This just in --
Republicans have won a victory in their challenge to the election of Washington Governor Christine Gregoire.On the other handChelan County Superior Court Judge John Bridges said Republicans will be allowed to use "proportional analysis" to prove their case that illegal votes swayed the election.
But Bridges said he was not yet accepting the statistical analysis as valid for the trial set to begin May 23. He said the Republican evidence is subject to a separate hearing where Democrats can challenge its scientific value.One day at a time.
UPDATE: Bridges gave a small favor to the Democrats that won't hurt the Republicans on substance but will only cause them to spend more money on research:
In granting a Democratic motion, Bridges said that any party alleging illegal votes will have to produce in court a copy of the voter's signature in a polling place book or on the envelope of an absentee or provisional ballot.Oy. It appears that the good judge gets too much of his news from the Seattle Post-Intelligencer. Once Bridges reads Dean Logan's deposition, he will see that crediting is an integral part of canvassing, done prior to certification. Posted by Stefan Sharkansky at May 02, 2005 09:44 AM | Email ThisBridges said crediting is a "post-election administrative exercise" and "does not bear upon the authenticity of election results."
This is great news but one piece is missing and that is whether the proportional analyis will be broken down by county or by precinct.
I think the precinct level is very important as there is news that democrat party operatives have found unverified provisional ballots in or near Pullman. These could attributed to students who may have voted overwhelmingly for Gregoire.
Should the analysis be carried out at the county level, these provisionals could be unfairly weighted towards Rossi whereas if they are carried out at the precinct level, they could be weighted towards Gregoire.
In my opinion it should be common sense to think the precinct level is the more 'coarse' level that tends toward a more accurate analysis.
We'll see.
Posted by: Michael on May 2, 2005 09:55 AMGod love you all here at SoundPolitics.com
Posted by: Son of Liberty on May 2, 2005 10:01 AMPlease someone tell me if I understand this issue correctly.
First, the Democrat precinct poll watchers were given extra ballots. True or false.
Secondly, there was no acounting for those ballots. True or false.
Thirdly, there were hundreds more votes than voters. True or false.
What I understand is that there are considerably more votes than voters and the Democrats had extra ballots. Why is this not THE major issue?
Posted by: Ron on May 2, 2005 10:11 AMBridges said crediting is a "post-election administrative exercise" and "does not bear upon the authenticity of election results."
Is it Postman or Bridges who doesn't know what "crediting" a voter with voting in the election means?
If the statement is meant to say that voter lists published after certification aren't good enough to prove whether a particular felon voted, that's fine. Naturally, the list could contain an error regarding one specific individual -- so one must look at the poll book or absentee ballot envelope to see if that person cast a ballot.
If on the other hand it is meant to say that nothing done during canvassing to track the number of ballots cast -- that is, by crediting voters in the poll books and in the database when accepting as valid an absentee or provisional ballot -- then what is the evidence one would use to show that more ballots were sent through the vote tabulation than had been accepted?
Posted by: Micajah on May 2, 2005 10:27 AMThe logic in your post is too clean, clear, obvious and pure to allow it to be used as the point of judgement -- needs to be much more convoluted and distorted and obfiscated so that only the entrenced genious' of the ruling class can devine the "truth"
Posted by: Bill on May 2, 2005 10:39 AMOn a serious note:
I do believe the proportional analysis will be presented in two ways. Rossi's team will present one set of facts which includes by the precinct by precinct analysis and will be able to defend how they came to their conclusions very easily, just as a CPA can show where a company really stands by looking at the internals rather than the bottom line. The Dem's will present their indefensibile conclusions based on whatever helps them look best from county to county. It will be up to the judge to determine which one is the best evidence from which to draw a conclusion.
Additionally, no one should assume that the Rossi team will present a narrow case. It will most likely be an overwhelming case of illegal votes cast by non citizens/felons/dead voters, plus ballot stuffing, plus poor tracking of uncounted (finding legal ballots) votes, plus counting unverified provisional ballots, plus illegally certifying before being sure of results, and possibly much more. You want nuclear option--Dem's you got it.
Posted by: Larry T on May 2, 2005 11:01 AMNailed it on the head. The Rossi case needs to be so overwhelming that the cumulative effect is Hiroshima.
(uh oh...now the PC crowd is gonna jump on me for that insensitive analogy)
Posted by: Danny on May 2, 2005 11:07 AMHe know's the right's and wrongs of this case...He needs to apply a strict form of common sense now and smack the Democrats hands...There should be no more appeasement.
Enough is enough.
Posted by: Deborah on May 2, 2005 11:14 AMI doubt that Bridges is any more concerned about the result on appeal than any trial judge would be in any case.
Once a transcript of the judge's statements is finally released, we may see whether The Seattle Times got it right about the separate hearing stated in their article.
The judge has apparently refused to exclude the "proportional analysis" evidence proffered by the GOP. In a motion such as the one submitted by the Dems, they are, in effect, trying to suppress the evidence by getting a ruling ahead of trial that it is inadmissible.
Bridges says it is admissible, but that doesn't mean it will be enough to carry the day.
That's true of all admissible evidence. Until all the admissible evidence is in, and is considered, the judge cannot (and, of course, should not) say that it will be sufficient to prove the claims made by the party that offered the evidence.
If there is to be a separate hearing before the trial, I'm a little concerned. Such a hearing would mean that the admissibility isn't yet determined. The way The Seattle Times reported it, there would be a hearing to determine whether the proportional analysis is sufficiently reliable to be admissible. If the only question were whether it is sufficiently reliable to support the GOP's claims, then I believe that question would be decided at trial.
Posted by: Micajah on May 2, 2005 11:29 AMIt is dangerous for the Republicans because they can get mired in technical details and loose sight of the forest for the trees.
Dangerous for the judge for if he does not accept it the statistical analysis, he is left with
a) having no practical way to challenge any election.
or b) reversing himself on errors > margin of win is not sufficient.
It is least dangerous for the Dems, but they have to bring evidence to the table to challenge it.
I am not in favor of statistical, proportional analysis. It's critical assumption is that "errors" are random, not coordinated.
If proportional analysis is used to challenge elections, then a COORDINATED error to stuff the ballot box would work like this:
Party A wants their candidate A to win.
Put 10 extra ballots for A in a ballot box that leans 60/40 to Candidate B.
Loose 10 ballots for Candidate B from a ballot box from a precinct that leans 60/40 for Candidate A.
(Prior to the court challenge who would have thought the above was possible. Thanks to the KC reports Sharkansky dug out we know it happened several times over!)
Ok. If the stuffing is not found out, Candidate A benefits by a 20 vote margin
If the ballot counting IS found out, here's what happens.
10 extra votes in Candidate B's precinct. Proportional fix: Candidate B looses 6 votes, Candidate A looses 4 votes -from the canvas which included the 10 stuffed votes for A.
10 missing votes in Candidate A's precinct. Proportional fix: Candidate A get 6 of the missing votes, and cadidate B get 4 of them.
Net result from the two precincts: Candidate A is awarded a net 2 votes, Candidate B looses a net 2 votes. Margin of victory over the canvas is increase 4 votes in the favor of candidate A.
Hold the phone, there is still something up the sleeve! We started the proportional analysis with a canvas that INCLUDED the 10 stuffed ballots and does not include the 10 "Lost" ballots.
The stuffing "errors" increase the margin 20 votes in favor of candidate A. The porportional fix increased the margin to 24!.
If you use a proportional "fix" and IF a party wants to stuff the ballot box as described, then you WANT to be challenged. The statistical fix will be in your favor.
The statistical study is scientifically valid if and only if the errors are random.
The irony may be that the Dems can win by showing the judge that the errors are statistically NOT random. They know the errors are not random because they made the 'errors'.
Posted by: StephenR from Houston on May 2, 2005 11:33 AMYou're probably right about the risk of coordinated voting fraud as a means to take advantage of "proportional analysis," but we're stuck with the laws we have -- for now.
When our law talks about identifying which candidate "appears" to have benefited from illegal votes, then we have to talk about it.
Far better would be a law which unambiguously sets the standard for annulling an election without regard to how the votes "appear" to have been cast. It's fine to authorize the court to consider evidence of how they were actually cast, just in case such evidence is available; but it isn't all right to require production of such evidence when it isn't available (except via something like "proportional analysis" or asking people who cast illegitimate ballots to tell us how they voted).
When it's clear that far more ballots were sent through vote tabulation than the number of valid ballots issued and cast, then the important question ought to be: Could that number of illegitimate ballots have caused a person to be declared the winner even though that person didn't actually get the highest number of legitimate votes? If the answer is "yes, that's a big enough number to cause such a result," then the election ought to be annulled.
I believe RCW 29A.68.070 already states such a standard with regard to errors by election boards: "such as to procure" can be interpreted to mean "great enough to cause." But, that standard isn't stated with regard to "illegal votes."
It's better to have Lt.Gov. Owen assume the office of governor for a little less than a year while awaiting the next general election, for example, than to have someone put in office as governor who hasn't been duly elected.
Posted by: Micajah on May 2, 2005 11:51 AMYou are saying that the "illegal" votes would be proportional, but the canvas errors (ballotless voters, voterless ballots) may be to a different standard?
A very partisan canvassing board certified the election, even in the presence of a lot of information regarding significant problems.
The result is a statistical tie, but due to the archane nature of our laws, we simply stop examining the accounting at certification and then retabulate once by machine and then once by hand.
In the end, the thing that produced the variance in outcome between machine and hand recounts was not any difference in tabulation, but the introduction of more ballots that the partisan elections officials found to be counted for Gregoire. At that point, accounting and canvassing had already ceased.
There's no way of knowing what the actual results of canvassing and accounting were because the documents are irreconcileable and apparently they will not be used as grounds for contest.
Meanwhile, we've got an uninspiring and frankly very ugly union controlled puppet in office who has already gone back on the word of many of her campaign promises.
The state will suffer with the current statist vector until 2006 and 2008 and then I suspect that the WA electorate as a whole will speak loudly at the polls.
Kudos to Stefan for all of this hard work.
Hopefully Ms Gregoire will find the real killer someday.
Posted by: Jeff B. on May 2, 2005 12:19 PMBridges stated in court that there were two standards -- one for "illegal votes" and the other for errors, irregularities, misconduct, etc., on the part of anyone involved on an election board at the precinct level or otherwise in the canvassing and certification of the election.
I don't know whether today's hearing will modify or clarify what he said in February.
From the transcript of the Feb. 4, 2005, hearing on motions:
Page 28 – THE COURT: Counsel, if you’ll permit me, I’d like to elaborate with one more sentence on my decision with respect
Page 29 – to these causes, and I think I tried to be as plain as I could with respect to what I felt the statutory standard was relating to illegal votes. With respect to misconduct, whether that misconduct falls in 020 or 011, I think the standard is 29A.68.070, and so I want to say that so you folks have some sense of what I think the ultimate standard of proof is and what the petitioners have to show.
"The Rossi case needs to be so overwhelming..."
Remember the OJ trail, overwhelming evidence, but they thought they did not need it. Show everything, everything, drop all your bombs. Shoot down everything, hold back nothing.
Get ready for 'Lawyer Talk', "The meaning of is 'is'." It will get mighty slippery in there
Personally, I am just very happy that the case is being heard and that we will be able to get a resolution on this.
I don't think this is going to be as Red State, some conspiracy that could be compared to OJ.
Either we address this or we don't.
One thing is for certain is that all of us have spent entirely too much time and energy on this situation already.
Posted by: PRR on May 2, 2005 01:24 PMI just hope the GOP has a lot more ammunition.
Posted by: Tucker on May 2, 2005 02:02 PMI should know better than to accept the media's first draft (liberal friendly) version of what transpired today....
..I guess it's time to roll up the sleeves...Get all of the evidence categorized and ready....(But don't post it here...heh)
OK...Let's do this!
Posted by: Deborah on May 2, 2005 04:10 PMMisconduct of board -- Irregularity material to result.
No irregularity or improper conduct in the proceedings of any election board or any member of the board amounts to such malconduct as to annul or set aside any election unless the irregularity or improper conduct was such as to procure the person whose right to the office may be contested, to be declared duly elected although the person did not receive the highest number of legal votes.
Micajah, I'm not sure this section either can or cannot be interpreted as allowing proportional distribution of PB's fed through the Accutrons.
But it does seem to provide Rossi more wiggle room than would just the felon vote totals.
Posted by: Deadwood on May 2, 2005 05:33 PMBut I think trying to predict anything at this point is near impossible. What we know today is that the judge ruled in favor of the GOP contest, and then made it clear what the next hurdle is that they have to clear. In my opinion, he's helping the GOP out by ruling in their favor step-by-step and then giving them the time to make sure that they can adequately prepare for the next phase of the process. By the time we get to trial, the Republican case should be so completely solid, based on what Bridges has told the GOP that he wants to see, that the ruling will be easy for him to make.
By the way, on a semi-related note, the Mayor of San Diego will be stepping down after being ranked as one of the worst mayors in the country for his performance in that office. About his resignation, San Diego blogger Citizen Smash writes:
Once the resignation takes effect, Deputy Mayor Michael Zucchet, or another person to be designated by the City Council, will take over as Acting Mayor. A special election will then be called for November, to fill out the remaining three years of Murphy's second term.
Another high profile position in which an official is leaving office and will likely be replaced by a special election this coming November for the remainder of this term. It may help Rossi's case to be able to point to a situation that necessitates a special election so close to home.
Posted by: JRR on May 2, 2005 06:14 PMThat RCW section doesn't have anything to do with "proportional analysis," if it means what I think it means. It doesn't require the petitioners to identify the candidate who got the votes resulting from neglect, error, etc. It simply requires that the errors were "such as to procure" the declaration of the wrong person as the winner.
Trouble is, no one has talked that way for a hundred years. What the dickens does "such as to procure" mean? I think it means "great enough to cause," in other words, the number of illegitimate ballots counted as a result of errors by election officials is bigger than the margin of victory.
I haven't seen anything from Bridges that hints at the way he interprets that section. He has only said what I quoted from the transcript -- that he believes it is the standard of proof for things other than "illegal votes."
I don't know whether Bridges would accept the idea that illegitimate ballots inserted into the Accuvote machines aren't just "illegal votes" (if they are "illegal votes") that have to be parceled out to the candidates as they "appear" to have been cast. He might accept the idea that they are in the vote count through neglect, so don't have to be handled under that extremely restrictive law about "illegal votes." They were folded before being handed to voters, and thus recognizable by their distinctive creases, but Logan's gang never attempted to find and remove them from the ballot storage boxes. Their canvassing crew knew with reasonable certainty how many would be found in each box, so why not peek into the box and see if the creased ballots you think are in there are really in there? Having found just the right number of creased ballots in the box, wouldn't it be neglect not to remove them?
If people insist on knowing which candidate got the most votes on those creased ballots, look at them and count the votes on them after pulling them out of the boxes -- just for the sake of curiosity.
Logan's gang claims to have subsequently identified the probable registered voters for many of those illegitimate ballots, but that's not relevant to me. Those ballots didn't belong in the boxes. What didn't belong in the boxes ought to have been removed from the boxes -- and then the votes on the remaining ballots should have been counted for each affected precinct.
Posted by: Micajah on May 2, 2005 07:28 PMMakes you wonder WHY the Demcorats in this nation are so against efforts to ensure that only LEGAL voters are allowed to vote??? It gives me the chills! Makes me wonder how many illegal votes were cast across this nation last November with the assistance of the Motor Voter act and the efforts of sleazy organizations such a Moveon.Org??
Does anyone remember that website that was up after Bush won? The "I'm Sorry" site that showed weeping liberals from not just our country - but all over Europe and the world posting pictures of apology for the loss of their Presidential Candidate Kerry?
My first thought was that there must have been quite an international effort to sway our election results! Now - with what's been uncovered in King County ...I'm convinced!....
We live in disturbing times....
“Was such as to procure" means the misconduct did cause the person who did not get the most votes to win.
‘Was such as to procure’ means that their actions must indicate they were conducted to obtain a win. It would need to be shown that their actions or lack of actions were with intent to cause a win, not just that their actions caused the win. This is another flaw in our laws. They presume that those that conduct our elections are objective. They should be rewritten to account for human nature and assume bias.
Isn't she just an odd one?
Gregoire insisted that she was not involved with the contest - She said it was the Democrat *party* who was going to tangle with Rossi..She tried to make herself out to be nothing more than an interested observer to it all....Yet - there she was with John Kerry this weekend...raising money for legal expenses incurred fighting the Rossi contest!
Not worried? hah! By her mere presence in front of a news camera... she's very worried right now...And she should be.
Under 1, by special effort I would take procure as active work done to procure the election or procure an illegal vote that would alter the outcome of the election. In this case the special effort used to gain ballot signatures for only one candidate after an election deadline would be a special effort. Another would be a felon voter drive. Both required special effort.
Under 2, to bring about, This addresses procedure. In this case KCE failed to follow procedure which increased the effective result for gregoire. Failure to follow the procedure can cause something to come about as assuredly as following procedure. Imagine how the election would have been different had KCE followed procedure.
Under 3, guess what present KCE got for the statewide disenfranchised voters from the illegal, felon, and immigrant votes. You got it. A S@@ual favor of gutter slang lineage.
Also, procure can mean any of the applicable following terms.
come by
cater
buy
acquire
receive
accumulate
annex
appropriate
find
gain
get
glom onto
grab
induce
obtain
persuade
pick up
solicit
Funny, those terms usually appear in a derogatory sense in the reports investigating this election with primary inferrence on KCE results.
Posted by: Mark Beyer on May 3, 2005 12:51 AMLet me explain briefly. First off, with any statistical analysis, the result you get is precise with a given range. Two standard deviations provides you 95% accuracy. Three standard deviations, provides you 99% accuracy, and four standard deviations is 99.999% (I believe). Therefore, when it is stated that a precinct voted 60% to 40% for one candidate, what is missing from that number is the degree in accurracy of that number and the range. Therefore, is it 60% +/- 5% or what? Without all the proper statistical numbers, the numbers thrown out are meaningless.
My second point is that a valid analysis needs to demonstrate is that the population sample is representative of the population in general. This is where I have the most problem with the current argument with felons. Felons do not represent a normal cross section of the population. Therefore, unless you have accurate sample of how the felon population voted, I don't see how you can argue that the proportion is accurate. This is where the Republican's argument is the weakest. While I agree that one could argue proportional analysis for provisional ballots, I don't see the same numbers applying to the felon subgroup. The felon subgroup is predominantly male and middle to lower class. I would suppose if one could figure out how males of middle or lower class voted, you could apply this proportion, but since this data is not available, the analysis is very weak. I expect this is one area of attack the Democrats will use.
Maybe someone who knows more on the Republicans data can shed light on the two issues I raised regarding the statistical analysis (e.g., degree of accuracy of proportions to be used and how to relate that the precinct level voting is representative of how felons voted).
OBTW, since this is my first time voting. I would like to add one other comment. If what Rossi is claiming is true (i.e., that there were so many errors that we don't know who won), then why do many still claim that he won the first two vote counts? The errors being introduced affect the first two vote counts, just as much as the third vote count. Additionally, we know that the first vote count had an error with Gray's Harbor count and the first and second count did not take into account the King County ballots that never completed canvasing and were finally included in the third count. If these two errors didn't occur with the first two counts, Rossi wouldn't have won any of the counts. Therefore, what's with the claim that he won the first two counts? He contradicts himself with his lawsuit that the counts were so inaccurate that we don't know a winner.
Cheers!
Posted by: tc on May 3, 2005 06:47 PMThe issue of sample accuracy is minor. The standard deviation is easily determined from the numbers of total and illegal numbers, and it will be relatively small. Unless the bottom line result that the Republicans come up with is within 10 or 20 votes (that's a guesstimate on my part), the margin of error will be irrelevant. If the result is within such a small margin, then everyone can debate whether the Republicans need to show that Rossi should have won with 50% certainty, or 90% certainty, or 25% certainty, or whatever. But I doubt it will come to that.
The question of whether felon voting patterns can be accurately projected based on the whole precinct is the $64,000 question. I have sympathy for the judge here, because the answer is clearly no, and yet if he says that, then the Republicans are left with no reasonable way to win their protest. And the presumption must be that since the legislature specified how to protest, they intended it to be possible.
Arguably the judge could use the demographic breakdown (age, sex, income, religion, etc.) of felons and other voters, as the Democrats have asked, but there is one big problem with that. Demographic voting data is available only through polls conducted by nongovernmental organizations. The government has official data only on votes by precinct. If you think King County elections are haphazardly administered, just try scrutinizing the procedures of pollsters. I can't imagine a judge deciding an election based on polls. It *is* conceivable that a judge could use polls as a way of saying that one side or other has made a reasonable case that the election results should or shouldn't be thrown out, but even that gives a lot of power to the pollsters.
So the judge is left with two bad choices: Use official, relatively undisputed precinct data that is clearly inadequate, or accept a somewhat more meaningful analysis using more thorough but less reliable pollster data. How to weigh quality of data vs. quality of analysis? The law doesn't say. I'd lean toward more thorough analysis, but it's scary to lean down a slippery slope, especially when at the bottom of the slope is an ocean with Sharks...
And even if the judge accepts demographic analysis, the parties can debate forever exactly how to do the analysis using pollster data that is not only unofficial and inaccurate, but also far from exhaustive. I'm sure both parties are trying all different analyses to see how to make the numbers come out best, and the judge will need to weigh their arguments. Again, it would be tempting for the judge to say this is impossible, but the law clearly implies that it is somehow possible.
Maybe Rossi and Gregoire can just agree to share the office for four years?
Posted by: Bruce on May 3, 2005 09:15 PMI'm not sure whether the margin would be enough to change the result of the election, but it would be close.
Posted by: Bruce on May 3, 2005 11:53 PM