R-71: 9,000 new signatures counted ?>

R-71: 9,000 new signatures counted

Signature checkers for the state Elections Division have scrutinized almost 9,000 new Referendum 71 signatures, bringing the  cumulative check total to over 88,000.  

Over 77,000 signatures have been accepted and more than 10,000 rejected on various grounds, bringing the error rate so far to just under 12 percent.

sigcheckThe daily update showed these cumulative totals: 88,191 checked, with 77,637 accepted and 10,554 rejected.  Sponsors, Protect Marriage Washington, need 120,577 valid Washington voter signatures if they are to gain a November statewide ballot spot. They are attempting to overturn the newly adopted “everything  but marriage” law that expands state rights and responsibilities to state-registered domestic partners so that they equal those granted to married couples.

The rejection rate is 11.97 percent.  Sponsors can sustain an overall error rate of no more than 12.4 percent if they wish to make the ballot.  The error rate has been edging up, as traditionally occurs with initiatives and referenda as the duplicate signature rate rises. Thus far, 867 signatures have been rejected as duplicate or triplicate. At the same time, checkers are taking a new look at the signatures that were previously rejected because they weren’t found on the state voter database. Master checkers are reviewing to see if some of those people were newly registered and not on the version of the database the checkers were using.

The rejections thus far include 8,822 people whose registrations were not found, 821 whose petition signature did not match the one on file, 887 duplicates and 44 cases where checkers have asked the voter’s home county for an electronic signature that can be compared with the signature on the petition.

The checkers have now look at about 400 “volumes” or batches of 15 petition sheets, each bearing between 1 and 20 signatures.  There are 623 volumes total.

20 thoughts on “R-71: 9,000 new signatures counted

  1. Dear Secretary of State Sam Reed:
    Please allow all those citizens of this state who signed R-71 to have their signatures counted accurately. In short, SLOW DOWN and do things the right way not in haste.


  2. There were several comments posted earlier today (in response to the immediately prior SoS blog report) regarding two important issues: the lack of double-checking of every signature that was initially accepted (but no master/senior checker ever reviewed the initial acceptance to see if it was properly accepted), AND the problem of signatures from persons who were not registered to vote at the time they signed the petition. These are critical issues, and I would encourage readers who posted their comments earlier today on these issues to look at prior posts by me and Susan B regarding these issues — and the answers given by David Ammons to our posts/questions about these issues. In particular, and please excuse the lengthy comment below, but here is exactly what I posted some days ago — and, as wonderful as the SoS staff has been in answering questions, with all due respect I don’t believe their responses to the comments below were adequate to address the concerns that I expressed.


    (1) MY AUGUST 17 POST: Double-checking 200 ‘initially accepted’ signatures will hardly tell anything because the sample is tiny (00.14% — about 1/10th of one-percent).
    For example, the number of disqualified signatures due to duplication is now about 0.72% (of the approximately 70,000 counted so far). Therefore, if one were to check for the number of signatures disqualified for duplication in volumes 192 (297 signatures submitted), 177 (282 signatures submitted) and 152 (300 signatures submitted), one would expect to find 1 or 2 signatures disqualified for duplication among these approximately 900 signatures — yet there were ZERO signatures disqualified for duplication in these 3 volumes. However, one could not reasonably conclude that because there were ZERO duplicates in these 900 signatures, therefore there are no duplicates among all of the submitted signatures (about 138,000). The same analysis could be made in regard to signatures for which there are other potential grounds for disqualification — e.g., no match, or the voter isn’t registered, etc.
    In the above example of checking 900 signatures on 3 petitions and focusing only on the disqualification for ‘duplicates’, we are dealing with such small numbers — checking only 1/10 of one percent (900) of all 138,000 signatures, and an overall proven duplicate rate of about 1% (based on all 138,000 signatures) — that the result of checking only 900 signatures is meaningless.
    The SoS should be commended for double-checking some initially-accepted signatures — but the sampling has to be at least several thousand signatures, not 200. Otherwise, it is an exercise that really proves nothing.


    (1) MY AUGUST 14 POST: Dave (Ammons), I thought there was a deadline by which a certain number of valid signatures had to be submitted? If someone is not a registered voter on the date that the petitions are submitted, and if their registration is permitted to take place AFTER the date the signatures are submitted, then doesnt this effectively extend the deadline for submitting signatures? As I understand the answer you’ve provided from the Elections Division, if the deadline for petition submittal was August 1 (hypothetically), and there among the submitted signatures were 100 of persons who actually filed their registration to vote on August 3, then as long as those 100 people show up as registered by the date the checkers check these 100 signatures (say August 15, hypothetically), these 100 signatures will be counted???
    I hope I’ve misinterpreted what you’ve post from the Elections Division — otherwise, this makes a mockery out of the submittal deadline of August 1. It means: (A) I can get hundreds of unregistered people to sign the petitions by the submittal deadline, and then get them to register after August 1; and (B) I can register hundreds of others to sign even though I don’t know if they are registered and then, after I submit the petitions, I can check the group about whose registration I was uncertain and if they were not registered to vote when they signed I can get them registered — and as long as get the registrations of (A) and (B) completed before the checkers reach their names, they are valid! In other words, I just need to “win the race” with the checkers.
    If this is really true, so much for the deadline to submit signatures — it is a moving deadline, which seems inconsistent with state law. Does the law really allow this?

    (2) MY AUGUST 17 POST: Thanks (to Dave Ammons) for your response to my posting (and that of Bill W) on the August 13 blog regarding the acceptance of signatures from persons who were not registered to vote when they signed the petitions. You said:
    “Our policy says that if the new voter fills out the registration card, he/she has done all they can do, and that the subsequent process of the registration being received and processed at the county elections office should not count against the new voter. This has been the policy for some time, and was not dreamt up for R-71. This question, like all of the others associated with R-71 check, is subject to judicial review. I suspect no judge would rule against the new voters.”
    First of all, I’ve never said or even hinted that the process of qualifying signatures is being “dreamt up for R-71.” If others have said that, I disagree with their statements. I think all of you at the SoS office are doing your very best. However, what we are all finding out is that potential problems with long-standing procedures like this one (and others) really didn’t make a difference in the past when the signature counts weren’t close in prior referenda (or initiatives) — but they could make a big difference now.
    Second, as to the specific concern about counting signatures of persons who weren’t on your voter rolls (or whose registrations to vote weren’t even submitted BEFORE the petitions were submitted) — I don’t believe you’ve really answered the concern that this process effectively extends the deadline for submitting the correct number of signatures. The legislature set a specific deadline for petitions to be submitted; it didn’t seem to leave a loophole that the deadline could effectively be extended by days or weeks as the result of counting signatures of persons who weren’t registered voters when they signed, or when the petitions were submitted.
    Third, and related to this same point about the legislature having set these procedures: on Friday, the SoS lawyers made a filing with the federal court in which the SoS submitted affidavits describing the signature process, and the SoS quoted from state law. The quotes were to Wash. Rev. Code Section 29A.72.130 which says anyone who signs a petition is attesting that “I am a legal voter” (NOT, I “will become” a legal voter, or I “am filing a voting registration form”), and to Wash. Rev. Code Section 29A.72.130 which says that it is punishable by fine or imprisonment if someone signs “when he or she IS not a legal voter” (NOT, “he or she is not eligible to become” a legal voter). In fact, the petitions themselves that were submitted contain similar warnings to people not to sign if they are not legal voters. With the state law so clear, regardless of what long-standing procedures the SoS has followed, the fact remains that if Mary Jones or Tom Smith signs a petition on July 10 saying that “I AM a legal voter”, and if his/her new voting registration isn’t even sent into the SoS until July 15 — or, even worse, until after the petitions are filed — it simply is inconceivable that the person could be considered a ‘legal voter’ WHEN S/HE SIGNED THE PETITION, which is what the legislature has required. You’ve said that you suspect a judge would not find that there is a problem here, but can you cite one single court decision — from any state — where the court ruled that when a law says a person must be a legal voter at the time s/he signed a petition, the law allows that person’s signature to be counted even though s/he didn’t file his/her voter registration until days after signing the petition? I would find it interesting to read any such court decisions if you know of them — as they would appear to be contrary to the direct language of what Washington’s statutes seem to say.
    Again, thanks for all the hard work that is being done, and for your responses to our comments.

  3. Oh a second question. Are the “junior” checkers now utilizing the “new and improved” “up-to-date” database? Or are they still using the “old-clunker” database?

  4. Do you stop counting when you reach the error rate of 12.4% or do you complete the job? Who is paying for all this counting? Shouldn’t the people who submitted the signatures be forced to pay at least part of the bill? If this doesn’t make it on the ballot, can it be resubmitted or is there a closed window for this like in Maine? Sorry for all the questions, but coming from a state without referendums, I find this all a bit quirky and open to abuse.

  5. Thanks for these updates, I have learned so much about this referendum process through the blog postings.

    If a “not found” signature is double double checked and then found to be on the new voter database, before it is accepted, what is the process to ensure it does not have duplicates as well?

  6. myself & my husband Richard Perry are both registered voters of Washington state. We would ask that your staff spend counting the votes in R-71 in a mannor that will give time to count the votes correctly. This should be given equal time just as anything else & not pushed through in a hurry. Shirley Perry

  7. David,

    Is a signature counted as valid if the signer was not registered when they signed but did register later?

  8. In a show of bipartisan support, I agree with the sponsors’ calls for slowing down.

    Not master-checking the accepted signatures in the interest of a hasty conclusion to this process may prove to have altered the error rate by up to several percentage points. I understand your team is getting weary, but it is truly vital that we ensure that all signatures — including those already accepted — are master checked so we achieve the accuracy and rightness we are looking for in this process.

    I believe that the sponsors would not want to have a signature accepted that was not completely vetted, and I do believe they would wholeheartedly accept and approve of an vigorous master-check of accepted signatures. They wouldn’t want to reach the ballot with wrongly verified names, would they? No honest and upright campaign would ever want that. Any outcome is acceptable if done in an accurate way, and this includes double-checking that no signature was accepted in error.

    So please: Slow down, master check all signatures, and let’s get this one right. We need to truly discover the intent of Washington voters. Anything less would be an insult to this state’s great legacy.

  9. David:

    When you post today’s results, could you please indicate the extent to which the posted results reflect the shifts that will result from the “live database” check? I know yesterday’s results did not reflect any of these shifts, and if you could confirm this for today’s results, that would be appreciated.

  10. I agree with AJ above. The accepted signatures need to be subjected to the same kind of careful re-checking that the rejected signatures have been subjected. So please master check all signatures.

  11. To repeat my request: do the totals just posted a few minutes ago reflect any or all of the modifications that resulted from the live-database recheck?

  12. Also if the live-database modifications are now complete, it would be great to know what the total number of rejected-to-accepted shifts there were. Your post yesterday, indicated that it should be in the neighborhood of 12% of the no-match category, but it would be of interest to know what the final number was.

  13. Could you please let me know why the comment that I submitted several hours ago is still awaiting moderation?

  14. Hi Tim
    You and others have requested that we not rush our checkers.

    Our consistent direction to checkers has been to be as careful and methodical as possible in doing checks. We have never requested the checkers or master checkers to hurry up or speed up the process.

    That said, the production rates do change as a result of several dynamics. When we doubled the number of checkers from about 15 to about 30, production went up. When we initially hire and train new checkers, their production on days 1 and 2 are typically less than on days 4 and 5 when they have become more experienced and adept at searching the database. And, yes, it is possible, that later in the process as fatigue and the intense scrutiny of observers sets in, that production rates can change.

    But, the bottom line is that we have never encouraged the checkers to speed up or hurry though this check. We want the most careful and accurate check possible in this very close check.

    Thanks for your question.

    Nick Handy
    Director of Elections

    Hi Dave and Jim,
    You and others have asked about the registration data versus the petition check date.

    The issue is what happens when a signer signs a petition but does not get registered until later. The question is whether we will count the signature if the registration occurs after the date the signer signed the petition.

    The operative date for this check, and all previous checks in recent history, is the date of the check. We use the date of the check for the following reasons.

    We do not know what date a signer signs a petition. For the typical initiative petition that begins in January and is filed in June, we simply have no idea on what date any signer signed a petition. This information is not required by state law and does not appear on the petition.

    A further factor is that initiative and referendum gatherers typically carry voter registration forms with them. When a person declines to sign because he or she is not registered, the gatherer provides a voter registration form and encourages the person to both register to vote and sign the petition.

    To encourage and promote voter registration, the state and our office has actually encouraged and promoted this practice for many years. As a result, many people register to vote so their name will be counted on a petition. But, those signers were not registered on the date they signed the petitions, and may not become registered for several weeks later as the forms are later submitted to the state and processed by the counties.

    We do support the policy behind this decision. At the time of the check, we know that the person is a validly registered voter in the state of Washington. The only signatures that are counted for the petition are signatures of validly registered voters. And, signers were not rejected simply because their registration form had not been processed by the state yet.

    We do understand the legal issues that surround the state’s interpretation on this issue, but we are not going to change in the middle of this petition a longstanding practice and procedure that has not been changed by the courts or the legislature or our office over many years of initiative and referenda checks.

    Thanks for your question.

    Nick Handy
    Director of Elections

    Hi Bianca,
    You ask if we will stop at the point a sufficient number of names are approved or if we will complete the check of all names.

    We are planning to check all names submitted even if that means going beyond the minimum required. We are anticipating litigation at the end of this check because it is so close and we want the reviewing judge to have all information necessary to an informed decision.

    Thanks for your question.
    Nick Handy
    Director of Elections

    Hi Kevin,
    You ask what procedures are being used to ensure that signatures found in the “recent registration check” also known as the “third check” are not duplicates.

    This is an excellent question and I do not know exactly what procedures are being used for this but I can assure you that our supervisors managing the check are well aware of this situation and have instituted procedures to ensure that a duplicate check is being run on these.

    Your question is a good one because of course the recent registration check is being done on the live database and the first two checks are being done on a copy of the database printed in late June.

    Incidentally, we are now in a phase of the check where triplicates are beginning to show up.

    Thanks for your question.

    Nick Handy
    Director of Elections

    Hi Susan,
    You ask that we be sure to show the results of the “recent checks” in the data.

    We posted our initial data for the recent checks on Friday and blogged on the volumes which had been updated. I’ll check with our staff today to see how they plan to show this on the spread sheet that reports results by volume.

    Thanks for your question.

    Nick Handy
    Director of Elections

  15. Do you have any court decision, from any state, anywhere in the US, which says that it is legally permissible to count as valid signatures the signatures of persons who are NOT registered to vote at the time when they sign a petition — but they subsequently become registered voters sometime before their name is actually checked by the SoS??

    You say that this is the long-standing policy in WA, and maybe that’s true because it hasn’t made a different in past countings. But in the close case here on R71, it sounds like if the petition exceeds the number of required signatures by 1,000 — for example — and if the number of signers who were not registered when they actually signed the petition is 2,000, then you will count all 2,000 signatures as long as the signers’ names appear registered when you reach their names for checking. So instead of failing to qualify by 1,000 votes, the petition would actually qualify by 1,000 votes — even if those 2,000 people were NOT qualified voters when they signed.

    For sure, there will be a lawsuit over this interpretation of WA law. Not only as a practical matter does this illegally extend the date for filing petitions, but the law is also clear in saying that WHEN persons sign, they must be legal voters — not that they can sign and then become legal voters later. Again, can you cite a single court case that allows such a result in light of the specific requirement in the law that a signer be a legal voter AT THE TIME OF SIGNING?


  16. Hi J Scooter – We are not aware of any case law that says such signatures are invalid, and we don’t know of any case law that says such signatures are valid.

    To enforce a rule that a person must be registered to vote before the person signs a petition would require each person to confirm that he or she is registered to vote while standing in front of Safeco Field or in front of Wal-Mart or one of the many other locations where petitions are circulated. It also would require that there be some information provided as to when the person signed. No information is provided on the petitions as to when the signers signed.

  17. Brian, the problem is that even if someone registered to vote on August 1st (where they were previously not a registered voter) yet they signed the petition before the deadline date (July 25th,) their signature is counted. This indicates they were NOT a registered voter when signing the petition.

Comments are closed.

Comments are closed.