The King County Elections scandal, along with its many other disturbing facets, is a kind of an accounting scandal, not unlike Fannie Mae or Enron. Because the numbers simply don't add up. After three "certified" vote counts, the numbers have never added up and they never will add up. Even after they cook the books / "bake the cookies" and "finalize" the voter list, the numbers are not going to add up. [See the extended entry for the apropos source of the phrase "Enron Math"!]
I cannot say whether this accounting scandal is attributable to deliberate fraud like Enron was. At the very least it looks like willful ignorance if not criminal negligence. Simply put, none of the three counts should have been certified by the canvassing board. Let's take a look at the polling place ballots for a manageable example.
This table shows how the count of polling place ballots has changed over the three vote counts vs. the number of polling place voters identified in the county's voter list.
| Count | Polling Place |
|
| Initial Count | 305,583 | |
| Machine Recount | 305,616 | +33 |
| Manual Recount [est.] | 305,636 | +20 |
| Voters | 303,834 | (1,782) |
I say that the number of polling place ballots for the manual recount is approximate, because the final precinct canvass does not break down the totals into absentee, polling place and provisional numbers as did the canvasses of the earlier counts. This is also consistent with a recount worker's report that all three types of ballots were commingled during the recount. [see the extended entry for the recount worker's e-mail]. This commingling makes it a lot harder to discover and/or explain any discrepancies, and therefore would help conceal errors or fraud.
Most serious is the discrepancy of 1,800 or so more polling place ballots than polling place voters. If we compare voters and ballots precinct-by-precinct, is revealed to be even greater than 1,800. We don't know exactly how many polling ballots per precinct there were in the hand count, so we'll compare the precinct voters with the precinct ballots from the machine recount. This is close enough to illustrate the various counting problems.
The punchline is that there are 1,081 precincts with more polling ballots than polling voters, for a total of 2,644 voterless ballots. On the other hand, there are 445 precincts with more voters than ballots, for a total of 862 ballotless voters.
Some of the errors that cause these discrepancies can be identified and classified:
1. Ballots attributed to the wrong precinct
The two precincts with the largest discrepancies on either end happen to be Seattle's 37-1825 and 37-2686. As it turns out, they share a polling place. Here's how the numbers came out:
| Precinct | Absentee Ballots |
Absentee Voters |
Provisional |
Provisional Voters |
Polling Ballots |
Polling Voters |
| 37-1825 | 290 | 293 | 23 | 18 | 156 | 108 |
| 37-2686 | 248 | 252 | 26 | 30 | 104 | 145 |
2. Ballots misattributed by type.
This is the unincorporated "DENVER" precinct near Federal Way
| Precinct | Absentee Ballots |
Absentee Voters |
Provisional |
Provisional Voters |
Polling Ballots |
Polling Voters |
| DENVER | 315 | 314 | 36 | 13 | 148 | 169 |
3. More ballots than voters for no apparent reason
| Precinct | Absentee Ballots |
Absentee Voters |
Provisional |
Provisional Voters |
Polling Ballots |
Polling Voters |
| 37-1563 | 133 | 132 | 24 | 22 | 64 | 18 |
4. Loosey-Goosey Counting
This is Seattle precinct 43-1976
| Count | Absentee | "Add-on"/ Provisional |
Polling |
Total |
| Initial Count | 272 | 82 | 225 | 579 |
| Machine Recount | 276 | 82 | 226 | 584 |
| Manual Recount |
? | ? | ? | 581 |
| Voters | 272 | 72 | 219 | 563 |
The most serious problem is that there are at least 3,539 ballots for whom there are no identified voters. Even after you adjust for the most obvious misattributions, there are still thousands of unexplained discrepancies in the vote count. King County says that the voter list is preliminary and will be "finalized" this week. They'll probably be able to attribute some more votes to some more voters, but not all of the errors I've identified are fixable. The incorrect attribution of ballots to precincts can not be changed at this point and we'll never be able to reconcile the numbers. Although some of the above accounting foul-ups are merely internal reconciliation errors, they are still serious, and they make it more difficult to detect and correct outcome-changing tabulation errors and fraud. They also point to underlying systemic problems, such as poorly trained and incompetent workers. This adds weight to the perception that the canvassing board and elections management are a bunch of political hacks who are more interested in teasing out a partisan outcome than they are in any sort of quality control.
It's an Enron style accounting mess. The numbers are meaningless at this point. Only a fool or a partisan Democrat could have any confidence that these numbers are good enough to tell us who won the election. Throw them out and give us a new election. With FEC supervision of the King County revote.
Here's the op-ed where I found the term "Enron math":
If you haven't come across it yet, prepare yourself for the new math -- Enron math.it concludes:The theory is simple: Any way you tally the numbers, they don't add up for Northwest residents.
So we expect a long, hard fight before we find the truth about Enron. In the meantime, millions of people will pay higher costs without knowing if those costs are just and reasonable. And investors, who followed all the rules, will have been ripped off just like a naïve person who lost money in a pyramid scheme.Substitute "King County Elections" for Enron and both the lead and conclusion still work. The author of this op-ed? Attorney General Christine Gregoire!That's Enron math.
...
The email from a manual recount worker:
I was a Republican counter in King County during the recount. You probably already know but if not this info might be helpful. We first counted the poll ballots (we did the absentee and poll ballots separately). When the counters received the box of ballots we also received an envelope of the provisional ballots for that precinct. At that point the provisional ballots were mixed in with the poll ballots. During our training quite a few of us mentioned that the provisional should not be mix in with the poll ballots. King County did not care.Posted by Stefan Sharkansky at January 02, 2005 11:35 PM | Email ThisAlso another problem I brought up is that I noticed that some absent ballots were mixed in with the poll ballots. Poll ballots use black ink for the precinct number and absentee ballots use red ink. King County will tell you that this is normal to see absentee ballots in with poll ballots. What King County means is that county employees would use extra absentee ballots to make duplicate ballots for whatever reason so the ballot would be read by the machine. Remember that these "dup" ballots were filled out by King County employees. The King County employees would know how to fill out the ballot so the machine would read it. Guess what, some of those absentee ballots in the poll ballots were not filled out correctly and definitely looked like they were filled out by someone other than a county worker. These same ballots did not have red ink on them as well. Whenever a King County employee altered/enhanced a ballot they had to write on it with red ball point pen. These absentee ballots in the poll ballots did not have red ink on them. So in my opinion the poll ballots were simply compromised. Yes the rumors we heard as counters said that the numbers were off on the light side. No one seemed to have a good explanation as to why.
I can't believe that any Democrat can look at this info with a straight face and say that there's not a real possibility of something that at bare minimum, swayed the election.
I don't see how King County is going to explain away all of this, and even if there's not strong enough evidence to win a court case, there's more than enough evidence to win in the court of public opinion.
There is no way that the media will be able to keep the lid on all of this, especially with bloggers going full blast, and especially when the national blogs pick up on this story after King County releases any final numbers.
And there is no way that any rational Washingtonion will have even the slightest trust or confidence in Gregoire if she manages to slither in to office.
If you are a Democrat and you are reading this, sign the petition at www.revotewa.com because your going to need this revote to keep your trust afloat when the masses realize what King County has done and turn against your party.
Posted by: Jeff B. on January 3, 2005 12:06 AM"The punchline is that there are 1,081 precincts with more polling ballots than polling voters, for a total of 2,634 voterless ballots. On the other hand, there are 445 precincts with more voters than ballots, for a total of 862 ballotless voters."
2,634 minus 862 is 1,772 voterless ballots. However, this should be 1,782 voterless ballots, based on 305,616 ballots from the machine recount, less 303,834 actual poll voters.
Unlike King County's ballot and voter accounting system, there is absolutely no way your computer analysis should be off by one ballot, much less ten.
Posted by: Richard Pope on January 3, 2005 12:34 AMhttp://polipundit.com/wp-comments-popup.php?p=5838&c=1#comments
See #5 and subsequent. I think Paul Hooson has met his match.
Posted by: max on January 3, 2005 12:56 AMLet's see if a tiny url link to PoliPundit will work as a live link. Otherwise, nevermind....
Posted by: max on January 3, 2005 01:00 AMStefan has corrected 2,634 voterless poll ballots to 2,644 voterless poll ballots and the numbers balance again.
Keep in mind that Stefan has only analyzed poll ballots. The 1,782 net voterless poll ballots only account for a little over half of the 3,539 net voterless ballots in King County.
Stay tuned as the Shark tears his teeth into the fraudulent numbers regarding absentee ballots, and provisional ballots ...
Posted by: Richard Pope on January 3, 2005 01:19 AMEven WITHOUT any additional discrepancies the Shark may lay bare WRT absentee and provisional ballots (you can bet there will be more to come), I would respectfully propose consideration of and comment on the following:
Just this "gang of 3506" polling ballot errors alone; that should have been fully investigated and addressed by the KC canvassing board PRIOR to final certification; would seem to be ample justification under the existing RCWs for someone to file an election contest in State court and / or an equal protection case in Federal court. Since any registered voter can file, AFAIK no reason it has to wait for the Rossi campaign or the WSRP to do it; and Rossi / WSRP options would not be prejudiced by others filing first.
Might it be a good idea for "somebody" to file an election contest or an equal protection action as soon as possible, using data already dug up by the Shark for starters ??... Would a potential benefit of doing so perhaps be to give the WSRP more time to grind out details; while giving our (R) minorities in the State House and Senate some ammunition to try and delay formal certification of That Woman as Gov by the Legislature ?? Everybody is talking about the possibility of there being an election contest, but until somebody actually FILES one, talk is all there is.
One reason I mention the above: I got email reply from a serving member of the Legislature indicating that when certification of the race for Gov comes before lawmakers for approval in a joint session on 11 January, there will be a serious effort to defer that action. Given (D)s are in control of both houses that will likely be a steep hill to climb, but I am encouraged that there are legislators willing to push it.
Anyway: Only 8 days and counting until 11 Jan. If there is going to be any kind of a case officially on file prior to that 11 January vote, it's gonna have to be filed by somebody real soon...
Anybody know a good lawyer familiar with this area ?? I would be willing to kick in $100 for starters to support a near-term filing. I'm guessing that it would not take a huge number of $100 donations before at least an initial filing could be covered (actually pursuing a case to a decision would of course be a very different story).
Methow Ken
In several counties in Florida, 77% of registered voters are Democrats. Yet anywhere from 90% Republican to 142% Registered Republicans voted.
If 77% of a county is registered Democrat, then how did 142% of the vote come from Registered Republicans? The Registered Republicans should had been 32% not 142%.
If you are un-willing to investigate Voter Fraud in Florida, then all I can say to you is this...
Get over it, Move On.
What in hell does Florida have to do with the election of a governor in Washington? Inquiring minds want to know.
Keep up the good work Shark! Thousands and thousands are watching this fiasco unfold and rooting not necessarily for Rossi, but for justice.
Posted by: Chuck on January 3, 2005 05:17 AMShe should step down immediately.
Posted by: South County on January 3, 2005 05:56 AM"It's not just that marks on paper are being interpreted differently from one count to the next, it's that entire pieces of paper materialize, vanish and reappear, over and over again as if there's no baseline for reconciliation, such as the number of voters who voted in that precinct.
The most serious problem is that there are at least 3,539 ballots for whom there are no identified voters. Even after you adjust for the most obvious misattributions, there are still thousands of unexplained discrepancies in the vote count. King County says that the voter list is preliminary and will be "finalized" this week."
Looking (again) at just the last sentence, I ask: "How is it the King County Board reported this as a certified count and Sam Reed verified their assertion?
MAGNUM SERPENTINE - You've got some pretty heavy stats there. I suppose you can provide the back up numbers?
I think that's a point which needs to be hammered over and over again.
Posted by: ScottM on January 3, 2005 06:40 AMDitto for the Attorney General's (!) office. With emphasis on the *independent* part.
Posted by: Boonie on January 3, 2005 06:41 AM"If you are un-willing to investigate Voter Fraud in Florida, then all I can say to you is this...
Get over it, Move On."
We are residents of Washington State. It would be prudent to take care of issues in our own state that affect us right now, instead of trying to take care of other state's issues before we take care of our own. This issue affects us in so many ways, and here you try to grasp at ropes of sand with the Florida Fraud issue "card". Geez, how could you want to deal with Florida Fraud, when you don't even want to deal something in your own home state! F.Y.I., I am a Libertarian, and if the tables were turned around, I would want the same investigation, same possible revote due to fraud on Rossi's part. If it were a Libertarian candidate guilty of such, I would demand an investigation and revote. Fraud is not acceptable in any instance.
Your being selective as to dealing with potential fraud is saddening.
And, as Richard has pointed out, there is no fraudulent over-voting by Republicans in Florida. For decades most democrats vote for republican presidential candidates there, particularly in Florida's northern counties.
your genius do not stop doing what your
doing what your doing.the dems are scared
to death of you.Keep it up!!
your a genius do not stop doing what your
doing what your doing.the dems are scared
to death of you.Keep it up!!
What in hell does Florida have to do with the election of a governor in Washington? Inquiring minds want to know.
Nothing at all, of course. Magnum just demonstrates the urgency with which partisan tools hope to change the subject before too many people catch on.
Posted by: Matt J Kurlander on January 3, 2005 09:16 AMThose Florisa counties are dixiecrat counties. They register Democrat & have been voting for Republicans for years. Even Bob Dole won in these counties back in '96.
There is plenty of information on this subject if you take the time to research things rather than re-spout factless loony conspiracy theories.
Posted by: Kingdome on January 3, 2005 09:34 AMChange the system so that each precinct or county counts it's votes, but that all counts are kept secret until everyone is finished counting.
Posted by: flenser on January 3, 2005 09:40 AMaren't we all glad we didn't just shut up and go away like all the big mouths in the media suggested? Nice job Schram, nice job Times, reminds me I have a phone call to make.
press on
speaking of not shutting up as the media suggested, at 4wethepeople.com/wa.htm you can order orange ribbon stickers today, the orange magnetic ribbons are being worked on. No, we will not shut up.
I'm guessing the phone call you had to make was to cancel a subscription to a news paper?
Press On, we will. Don't forget rally outside of Ft. Lewis, tommorrow @ 4p.m. and next Tuesday the 11th, in front of the capitol (wear orange), time to be announced.
Orange Robyn
Posted by: Orange Robyn on January 3, 2005 09:51 AMthis is a good point ctmohr brings up. does anyone know the answer to this??
Posted by: darcy on January 3, 2005 09:52 AMTwo reporters from one of Florida's major newspapers spent three days recounting the votes. The results? Bush did indeed win those areas by a huge margin, fair and square. The democrats in that area are dixiecrats who apparently, in large number, could not stomach the liberal senator from Massachessets.
It is unfortunate that Mr. Serpentine and others choose to ignore the findings of that particular recount and continue to make wild and baseless charges. There is absolutely no comparison between the situation that occurred in Florida and the apparent fraud in King County and they know it.
Posted by: nora on January 3, 2005 10:22 AMAs I've heard, the prior elections director quit after repeated serious mistakes(voters not receiving a ballot, voters receiving the wrong ballot, etc.)
Also, as I've been told, the prior elections supervisor was fired when she was caught trying to cover-up mistakes and then lied about why the mistakes occurred.
Posted by: Susu on January 3, 2005 10:34 AMOne question that comes to mind. As hopelessly muddled as the voter/ballot counts seem in King county, are most other counties in the state able to reconcile their totals?
Posted by: David Bolotin on January 3, 2005 10:45 AMI'm not sure how many precincts and ballots they counted in the Florida panhandle, but it does make one wonder how KC could be so slow.
Also, it took KC seemingly forever to count in counts 1, 2 and 3. Yet, the day after the Supreme court decision, the 3 person canvassing board counted and analyzed 700 something ballots in 4 hours or so when they knew they were racing towards their apparent victory. Were the other counts slowed down in a calculated way to ensure King County would be able to report last, and, if needed, adjust the way in which they were counting to guarantee Gregoire's victory?
Just another thing that makes you go "hmmmmmm".
Posted by: nora on January 3, 2005 11:06 AMNot everyone returns all five of the punch cards in Franklin County. I think they get the number of ballots cast from the number of people returning the first punch card, which may or may not be the most popular one.
If you wanted to know the number of actual people voting in Franklin County, you would get the list of actual voters. Since not everyone returned the punch card with the Governor's race on it, there should actually be fewer ballots counted for Governor, than actual people voting.
Posted by: Richard Pope on January 3, 2005 11:11 AMIn San Mateo County, where I worked as a precinct officer, each precinct has its own ballot box. Voters are given ballots and are supposed to drop the ballots in the ballot box themselves. Polling places with multiple precincts therefore have multiple ballot boxes. If it's slow, it's easy to ensure that they drop the ballot in the right box; if it isn't slow, it's terribly difficult. We had *two* ballots dropped in the wrong box in our polling place. (We discovered it in the audit at the end of the night, and notified the county elections department in our reports - so there's no excuse for the lack of discovery).
I don't understand how polling place ballots could be misclassified as provisionals. It's trivially easy to misclassify provisionals as polling place ballots, however: since most voters (but not provisional voters!) are supposed to drop the ballots in the box themselves, provisional voters can - and do - accidentally forget to return the provisional ballots to the poll workers, and instead
Posted by: aphrael on January 3, 2005 12:43 PMYou don't seem to have completed your thought, which is just as well because you seemed to be headed in the wrong direction. When provisional ballots are handed out, they are folded by the election judge which causes them to be rejected by the Accuvote machine. The only way for a provisional ballot to be counted is to be placed into a security sleeve, slipped into a provisional ballot envelope, and put in the side pocket of the big black box.
Posted by: Huckleberry on January 3, 2005 12:57 PMEach ballot has a bar-code that uniquely indentifies the precinct. When you feed them through a machine, it doesn't matter where (which precinct) they are from - the machine counts them without problem.
In the hand count, all these ballots must be separated by people whose eyes are ready to burst into flames from looking at thousands of ballots every day, and then summarized accordingly.
Machine - good. Hand - bad.
Posted by: Larry on January 3, 2005 01:00 PMJust when I'm thinking there is a lull in your investigation........WHAM! You bring more out - both barrels!
You inspire us all in Washington State! Keep it coming!
Posted by: Deborah on January 3, 2005 02:41 PMIs that voter fraud as well or is it OK in counties that voted for Rossi?
Posted by: Doug A on January 3, 2005 11:57 PMJust because I have a blog primarily devoted to fart jokes doesn't make me an idiot.
Posted by: Matt J Kurlander on January 4, 2005 09:04 AM