cid

Posted by: CidinDupont on January 24, 2005 01:11 PM
54. Scottd: It seems to that as citizens (not Republicans or Democrats) we should demand both accountability and truth from those who wish to serve us in government. The burden of proof is on the politicos (both sides) to prove the election was honest, not on us to prove it was a cheat. And yes, the Seattle papers are asleep at the switch here. SS should be nominated for a Pulitzer.

Posted by: DCPI on January 24, 2005 01:20 PM
55. scottd, in that RCW it also requires the auditor to keep the record of votes for the previous five votes, and under a seperate rcw 29a.08.540
they must keep the voter registration informaion of cancelled registrations (which includes the record of votes as required in the first rcw I quoted) for a period of 2 years. Shouldn't that include the deleted voters you speak of?

Posted by: ChuckJ on January 24, 2005 01:37 PM
56. There's no way King Count election officials thought that someone like Stefan would challenge, let alone broadcast, their counted (or uncounted) votes.

These guys count on lip service and media silence to get what they want.

Justice is not an issue for them.

Posted by: TADD on January 24, 2005 02:22 PM
57. ChuckJ: Shouldn't that include the deleted voters you speak of?

Yes, it would. The RCW states that cancellation records are stored in a separate file. Has Stefan requested that file?

Posted by: scottd on January 24, 2005 02:37 PM
58. "...the good of the state." Sounds somewhat Stalinistic...hmmmmm

We're going to miss headless lucy after Rossi is restored to his rightful position as Governor. Maybe she/he/it will come back to us as the north end of a south-bound hyena.

Posted by: Susu on January 24, 2005 02:41 PM
59. Gee, I'm away for a few minutes, and fights break out. Everyone breathe, count to ten and smile, please.

scottd (and my fellow Democrats on this board):

If I'm reading you correctly, your objection to the initial post is that the raw data isn't worth the paper that it isn't printed on. That being the case, any interpretation of that data can be no more valid than the source, and that making statements about who "would have won" based on the flawed data is at best nothing more than a math-class exercise, with no real meaning.

Do I have the gist of your point?

Stephen (and the Republicans on this board):

If I'm reading your posts correctly, while acknowledging that the data is so poor as to be almost meaningless, you make the point that it is the best currently available, it is being provided by people that are presumably Gregoire supporters and that your analysis of it gives Mr. Rossi a win.

You also apparently believe that while the analysis is not definitive as to the winner of the election, it is nonetheless interesting and thought-provoking.

Do I have the gist of your point?

Now, assuming I just got a "yes" on both those questions, can we all shake hands and play nice now? (Me included, as I got a bit testy in some of my earlier posts.)

Posted by: John Barelli on January 24, 2005 02:52 PM
60. And could we also ask you to spell Stefan's name correctly? Unless of course you are refering to someone other than Mr. Sharkansky. Good Grief!

Posted by: Susu on January 24, 2005 02:58 PM
61. And could we also ask you to spell Stefan's name correctly? Unless of course you are refering to someone other than Mr. Sharkansky. Good Grief!

Mr. Sharkansky, you have my apologies. Having people misspell my name annoys me, too.

Posted by: John Barelli on January 24, 2005 03:03 PM
62. Scottd,
I don't know if stefan has requested that file. I guess my point is that King County record keeping seems to be pretty sloppy. All the information needed to make a judgement on where the votes came from should be included with the information that Stefan had requested. This information already exits in the public files(as required by law) and obviously King County should have it and give it out when asked (as required by law). If they have this information why then are they unable to follow their own paper trail to give a valid reason for the voterless ballots? Other than to say this happens every election they haven't given a reason, and they do owe us and explanation. It is their JOB to keep track of this information accurately. Stefan has done the best he can with the information he was provided, is it perfect? No, but neither is the information King County has and they seem to be satisfied with "it happens every election". So for me, I will accept what Stefan has to say until King County has the gumption to look into this further themselves.

Posted by: ChuckJ on January 24, 2005 03:19 PM
63. The RCW states that cancellation records are stored in a separate file. Has Stefan requested that file?

Why, yes, I requested it on Jan. 10. I have not received it yet. Do you not grasp the significance of what is happening here "Scott"? The county cannot explain its numbers and it is insufficiently responsive in releasing to the public the information it is required to release. It is not a law abiding entity. It is a rogue agency. I'm starting with the (charitable) assumption that the information which has been provided is valid information, and deriving the only conclusions that can be derived from that information.

If you question my conclusions, then you are inherently questioning the validity of the county's data. If that's the case, then frankly, we're not that far apart. If you are a person of integrity you'll join the rest of us to demand that unless the county can reconcile its numbers, then the election should be tossed overboard.

thanks!

Posted by: Stefan Sharkansky on January 24, 2005 03:32 PM
64. Stefan,
Thanks for the clarification on the question. And well put.

Posted by: ChuckJ on January 24, 2005 03:41 PM
65. ChuckJ:

It seems that there is at least one thing we agree upon: King County has done a monumentally poor job of explaining its procedures and records. Part of that may be defensiveness or sheer buttheadedness -- but that's no excuse. KingCo has a duty to maintain voter confidence in the election system and they have failed in this.

It may even turn out that KingCo has, in fact, materially departed from its legal obligations. So far, that hasn't been demonstrated -- but we may get a better picture as more data are revealed during the election contest proceedings.

However, it's not a solution to respond to incomplete disclosure on the county's part by deliberately misinterpreting the data they have provided. Stefan knows that he does not have a complete list of voters, and now he knows that the potential size of that margin of error is at least as large as the conclusions he is drawing. This adds nothing to the discussion but noise.

"Stefan": No need for the scare quotes. Scott is indeed my name -- but you can just call me scottd if you want to distinguish me from other Scotts on your board.

So far, I don't "inherently question the validity of the county's data" -- and that's not an inevitable corollary to questioning your analysis. It's quite possible that the voter lists they have given you are indeed accurate representations of the voter rolls at the time of their production. As we both know, that data would be insufficient basis for the analysis you have performed.

Posted by: scottd on January 24, 2005 03:58 PM
66. As we both know, that data would be insufficient basis for the analysis you have performed.

Um, no, "Scott". Neither of us know this. You have not even seen the data, so you know nothing about it. I've invested hours of my time researching it and I know that my analysis is consistent with the data that exists. It's very easy for anonymous trolls like yourself to simply make up statements like that without having any factual basis for making them. If you want to do some data mining and challenge some of my conclusions by doing better analysis than I've done, please do. I welecome the opportunity to learn something new from you. Until then, why not go show off at one of the Gregoire fan blogs where your baseless hallucinations will actually raise the quality of the discussion?

Posted by: Stefan Sharkansky on January 24, 2005 04:07 PM
67. Scottd,
I disagree with your assumption that Stefan has deliberaltly misinterperated the data. I also would question you on what to do with the incomplete information given to him. If King County says this is what we have and all we will give you, is he to give up and say ok I won't look at it at all beacause you won't give me what I need. I think Stefan acknowledges the fact that the information is incomplete, and that is what I and many others have a problem with. Saying that incomplete information should not be analyzed is lame at best. Right now that incomplete info is all we have and by reviewing it Stefan adds to the discussion by revieling even more discrepancies in the information provided.

Posted by: Chuck j on January 24, 2005 04:16 PM
68. However, it's not a solution to respond to incomplete disclosure on the county's part by deliberately misinterpreting the data they have provided. Stefan knows that he does not have a complete list of voters, and now he knows that the potential size of that margin of error is at least as large as the conclusions he is drawing. This adds nothing to the discussion but noise.

I think the Shark is going on the assumption that valid voters should appear in at least one version of the data file that he's had. The only reason that a valid voter might not appear in at least one such file would be if the voter registered before the election and the registration was then immediately cancelled before it could show up anyplace.

I suppose it's vaguely conceivable that if somebody moved shortly after an election such a thing might happen. It does seem rather unlikely, though, given how many registrations still exist for people who moved a long time ago. Further, while it would be plausible that this would account for a few of the 'mystery' voters, it seems very unlikely that this would account for any significant portion unless there was something fishy going on.

Posted by: supercat on January 24, 2005 04:18 PM
69. Stefan, how do you know that the 636 duplicate record people didn't vote twice? Thousands of people were mailed duplicate absentee ballots. Some of them may have completed more than one of them. Any discernable pattern to the duplicate record people?

Posted by: Richard Pope on January 24, 2005 04:30 PM
70. Following up on Richard Pope's question. I was thinking today about the duplicate ballots that were sent out. First, the article said that some may have received up to four ballots and the only thing King County did was tell the people to throw away all but one. (sigh)

The news article stated that:
"County officials said the county's election system does not allow more than one vote per person to be counted."

However, even if the ballots could not be counted by the machine, they could be counted in a manual recount.

Am I missing something here?

Posted by: Gil on January 24, 2005 04:43 PM
71. "can we all shake hands and play nice now? (Me included, as I got a bit testy in some of my earlier posts.)"

" It's very easy for anonymous trolls like yourself to simply make up statements like that without having any factual basis for making them."

"why not go show off at one of the Gregoire fan blogs where your baseless hallucinations will actually raise the quality of the discussion?"

still testy, eh.

Posted by: ralph on January 24, 2005 04:44 PM
72. Gil deftly points out:

"County officials said the county's election system does not allow more than one vote per person to be counted."

I can't claim to be an expert, but I think what has been shown over the course of this fiasco is that the County HAS NO SYSTEM other than to ask you to "Please refrain from casting more than one legal vote." Thus, they can make the ridiculously true statement above...

Posted by: smegma on January 24, 2005 04:53 PM
73. ralph determines:

"still testy, eh."

Brilliance and edification lost...

Posted by: smegma on January 24, 2005 04:59 PM
74. "But as fair as I can be, I can't come up with a good explanation for precincts that have more voters than ballots."

One of those precincts is just up the street from me...another is in my sister's area - and yet another is in my brothers neighborhood!. These are somewhat *conservative* communities. Very strange that they would be missing ballots....

Posted by: Deborah on January 24, 2005 05:00 PM
75. Supercat:
I would agree that nearly all valid voters are likely to be shown on the Nov. 2 list. Most of them will have been registered for quite a while.

The problem I pointed out was that it appears that King County deletes roughly 5000 voters per month from its rolls. It's quite possible that a large number of voters on the Nov. 2 list were deleted before the Dec. 29 list, so Stefan would have no way of including them in his reconciliation unless he had a list of those cancellations. This is just one source of error that could easily result in thousands of "voterless ballots".

Attempting to draw inferences from incomplete data is fine, but it's meaningless if you don't make some attempt to quantify the likely sources of error.


Posted by: scottd on January 24, 2005 05:54 PM
76. The problem I pointed out was that it appears that King County deletes roughly 5000 voters per month from its rolls. It's quite possible that a large number of voters on the Nov. 2 list were deleted before the Dec. 29 list, so Stefan would have no way of including them in his reconciliation unless he had a list of those cancellations. This is just one source of error that could easily result in thousands of "voterless ballots".

I don't know exactly how the Shark is figuring people who were in the November 2 file but not in any later one. I guess it would be possible that some of those people voted, but even if one assumed that to be the case I don't think that would explain anything near all of the mystery voters.

Another thing I'm a little puzzled by is how exactly King County is running their databases. The county is required to keep track of whether people voted in the last 5 elections; even if a person becomes ineligible to vote in King County, the county must still keep that information.

It would seem that the logical way to do things would be to have a database keyed on voter name and birthdate, which would indicate where the person was registered, when and how their registration request was made, when it was processed, at what precinct and by what means (if any) they are eligible to vote, and at what precinct and by what means (if any) they have voted in the last five elections. If a voter becomes ineligible, they would be flagged as such (not eligible to vote by any means) but the record left in the database.

Although it would be possible to use a separate database to keep track of people whose registrations have been cancelled, this seems like an overly complicated approach given that it would necessitate searching the cancellation database every time a voter is registered to determine whether it was necessary to import their voting history. In addition, this sort of approach would introduce risks of the databases getting out of sync and would generally be a big mess.

Maybe that's how King County does things--I have no idea. But it would seem that being able to produce a simple unambiguous list of people who actually voted in a given election (and where and how they did so) should be a sufficiently obvious, common, and necessary query that a database should support it inherently. Having to issue separate queries to retrieve some records from each of two or more separate databases introduces a variety of risks unless all queries can be done simultaneously. Otherwise any records that are moved from one database to another between queries could appear twice or not at all.

Given that the county is required to keep records of who voted, as well as where and by what means they did so, and given that the county is required to make such information available to the public, is it unreasonable to demand, based on those requirements, that the county supply that information in some complete form?

Posted by: supercat on January 24, 2005 06:30 PM
77. Stefan:
My evaluation of your methods is based only on what you've reported. Maybe I got it wrong, so I'll ask you directly: Do you have a complete list of voters from the Nov. election -- including voters who have been deleted from the current roll? Or are you working off your own reconstruction of such a list? If it's the latter, how have you accounted for the thousands of voters that are likely to have been deleted between Nov. 2 and Dec. 29? Even if you could come up with their names, how would you know their voting status in the last election?

My assertion has been that the reconciliation you are attempting is meaningless without that list (for obvious reasons), and that the method you've described for reconstructing the list is inadequate. It's inadequate because the magnitude of likely errors is at least as large as the voter/ballot discrepancies you are attempting to demonstrate. Show me where I've gotten that wrong and you'll have my apology -- plus, if you like, a promise to never again stain your blog with my presence.


P.S.: I've spent many years conducting technical reviews and I've found that most people are happy to clarify troublesome points in an otherwise sound analysis. When folks get testy, it's usually because they realize that parts of their analysis are weak. Attempting to divert attention to the questioner instead of the technical issue is lamest of all.

Posted by: scottd on January 24, 2005 06:33 PM
78. Supercat,
Not sure how King County is keeping track now, but things will be changing in 2006. Along with the Help America Vote Act came RCW 29A.08.651. This puts the state in charge of the voter list with a unique id number for each voter. This might help to solve some of these problems in the future.

Posted by: ChuckJ on January 24, 2005 06:43 PM
79. Another thing I'm curious about, btw: what sort of chain-of-custody does King County maintain over records that would indicate who voted? It would seem like maintaining chain of custody over such records would be an essential part of assuring a fair election, but in King County I don't think anything would surprise me.

Posted by: supercat on January 24, 2005 06:54 PM
80. scottd, King County has presented a moving target with their changes to the lists of registered voters and voters credited with voting in the general election. Before you presume that they dropped lots of people from the rolls in December, keep in mind that King County explained their inability to produce anything more than a "preliminary list" at first because they had been so busy with the recounts.

I doubt they did very much of their routine tasks like maintenance of their voter registration lists while they were so busy.

One of their first tasks after the election was to reconcile their records before the next election.

I'm not saying they didn't drop some voters from the rolls, but I wouldn't expect them to have done as much in November and December as they otherwise would have.

Posted by: Micajah on January 24, 2005 07:09 PM
81. reading some earlier items:

John Barelli,

Peace, my friend. I agree, and I know (more or less) where you have been...you continue to bring something (IMO much needed) into the debate.

scottd,

I have such a headache from last night...are you related to torridjoe?

Enough said...

none@none.com is ridiculous, and I bought it...

Posted by: smegma on January 24, 2005 07:18 PM
82. smegma (and others): I didn't think lack of an email address would be an issue. On the other hand, I suppose it's just polite to provide a channel for offblog discussion without chewing up too much of Stefan's bandwidth. Click on my name for updated contact info.

BTW, don't read too much into my posts. I'm not a Gregoire partisan -- I just detest sloppy analysis.

Posted by: scottd on January 24, 2005 07:52 PM
83. Wow..this is amost interesting discussion. Scottd is doing a credible job of digesting data that is provided. Stefan shies away from critical analysis of some of his conclusions.

In other words, as I see it, Scottd is doing to Stefan's analysis what Stefan is doing to the King County folks. It's very interesting to see Stefan's response less than gracious.

I thought it was about the facts, the information. Apparently, only certain facts (imputed or provided) are worthy of above-board scrutiny here. Gives one long pause for all the data provided here.

Thanks scottd, you are doing an incredible service! Stefan..your service is remarkable too as is Goldstein's on horsesass.

But, in the end, as someone quoted John Adams a long time ago in this blog, facts are stubborn things. Hard when they get in the way, eh? Makes you want to rant!


Posted by: jim on January 24, 2005 08:22 PM
84. Smegma! I've enjoyed watching your conservative brethren and sisters line up squarely behind you and defend your right to use that name even though I'm sure many of them would strenuously object on moral grounds were I to use it. By Conservative popular demand we will have Smegma--- and lots more of it!

Posted by: headless lucy on January 24, 2005 08:29 PM
85. "Scott" or Scottd,

BTW, don't read too much into my posts.

Many of us don't I would assume...

I'm not a Gregoire partisan -- I just detest sloppy analysis.

I don't like it too, however, if King County is playing the game of smoke and mirrors then you can't accuse anyone of sloppy analysis. Incomplete, sure.

I will assume that a non-partisan Judge will not give King County the benefit of the doubt if they don't have their act together. Where there is smoke, there is fire. Stefan is following the smoke. I'm quite sure that if King County could provide Stefan the data to refute his numbers he'd be more than willing to analyze it, post it, and even applaud King County for their desire for real election integrity. *caugh* The simple fact is that they haven't come clean...

BTW, you seem to be a very concerned voter. Do a counter analysis. I'm sure [actually, not really, no, not at all] KC still has the information they gave Stefan.

Posted by: VaCSProf on January 24, 2005 08:53 PM
86. I was wondering which conspiracy theory would fit, but Heedless? My own fleshy micro-organism?

Bedtime, silly...

And jim,

"In other words, as I see it, Scottd is doing to Stefan's analysis what Stefan is doing to the King County folks. It's very interesting to see Stefan's response less than gracious."

You fail to take into account that King County deserves it's reputation. scottd is merely a refutation of intelligent thought in the guise of intelligence.

Posted by: smegma on January 24, 2005 08:58 PM
87. scottd,
Your logic is simply flawed. Stefan is indeed working from limited data, but only because King County either refuses to or is unable to provide accurate data.

The central issue is King County's ability to accurately track a database of voters, generate the correct number of ballots from that list of voters, conduct the election and then reconcile the number of ballots to produce at outcome.

The very fact that King County is unable or unwilling to provide a query of the total number of voters as of a given date, especially an accurate number as of Nov. 2nd is the whole reason why we are even having this discussion.

Challenging the validity of the data is a red herring. We know that the data is innacurate. Let's stick to the central issue here, if the data is innaccurate and therefore it's not possible to know the result of the election, then how can we elect a governor?

The fallacies you seem to be arguing from are:
From Ignorance
Anonymous Authority
Converse Accident
Post Hoc
Joint Effect
Insignificant
Begging the Question
Straw Man

and others.

Posted by: Jeff B. on January 24, 2005 09:08 PM
88. JeffB:

Incomplete data is incomplete data...regardless of the reasons for its incompleteness. It limits one's ability to draw precise conclusions, so any conclusions reached need to be accompanied by an estimate of the likely errors. If Stefan were intellectually honest on this point he'd say "King County data show somewhere between zero and 4000 more ballots than voters -- can't really tell from the data I've seen." Of course, that sounds less authoritative -- but that's the actual state of his knowledge.

Challenging the validity of the data is a red herring.

That's pretty funny. Are you suggesting that the validity of the data used to form an argument isn't an issue when evaluating the credibility of the argument?

A+ on your ability to list common logical fallacies. Now let's see if you can apply them to anything I've said.

smegma: I thought we'd made up last night. Ask JeffB where argumentum ad hominem fits in on his list of logical fallacies.

Posted by: scottd on January 24, 2005 09:39 PM
89. scottd,

Don't have time to apply every fallacy here. What I will say is this:

You are pleading From Ignorance as to whether or not Stefan's argument is credible as you yourself admit that the data is flawed. You appeal to Anonymous Authority that there could be from 0 to 4000 ballots, to Joint Effect substantiating Stefan's claim while refuting it. You must like Steely Dan.

However what you say is Isignificant and Begging the Question because the more important question that Stefan is answering is whether we can have any certainty in the result to declare a winner. Your arguments are a Straw Man for the important conclusion that most rational Washingtonions have drawn about the deterministic outcome of the election and King County's ability to conduct it properly.

Lastly, yours is but a Hasty Generalization to Stefan's well thought analysis of given data. Show how you come up with "0 to 4000" as Stefan has shown how he comes up with his numbers.

Posted by: Jeff B. on January 24, 2005 10:18 PM
90. Anonymous troll "Scott" is not so much asking serious questions to challenge my analysis as he is simply asking the same questions over and over again and ignoring the answers that are being delivered.

Did I receive a complete voter list from King County? I requested one on December 28th. http://www.soundpolitics.com/archives/003315.htmlI also requested a list of cancelled voters. They told me that voter records were deleted between Dec. 29 and Jan. 7. They never told me that voter records were deleted between the election and Dec. 29. If they provided me with complete information, then my analysis is valid. If they didn't provide me with complete information then they're in violation of the Public Disclosure Act (why would they conceal information if they have nothing to hide?)

If anonymous troll "Scott" wants to present new information for us to consider, we'll all welcome it. Asking the same questions which have already been answered does not add anything to the discussion.

Posted by: Stefan Sharkansky on January 24, 2005 10:38 PM
91. Here is a theory that needs to be tested. While I have absolutely no confidence in the record keeping and administrative abilities of King County Elections, I think that the vast majority of the voterless ballots could be explained.

The list of people who voted on November 2, 2004 was initially provided matched up against a database of King County registered voters as of December 29, 2004. Then an updated list was provided matched up againjst a database of King County registered voters as of January 7, 2005.

The major problem with using these lists is that the database of King County registered voters used for the November 2, 2004 was generated as of October 2, 2004 -- the cut-off date for freezing the registration rolls 31 days before the election.

(They did allow in-person registration at the King County Elections office until October 18, 2004, but there would have been no cancellations after October 2, 2004 due to people moving out of the county -- only for deaths and felony convictions.)

However, King County cancels over 5,000 registrations during the average month. People move to other counties or states. People die. People are convicted of felonies. Actually, I think the number would be a lot more than 5,000 people, since there are over one million registered voters in this county, and I would think that more than 6% of them in any given year move somewhere else, move to the joint, or move on permanently.

Just using 5,000 people a month, there would be somewhere around 15,000 people who were on the registration rolls on October 2, 2004 who would no longer be on the registration rolls nearly three months later on December 29, 2004.

There are roughly 3,000 so-called voterless ballots in King County. (Since there are also a bit over 800 ballotless voters, this results in a net discrepancies of around 2,150 more ballots than voters. But the ballotless voter problem is a separate phenomenon.)

If someone has the current voter registration files for two different dates, it should be fairly simple (relatively speaking) matter to compare the two files, and see how many people in the first file are not included in the second file. That way, an estimate of the average number of people removed every month can be obtained.

Assuming that 15,000 people were removed from the King County registration rolls between October 2, 2004 and December 29, 2004, it would not be too surprising that 3,000 of these people could have legitimately voted in the November 2, 2004 election. That would be a turnout rate of 20%, compared with the overall county turnout of 83%.

So I will pose the theory that the reason why King County has so many voterless ballots is that thousands of people who voted in the November 2, 2004 general election are no longer on the King County voter registration rolls.

It would be very difficult to test this theory against the absentee ballot or provisional ballot voters. You would have to take all of the empty ballot envelopes and sort them by precinct. Or do your own data entry job from the empty envelopes. (If King County did a complete data entry job from these, you could try to get the data that pertains to now cancelled voters, if that data exists in the first place.)

It would be very easy to test this theory against poll voters. Each precinct has a printed poll book with lots of signatures in it of people who actually went to the polls and voted. (Hopefully, of course, signatures of the actual voters, instead of impersonators.) There is also a numbered list for each precinct, generated by the election worker, as people sign in to vote.

Simply take the five polling locations that have the highest surpluses of poll ballots over apparent poll voters. Request King County Elections to provide a copy of the poll books and numbered voter lists for each precinct for these five polling locations. Count the names on each numbered voter list (very easy) and also in each poll book (more time consuming, but also easy). Compare these numbers of actual signed in voters to (a) the number of poll ballots reported for each precinct and poll location and (b) the number of actual voters listed on the 12/29/2004 and 01/05/2005 lists of actual voters among then currently registered voters.

Posted by: Richard Pope on January 25, 2005 08:21 AM
92. I also have a theory as to why the King County Admin Building polling location has such an apparent surplus of provisional ballots over provisional voters -- 64 of them according to Stefan's analysis.

Military and overseas voters can cast something called a federal write-in ballot. Under state and federal law, this ballot must be counted so long as the person would be qualified to vote, even if they are not registered. I.e. -- the person is at least 18 years old, a U.S. citizen, and last lived in Washington before joining the military or moving overseas. (How all these qualifications can be verified, of course, is another question.)

Sometimes these federal write-in ballots are cast because a registered voter fails to request an absentee ballot or the requested absentee ballot does not arrive in time. Other times they are cast by unregistered voters, but are accompanied by a voter registration form. (There is also a federal post-card type registration form that can be used by military and overseas voter.)

However, in many cases, these federal write-in ballots are cast by eligible voters who aren't registered and don't register concurrently with voting. In such an event, the ballot must be counted (if the eligibility is verified), but the voter is not added to the registration rolls.

Apparently, 251 voters fell into this category for the November 2, 2004 general election in King County.

My theory is that a lot of these 251 voters ended up getting assigned to the King County Elections office and precinct 1823 because there was not enough information provided to allow them to be assigned to a different precinct. Also, that these 251 federal write-in ballot voters (and perhaps all of the federal write-in ballots for that matter) get treated as provisional ballots.

Posted by: Richard Pope on January 25, 2005 08:32 AM
93. I noticed that scottd's farewell post has been deleted.

If this heavyhanded censorship resulted from actions of the operators of this board, then you are truly pathetic hypocrites.

Scottd is guilty of trying to present reasonable objections to your analysis, and you call him names, and avoid dealing with any of the legitimate questions that he has raised. He has behaved politely, whereas you have not.

pathetic.

censorship, the refuge of weak minds.

Posted by: ralph on January 25, 2005 09:10 AM
94. Anonymous troll "Ralph" -- Censorship is when government prevents someone from expressing an opinion. I am not the government and I am not preventing anybody from expressing their opinions. This is my website. I do not have an obligation to operate a free bathroom wall for anonymous people to post baseless allegations and spread disinformation. When I filter out such posts, it's responsible editorial discretion.

Posted by: Stefan Sharkansky on January 25, 2005 09:28 AM
95. I thought ScottD had stepped way over the line in his last posting. I think deleting it was a judgment call that could have gone either way. I don't think I would have deleted it, but I can't fault Stefan for doing so. It was a sound exercise of discretion on his part.

It is pretty obvious that King County doesn't keep their voter registration records worth a darn. Their software and programming is extremely deficient. Nor do they keep records -- such as the cancelled voter records -- in the way that the RCW requires.

Stefan has been quite straightforward in saying what limitations are presented with King County's data. Most of ScottD's postings have been fairly reasoned and polite. Maybe he's had a bad day or two recently or something. But it really wasn't called for when he kept harping on the incomplete data situation, and then kept calling Stefan names over it.

Posted by: Richard Pope on January 25, 2005 11:14 AM
96. "Censorship is when government prevents someone from expressing an opinion. I am not the government and I am not preventing anybody from expressing their opinions."

webster:
censor: to examine in order to suppress or delete anything considered objectionable

"four legs good, two legs bad"
- orwell

Posted by: ralph on January 25, 2005 11:19 AM
97. Nobody should be surprised that this is a right-leaning blog designed to provide fodder for right-leaning arguments. It's like talk radio--facts and data are used selectively to defend points of view.

It's the perogative of the owner to decide what speech can be posted...it's just hard to accept any data or conclusions where free speech, sharing of ideas and information is curtailed. Especially when that information contradicts supposed "fact-based" conclusions.

I encourage all who take the right-wing arguments (or the left-wing arguments from other sources) to do appropriate due diligence. Unfortunately, most of us are too focused on our beliefs/goals to take the data as it is. We rely upon like-minded individuals to feed us data for our arguments.

Tim Eyman was a master of this -- he, of course, took it one step further and ripped off his own supporters expecting their unwavering belief in his stated positions made him above reproach; above fairness checking. Fortunately, others checked. That's the goodness of our system.

Let's the checking continue. Let the debate continue.

Posted by: jim on January 25, 2005 03:27 PM
98. This was Posted by Richard Pope at January 25, 2005 08:32 AM --

Military and overseas voters can cast something called a federal write-in ballot. Under state and federal law, this ballot must be counted so long as the person would be qualified to vote, even if they are not registered. I.e. -- the person is at least 18 years old, a U.S. citizen, and last lived in Washington before joining the military or moving overseas.

So far as I can tell, the Federal Write-in Absentee Ballot cannot be used to vote for state and local offices in Washington.

I searched high and low for any hint in WA law and in the Federal Voting Assistance Guide that the "FWAB" could be used to vote for anything other than federal offices in WA. There is nothing. The Voting Assistance Guide lists 9 states that have authorized its use for state offices in addition to federal offices -- WA isn't one of them.

Here's where I posted what I found in federal and state law:

http://crokersack.blogspot.com/2005/01/washington-election-laws-vs-overseas.html

Oh, and one more thing -- the "FWAB" can be used by overseas military voters who are already registered as voters -- not by those who aren't registered. It serves to allow them to vote in federal elections in the states where they are registered to vote when the absentee ballot they had already requested hasn't arrived in time for them to cast their votes in those federal electiions. I found nothing to indicate that it could be used by people who have not yet registered to vote and have not requested an absentee ballot from the jurisdiction where they are registered.

If anyone finds any law, regulation or notice which even hints that the "FWAB" could have been used to vote for Rossi or Gregoire for governor in Nov. 2004 (in any county other than the 4 who sent it out in October as part of the agreement with the Justice Dep't), I would like to know what you find.

Proving a negative is next to impossible, so I cannot say for certain "there is no law." I can say "I cannot find any such law, and neither has anyone else to my knowledge."

Posted by: Micajah on January 25, 2005 04:34 PM
99. Micajah,

The form is SF-186. John Barelli and I went through an argument over it before, and we both concluded that it is ambiguous, and I still contend that the State is under no obligation to count any State office candidate write-in (including Governor) advocated by the voter.

My stepson was unaware of how to do the state/local write-in, but he did do the Federal...due to the Federal rules, I would assume his votes for those offices were counted (except perhaps, in KC?)...

Don't know if that helps, but that is what I have...

Posted by: smegma on January 25, 2005 04:59 PM
100. jim&ralph of shared computer fame:

"I encourage all who take the right-wing arguments (or the left-wing arguments from other sources) to do appropriate due diligence. Unfortunately, most of us are too focused on our beliefs/goals to take the data as it is. We rely upon like-minded individuals to feed us data for our arguments."

Funny, I recall reading somewhere, in an echo of your statement, that psychiatrists cannot apply their learning to their own maladies. They can diagnose dual-personality disorder in you, but do not recognize the behaviour within themselves.

So, when you proselytize such, I wonder if you are really talking about how little you have applied the magic in your bullet toward yourselves...Or simply are making a diagnosis on another...

Posted by: smegma on January 25, 2005 05:42 PM
101. Richard Pope: Here is a theory that needs to be tested. While I have absolutely no confidence in the record keeping and administrative abilities of King County Elections, I think that the vast majority of the voterless ballots could be explained.
...
The major problem with using these lists is that the database of King County registered voters used for the November 2, 2004 was generated as of October 2, 2004 -- the cut-off date for freezing the registration rolls 31 days before the election.

Interesting theory, but inconsistent with facts. The Nov. 1 file was not generated on Oct. 2. It includes people who registered as late as Oct. 18.

Other problem with theory -- all of the voters listed in the Nov. 1 are stil listed in the Dec. 29 file. No evidence of the phantom cancellations that you cautiously speculate about or that the various anonymous trolls hysterically insist must have occurred.

Posted by: Stefan Sharkansky on January 25, 2005 05:55 PM
102. All-lower-case-letters-smegma ;-)

Yes, Std Form 186 is the Federal Write-in Absentee Ballot.

Here's where you can read an online description of its use, including a list of the 9 states which have expanded its use to include election of state or local officers, rather than only federal offices:

http://www.fvap.gov/services/fwab.html

It can only -- only, no doubt or ambiguity about it -- be used to vote for federal offices, unless the state has authorized its use to vote for state or local offices.

WA law requires that all ballots be in a certain form, and that they all be issued by the county auditor. The FWAB meets neither of those requirements.

Federal law doesn't require the states to accept the FWAB for voting in state and local elections. It only requires them to accept the FWAB for voting for federal offices.

WA hasn't adopted any law or regulation which would allow use of the FWAB to vote for any office other than a federal office.

The Dep't of Justice threatened a lawsuit if WA didn't get its regular absentee ballots in the mail by Oct. 8, 2004. Four counties couldn't do that, so part of the agreement was that those four would mail FWAB's to their overseas voters.

Mailing the FWAB to overseas military voters is silly, since SF 186 is stockpiled overseas just so it can be used for its intended purpose -- to vote in federal elections when the requested regular abstentee ballot doesn't arrive in time.

But, since all ballots must be issued by county auditors, and WA has its own outer envelopes and oaths, apparently DOJ accepted the idea of mailing the FWAB from those four counties, so that they didn't have to disregard more WA laws than need be in order to get the ballots in the mail. (Those 4 counties then mailed the regular absentee ballots a few days later to those same voters, once the regular ballots were printed.)

Perhaps in those 4 counties the FWAB's that were mailed out by the county auditors could have been used to vote in the gubernatorial election, but the news reports didn't say.

In no other county could the FWAB be used by WA voters to vote for anything other than a federal office.

Anyone who claims differently needs to identify the law which says the FWAB can be used in WA to vote for state or local offices.

The current law is contained in 42 USC Sec. 1973ff-1:

Each State shall -

(1) permit absent uniformed services voters and overseas voters to use absentee registration procedures and to vote by absentee ballot in general, special, primary, and runoff elections for Federal office;

(2) accept and process, with respect to any election for Federal office, any otherwise valid voter registration application and absentee ballot application from an absent uniformed services voter or overseas voter, if the application is received by the appropriate State election official not less than 30 days before the election;

(3) permit overseas voters to use Federal write-in absentee ballots (in accordance with section 1973ff-2 of this title) in general elections for Federal office; and

(4) use the official post card form (prescribed under section 1973ff of this title) for simultaneous voter registration application and absentee ballot application.

-SOURCE-

(Pub. L. 99-410, title I, Sec. 102, Aug. 28, 1986, 100 Stat. 925; Pub. L. 107-107, div. A, title XVI, Sec. 1606(a)(1), Dec. 28, 2001, 115 Stat. 1278.)


Here's 42 USC 1973ff-2, which is cited in the above section:

Sec. 1973ff-2. Federal write-in absentee ballot for overseas voters in general elections for Federal office

-STATUTE-

(a) In general
The Presidential designee shall prescribe a Federal write-in absentee ballot (including a secrecy envelope and mailing envelope for such ballot) for use in general elections for Federal office by overseas voters who make timely application for, and do not receive, States, absentee ballots.

(b) Submission and processing
Except as otherwise provided in this subchapter, a Federal write-in absentee ballot shall be submitted and processed in the manner provided by law for absentee ballots in the State involved.

A Federal write-in absentee ballot of an overseas voter shall not be counted -

(1) if the ballot is submitted from any location in the United States;

(2) if the application of the overseas voter for a State absentee ballot is received by the appropriate State election official less than 30 days before the general election; or

(3) if a State absentee ballot of the overseas voter is received by the appropriate State election official not later than the deadline for receipt of the State absentee ballot under State law.

(c) Special rules

The following rules shall apply with respect to Federal write-in absentee ballots:

(1) In completing the ballot, the overseas voter may designate a candidate by writing in the name of the candidate or by writing in the name of a political party (in which case the ballot shall be counted for the candidate of that political party).

(2) In the case of the offices of President and Vice President, a vote for a named candidate or a vote by writing in the name of a political party shall be counted as a vote for the electors supporting the candidate involved.

(3) Any abbreviation, misspelling, or other minor variation in the form of the name of a candidate or a political party shall be disregarded in determining the validity of the ballot, if the intention of the voter can be ascertained.

(d) Second ballot submission; instruction to overseas voter

An overseas voter who submits a Federal write-in absentee ballot and later receives a State absentee ballot, may submit the State absentee ballot. The Presidential designee shall assure that the instructions for each Federal write-in absentee ballot clearly state that an overseas voter who submits a Federal write-in absentee ballot and later receives and submits a State absentee ballot should make every reasonable effort to inform the appropriate State election official that the voter has submitted more than one ballot.

(e) Use of approved State absentee ballot in place of Federal write-in absentee ballot

The Federal write-in absentee ballot shall not be valid for use in a general election if the State involved provides a State absentee ballot that -

(1) at the request of the State, is approved by the Presidential designee for use in place of the Federal write-in absentee ballot; and

(2) is made available to overseas voters at least 60 days before the deadline for receipt of the State ballot under State law.

(f) Certain States exempted

A State is not required to permit use of the Federal write-in absentee ballot, if, on and after August 28, 1986, the State has in effect a law providing that -

(1) a State absentee ballot is required to be available to any voter described in section 1973ff-6(5)(A) of this title at least 90 days before the general election involved; and

(2) a State absentee ballot is required to be available to any voter described in section 1973ff-6(5)(B) or (C) of this title, as soon as the official list of candidates in the general election is complete.

-SOURCE-

(Pub. L. 99-410, title I, Sec. 103, Aug. 28, 1986, 100 Stat. 925.)

Posted by: Micajah on January 25, 2005 06:29 PM
103. Micajah,

The RCW authority for counting absentee ballots from military and overseas eligible voters who are not registered is contained in RCW 29A.40.091. State law allows this, which is sufficient. Whether or not federal law requires this is not relevant.

However, if they don't submit a voter registration application, then their vote is only counted in that election and they are not registered for subsequent elections.

RCW 29A.40.091
Envelopes and instructions.
The county auditor shall send each absentee voter a ballot, a security envelope in which to seal the ballot after voting, a larger envelope in which to return the security envelope, and instructions on how to mark the ballot and how to return it to the county auditor. The instructions that accompany an absentee ballot for a partisan primary must include instructions for voting the applicable ballot style, as provided in chapter 29A.36 RCW. The larger return envelope must contain a declaration by the absentee voter reciting his or her qualifications and stating that he or she has not voted in any other jurisdiction at this election, together with a summary of the penalties for any violation of any of the provisions of this chapter. The return envelope must provide space for the voter to indicate the date on which the ballot was voted and for the voter to sign the oath. A summary of the applicable penalty provisions of this chapter must be printed on the return envelope immediately adjacent to the space for the voter's signature. The signature of the voter on the return envelope must affirm and attest to the statements regarding the qualifications of that voter and to the validity of the ballot. For out-of-state voters, overseas voters, and service voters, the signed declaration on the return envelope constitutes the equivalent of a voter registration for the election or primary for which the ballot has been issued. The voter must be instructed to either return the ballot to the county auditor by whom it was issued or attach sufficient first class postage, if applicable, and mail the ballot to the appropriate county auditor no later than the day of the election or primary for which the ballot was issued.

If the county auditor chooses to forward absentee ballots, he or she must include with the ballot a clear explanation of the qualifications necessary to vote in that election and must also advise a voter with questions about his or her eligibility to contact the county auditor. This explanation may be provided on the ballot envelope, on an enclosed insert, or printed directly on the ballot itself. If the information is not included, the envelope must clearly indicate that the ballot is not to be forwarded and that return postage is guaranteed.

Posted by: Richard Pope on January 25, 2005 06:38 PM
104. Stefan,

The November 1 file would have been updated that day by a careful elections office -- or maybe even by King County's elections office -- in order to try to update the precinct poll books.

I don't know when the poll books were printed, but somehow they need to make an effort to put a note in them whenever a voter has obtained an absentee ballot. They would, then (I think), have a reason to update their database right on up to the day before election day in order to have that information.

What I cannot understand is why King County doesn't seem to have maintained a list of the people they believed were eligible to vote in the Nov. election.

From what you've said, unless I've misunderstood, it seems King County began to make changes to their computerized list of registered voters and cannot readily produce a list of those eligible to vote on Nov. 2.

If that's correct, then that's strange. What did they use (if anything) to process absentee ballots that were received after Nov. 2 and ensure that those ballots weren't coming from people who had already voted? It seems that preventing double-voting requires an updated, annotated list of the people eligible to vote -- that is, annotated when the first ballot is received from each voter, so that the second and third ballots can be rejected.

So, where is that list of people eligible to vote, much less the updated, annotated list of voters on the list who were credited with voting?

Those are, I imagine, rhetorical questions which will never be clearly answered by King County.

Posted by: Micajah on January 25, 2005 06:46 PM
105. Micajah,

It wouldn't matter whether or not the Federal Write-In Absentee Ballot could be or was used to vote for state and local offices. We are trying to compare the total number of ballots counted (which includes FWAB's) with the total number of actual voters (which does not include 251 non-registered voters who cast FWAB's). Whether or not these people voted for Governor is not important.

Apparently, Washington has a very liberal and inclusive practice towards counting votes. There is absolutely no state law which says that an FWAB cannot be used for vote for state and local office or for ballot measures. Obviously, federal law does not mandate this -- only requires they be counted for federal office. Nor does the FWAB list any state or local office. However, the administrative practice in Washington, or at least in King County, appears to be to count state and local votes that are written in on these ballots.

http://archives.seattletimes.nwsource.com/cgi-bin/texis.cgi/web/vortex/display?slug=ballots20m&date=20050120&query=carol+cassady

King County military and overseas voters don't really seem to be Republican leaning, from what the GOP observers said in this article.

Military and overseas voters from other counties, especially conservative counties, and those with military bases, are far more Republican leaning, perhaps heavily so in certain counties.

Did these other counties count the FWAB votes for state and local offices, including the straight party votes, and the party offices for particular offices?

Remember, the FWAB cannot be counted by tabulation machines, and must be duplicated onto a regular ballot form. The ballot duplication logs and original FWAB's would have to be inspected.

This could give Rossi a net gain of over 129 votes, if military base counties such as Pierce, Kitsap, Island and Spokane did not count straight party and state office votes on their FWAB's.


Posted by: Richard Pope on January 25, 2005 06:51 PM
106. Richard Pope,

Say that again?

How does this sentence authorize the use of any ballot other than an absentee ballot which "has been issued" by the county auditor?

For out-of-state voters, overseas voters, and service voters, the signed declaration on the return envelope constitutes the equivalent of a voter registration for the election or primary for which the ballot has been issued.

I don't see how that language could be read as authorizing the use of the FWAB to vote for WA state or local offices.

I didn't cite and quote the federal law because I thought it prohibited its use. I cited it to prove that federal law doesn't require use of the FWAB in anything other than federal elections.

Posted by: Micajah on January 25, 2005 06:54 PM
107. Richard Pope,

I went back and read RCW29A Chapter 40 again, and it doesn't even hint that the FWAB can be used to vote for state and local offices in WA.

Here's its URL:

http://www.leg.wa.gov/RCW/index.cfm?fuseaction=chapterdigest&chapter=29A.40

Even the section in which you picked out that one sentence says in its own first sentence:
The county auditor shall send each absentee voter a ballot, a security envelope in which to seal the ballot after voting, a larger envelope in which to return the security envelope, and instructions on how to mark the ballot and how to return it to the county auditor.

It's obvious where the ballot is coming from -- the county auditor, not a stockpile of FWAB's overseas.

The use of the signature as the equivalent of voter registration is simply WA's way of implementing the strangest law in the land: Any "out of state voter" who isn't registered to vote in WA can vote by absentee ballot.

But, those non-registered out of state voters still have to request an absentee ballot from the county auditor.

Now, I obviously don't know what King County is doing; and I certainly wouldn't be surprised to find that they don't know the laws which govern their jobs (even more disgusted, yes, but not surprised). But, if they are counting the votes on FWAB's for state and local offices then they are acting without any legal authority that I can find.

The reason I'm trying to find the answer is that there ought to be a plain English answer that everyone can read and understand.

Reading into that one sentence about the signature a meaning which isn't there won't give us a rule that everyone in the state can be expected to follow. And, for those overseas who look at the Federal Voting Assistance Guide, they won't have even a chance of concluding that they can use the FWAB in WA for anything other than voting on federal offices.

None of this may matter at all for the Nov. 2 election, but I would like to see it adopted as state law (in plain English) that the FWAB can be used when the requested absentee ballot hasn't arrived in time.

Posted by: Micajah on January 25, 2005 07:38 PM
108. Smegma, as you poke fun at me about people checking facts, please take a little time to get your facts right.

You say Ralph and I are of "shared computer fame".

I don't know Ralph, never met him and certainly have never shared a computer with him. You are probably referring to Frank (from a different posting). A little time spent fact checking would have helped make, instead of break, your point. But that underscores the problem with many of the postings here. Too little fact checking, too much bloviating. And when facts get checked (and question the poster), the poison comes out.

To your content...you talk, in humorous ways, about whether I apply the same logic of fact checking to myself (I presume that's what you meant). Please re-read my post (and the part you quoted): I suggested those on the right AND the left do such fact checking. I have no illusions. And I stated such.

Again Adams: "Facts are stubbon things."

Posted by: jim on January 25, 2005 08:17 PM
109. Micajah,

I emailed Sam Reed over a week ago asking whether the FWAB could be used for anything other than federal offices. I received the following response today from Pam Floyd, assistant elections director. Like you I can find nothing written anywhere that specifically authorizes this ballot to be used for state and local offices, so it would be interesting to find out how many of the 1342 FWABs that King county received contained votes for governor (and also how many were counted).

Subject: RE: 2005-231: Does Washington State Allow Federal Write-in Absentee Ballots to be used for State Offices?


Mike,
Yes, Washington counties will accept and count any races that are written in by the voter on the FWAB or our own special ballot. I do not know why the modifying language is not in the guide, but that publication is only a guide. We have always accepted the FWAB, and any write-ins that a voter added. If I am able to edit that wording in the future, I will be happy to do so.

We cannot add items to this ballot (FWAB), such as candidates names, as most of the military using the FWAB obtain it from another source such as FVAP or their voting officer within their unit, rather than from the county elections office.

You ask if someone enters "Governor D" as a write-in, would it count. That would be a decision for the canvassing board. A staff person would not make that decision. The statute says specifically that a candidate's name will be written in, so I suspect that it would not count, but again, it would be a canvassing board decision.

Let me know if you have any other questions. We are always interested in improving processes whenever possible and appreciate your input.
Pam

Posted by: MCL on January 25, 2005 09:14 PM
110. MCL,

I e-mailed the SecState elections division a couple of weeks ago, and didn't get a reply. You clearly have more pull than I. :)

What "guide" is Pam Floyd talking about when she says she doesn't know why the "modifying language" isn't in it?

Maybe there's no language in whatever the "guide" is, because there is no law authorizing use of the FWAB to vote for state and local officers in WA. That would be the obvious reason for its absence from the "guide."

Granting that the guide is only a guide to the law's contents, what's the law?

I came across this page on the SecState's web site:

http://www.secstate.wa.gov/elections/military.aspx

It, of course, says nothing about the FWAB.

I had asked in my e-mail to their office for a cite to any law or regulation that authorizes use of the FWAB to vote for state or local offices in WA. Maybe I scared them off by letting them know I had already searched the RCW, WAC, and Washington State Register with no success.

It is so odd that even the SecState's office doesn't give a damn whether there is any law, regulation, or written instruction that they can point to as authority for what they think is happening.

How do they get the word around to county elections officials to be sure that they all do the same thing?

How many other unwritten "rules" are some of them following?

How is anyone supposed to know what the rules are, if the people who have the power to issue the rules don't bother to do their job?

Their motto must be: "Rules? We don't need no stinking rules!"

Posted by: Micajah on January 25, 2005 09:40 PM
111. Micajah,

I believe that Pam was referring to the "Voting Assistance Guide" on the Federal Voting Assistance web site (http://www.fvap.gov/pubs/vag.html). Chapter 3 has specific absentee registration and voting procedures for each state.

Allowing the FWAB to be used for state and local offices appears to be a policy as opposed to a specific written law (unless somebody can find one). I'm guessing that it may not have been clear to a lot of folks using this ballot that they could write in a whole list of state and local candidates. The question that I am still trying to figure out is how many of these ballots had votes for state offices (governor in particular), and if so, were they counted consistently state wide. King County had something like 9% of their overseas military ballots returned via the FWAB, which likely means that these voters did not get their normal absentee ballots. The FWAB ballot has specific lines for President, Senator and Representative, but the remainder of the ballot is just blank lines.

Posted by: MCL on January 25, 2005 10:20 PM
112.
MCL (and Richard Pope)

I asked again and got this reply from Pamela Floyd, the Assistant Elections Director at the SecState's office:

State law does not address the use of the Federal Write-In Ballot. But federal law does apply during a federal election.

However, we do have a statute that addresses the state's special write-in ballot that is available 90 days before the election. That law is RCW 29A.40.050. These ballots may be used for any race or any issue. Given that statute, the counties accept all races on either write-in ballot that may be used by a voter.

So there we have it: There is no law in WA that authorizes use of the FWAB to vote on state or local offices.

However, they don't need no stinking rules -- they can just "read into" the statute that pertains only to the special absentee ballot issued by the county auditors an authorization to accept a ballot which is neither issued by the auditor nor the same as the ballot issued by the auditor.

And, to add insult to injury, they don't put what they believe is happening into writing for the information of all.

Posted by: Micajah on January 26, 2005 12:58 PM
Post a comment
Name:


Email Address:


URL:


Comments:


Remember info?