R-71 update: Elections Division prepares new-look reporting ?>

R-71 update: Elections Division prepares new-look reporting

r71updateState election officials are revising and upgrading the reporting process for Referendum 71, hoping to provide a clearer picture of how the signature-checking process is unfolding. The first report with the new approach will debut Tuesday afternoon.

The plan is for a daily late-afternoon update that shows signature-checking results for each batch of petitions that have been completely checked, and then re-checked by a supervisor.  Each batch has 15 petition sheets, with varying numbers of signatures.  

The fluid nature of the signature-checking has made it difficult to make sense of daily “snapshots” and the daily rejection rates that were published previously.  As of Friday night, for instance, hundreds of initially rejected signatures have been reversed by the master checker and added to the accepted stack, often because the signer wasn’t initially found on the statewide voter data base, but was found upon closer check. 

The new approach will provide information on signatures checked, signatures accepted, and signatures rejected or still pending an electronic signature from the person’s home county. An updated website,  also will give totals, an error rate, and reasons for rejecting signatures. Rather than reflecting signatures that are in mid-point of being processed, the new report will mostly be an end-of-process view, and not the raw data along the way.

“We hope viewers will find this clear and accountable and provide greater confidence as we update our website,” says state Elections Director Nick Handy.  “We have a time-tested system we have used for years to check our initiatives and referenda, including master checkers and permitting observers from both sides to watch everything we do. We didn’t alter our process for R-71. We are using great care.”

The Elections Division hopes to complete the signature-verification process before the end of the month. Sponsors seek a statewide vote on Senate Bill 5688, the “everything but marriage” law that expands rights for couples registered with the Secretary of State’s domestic partnership registry. The law is on hold pending the referendum outcome. Sponsors need 120,577 valid signatures to get on the ballot. They submitted over 137,000 signatures on July 25.

21 thoughts on “R-71 update: Elections Division prepares new-look reporting

  1. Respectfully, calling a system “time-tested,” or asserting that something is being done because it’s the way it’s always been done, does not establish validity. If you always speed when you’re driving, the officer will still give you a ticket even if you tell him that you have always speed on this road, or even if you say your speeding is a time-tested system because you’ve never been caught. Perhaps this new public visibility to the system will reveal that only double checking rejected signatures and not accepted signatures is, perhaps, not the right approach and that it does not ensure a voter’s accesibility to the iniative and referendum process, but rather skews the process to favor one side of an issue.

  2. I’m not pleased about not updating the site until tomorrow afternoon, but I hope the results will be clear and worth the wait. :)

  3. I agree with Brian Spencer. Only re-checking the rejected signatures is not acceptable. There may be case law, but I hope this gets challenged. There are statistical reasons to double check both, and as Mr. Spencer says, the process now in place favors one side. This is no different than recounting all votes. I wonder what the Governor would say….

  4. If you have several databases of signatures, say a state database and county databases, and the first (faster) checker only looks at the state database, you should expect some signatures that are only in a county database to get falsely rejected. The (slower) master checker, consulting more databases, can correct the error. On the other hand, it makes no sense to recheck a signature that was already unambiguously found, and accepted, in the state database.

    This is not like a hanging-chad situation, where the presence of the signature is ambiguous and debatable, no matter who looks at it. For this system to work properly, though, it is important for the initial checker to only accept signatures that are unambiguously valid, based on whatever databases she is looking at. If the initial checkers are not doing that, then the system is truly one-sided and unfair.

  5. I agree with this on principle, but I bet the courts would not see this as an issue of two sides, but rather more one-sided as that of the people’s constitutional right to get something on the ballot. Either way, I would like to see this challenged to have that question answered.

  6. @Duane Williams:

    I believe you are making two errors.

    First, you assume that the junior checkers review one database, whereas the master checker reviews two. I don’t understand that to be the case, based on what I have read on this blog. I think the master checker reviews the same database, but supposedly is more careful due to his/her greater experience.

    Second, you assume that the initially accepted signatures are unambiguously valid. I see no reason to accept that assertion, since the same kind of human errors that can cause a junior checker to reject a valid name (e.g., looking at a wrong address, missing a middle initial) can cause the junior checker to erroneously match and accept a bad signature.

  7. Duane, from what I understand the signatures are accepted or rejected on the same system at the first level of checking. Therefore it is one sided to recheck only the rejected ones. The reason given yesterday is that the state wants to fault towards access to the ballot. “Faulting” to one side is bias. This should be challenged.

  8. Thanks to blog readers who continue to comment on our master-checking process. Re Duane Williams’ post, our checkers and master checkers are using only the statewide voter registration database (VRDB) when they check the signatures. The only time the county elections department is involved is when the voter’s electronic signature hasn’t been uploaded to the VRDB or when the signature isn’t legible. Typically the counties find many of the missing signatures and submit them electronically.

  9. I certainly want to see every legitimate signature counted and signatures are the only system we have to verify that someone really signed the petition.
    However, as mentioned before, sometimes people will sign for spouses, family members and friends thinking that “if they were here they would sign this” but doing this is illegal and the forged signature should not be accepted. This is the kind of thing a senior-checker might be checking for and hopefully the initial checkers are flagging EVERY questionable signature and not just accepting them because the address and other stuff matches but the petition signature happens to look different than the one on file.

  10. Thanks so much for being part of the conversation about the referendum process.

    Just a friendly reminder that we will not post any comments that contain vulgar, offensive, threatening or harassing language or personal attacks.

    State law does prevents us from posting comments that specifically endorse of oppose any measure, and content promoting or advertising commercial services, entities, or products.

    We are happy to hear your comments and questions, and look forward to answering them in upcoming posts and updates.

    Please refer to our agency’s blog use policy here: http://blogs.secstate.wa.gov/FromOurCorner/index.php/blog-use-policy

  11. I would like to expand upon some points that I made yesterday in a posting, now that we have a more detailed explanation of how the initial signature checks are done.

    First, it is now clear that even if there is “long-established case law” that the state is supposed to “conduct the process to facilitate access to the initiative and referendum process guaranteed in the Constitution”, there has never been any court or other ruling on the issue of double-checking ONLY rejected signatures and giving them a second-chance to qualify — and not double-checking accepted signatures to see if they were wrong accepted. So, in fact there may be unique circumstances here that support the claim that the “double-standard” of not double-checking all signatures is inconsistent with state law or the constitution.

    Second, there does not seem to be any good response to Brian Murphy’s concern that unlike prior signature drives where the signature list was public, in this situation it is impossible for any voter to know if his/her name was signed without authorization because of the federal court TRO against disclosing the signatures; for example, one cannot rule out that, WITHOUT AUTHORIZATION OR KNOWLEDGE, a parent signed his/her adult child’s name, or vice-versa, or one spouse signed another’s name (see Bill W’s comment above). (NOTE: As noted in my earlier posting, if nothing else, this is a reason for the State and other interested parties to ask the federal judge to dissolve the TRO — the so-called privacy / retaliation concern feared by some signatories must be weighed against the right of all voters to assure that their names were not improperly submitted, especially now that we know that the State double-checks ONLY rejected signatures and does not double-check any accepted signatures. One wonders if the federal judge who granted the TRO might not balance the equities differently, and dissolve the TRO, once he knows how the signature checking process is being conducted; and one wonders whether this is not a rare case where the TRO can be appealed and where the 9th Circuit would itself dissolve the TRO because the non-disclosure of names affects the signature-counting process and the harm created by the erroneous acceptance of unauthorized or forged (but authorized) names might be difficult or impossible to undo if that tilts the balance to qualify the referendum for a vote).

    Third, no one from the Secretary of State’s office has responded to my earlier suggestion that — at a minimum –the State should randomly double-check 20,000 accepted signatures and see what it finds. If no errors at all are demonstrated, then perhaps double-checking of all accepted signatures would be a waste of time. But if even 1% error is found in the accepted signatures, that itself demonstrates the need to double-check ALL acceptable signatures.

    Fourth, Brian Spencer and David Johnson have noted — as I have done in my earlier posting — that there are TWO sides to the policy / constitutional question of qualifying for a referendum: on ONE side the right to get on the ballot, and on the OTHER side the right not to have a validly enacted law improperly submitted to a referendum. No one from the Secretary of State’s office has yet explained WHY, once a law is validly enacted, those seeking to retain the law should not be given the same considerations during the signature checking process as those seeking to overturn the law. In fact, one could say that once a law is validly enacted, it denies equal protection to give the law’s opponents two bites at the apple to get their signatures qualified VERSUS not giving the law’s supporters the same two bites to assure that accepted signatures were properly accepted. In fact, and most troubling in this situation, the law’s supporters really have not even received one full bite of the apple because the TRO denies them the right to see the very signatures that the law’s opponents have seen and submitted!

    In short, unless the State conducts — at a minimum — a random double-checking of the accepted signatures to determine if there are ANY errors in the accepted signatures — there seems to be a strong statutory or constitutional flaw in the qualifications process that is being used, whether that flaw exists due to Bush v Gore or some other court decisions. It would be helpful to have someone from the Secretary of State’s office address the specific points above, especially why there isn’t a random double-check of accepted signatures AND why the State does not ask the federal judge to dissolve the TRO so supporters of the law can get at least one bite of the apple in judging the validity of the submitted signatures.

    Thank you.

  12. Washington should require a significantly larger amount of signatures for a referrendum that proposes to alter civil rights. 120,577 is too low a standard.

  13. J Scooter:

    Well said. I hope Mr. Ammons will consider passing these points along to Secretary Reed and/or to the SoS legal department.

    There really is no reason (other than adherence to tradition and a well-intentioned but misplaced desire to favor qualification of referenda) why this needs to be an issue.

    If the SoS master-checked all signatures, it would eliminate this issue entirely. The anti-R71 folks would be satisfied and the pro-R71 people could not complain, since all of the signatures then would be treated equally.

  14. David,

    Can you please clarify whether there is any specific order by which the bundles are being reviewed? Is it by the order in which they were collected, or is it random?

  15. Timothy– no, they were bound in volumes of 15 petitions, with no particular order, either geographically or date in which they were gathered initially. Each bound volume was assigned a number and are being assigned to checkers in no particular order– luck of the draw.

    J Scooter and David Johnson — essentially requiring a double check would double the time and taxpayer expense for the every-signature check. Usually, an initiative or referendum sponsor turns enough a large enough pad to permit random sampling and it never comes to every-signature checking. We have no intention of moving to a full double check, absent a change in policy imposed by the courts or adopted by the Legislature or voters by initiative.

    Shawn– the signature thresholds are set by statute — for initiatives, it’s 8 percent of the last vote for governor and for referendum, it’s 4 percent for referring measures the Legislature has adopted. (The Legislature may refer its own questions or legislation to the ballot on its own action, without signature requirement or governor’s signature.) Washington is considered to have a fairly easy access-to-ballot requirement — although most sponsors have found it’s a daunting job to get on the ballot, usually requiring a very expensive campaign that includes $300k or more to hire paid signature-gatherers, or else a very, very well organized grassroots campaign with a killer issue.

    In Washington, of course, we cannot amend the constitution by initiative. that must originate with a two-thirds vote in both houses, plus simple majority of the voters.

    Bill W–once the signer is found on the voter database, the signature match is the heart of the checking process. checkers are trained to spot phony signatures submitted by someone other than the actual voter.

  16. Mr. Ammons,

    While I appreciate the fact a double check would “double the time and taxpayer expense for the every-signature check,” this particular referendum would, if successful, essentially deny a specific group of persons a set of rights bestowed upon another group of people. I would hope the Secretary of State’s office would behoove themselves to consider all Washingtonians human beings worthy of consideration — regardless of the hours of counting or taxpayer expense — to at least move to a double check for accepted and rejected signatures.

    Are the gay and lesbian citizens of Washington state not worthy of consideration, at least until R71 does, or does not pass in November, if it goes that far? Perhaps if R71 is successful, such a motivation to deny double checking for the accepted signaures would be understandable. Until then, …

  17. “essentially requiring a double check would double the time and taxpayer expense for the every-signature check.”

    Sorry to be blunt, but so what? The process you are undertaking will determine whether millions, possibly tens of millions of dollars will need to be spent by the pro- and anti-R71 groups, private citizens, and the state of Washington. If that requires a couple of extra weeks to do the job right, why not do it?

    Also, if your office hadn’t started down the path of using master-checkers – thus “doubling the time and expense” – for one group of signatures, this wouldn’t even be an issue. But since you have done so, and since that practice is affecting the count, and given that it appears that validation may turn on a few thousand signatures, this is a big issue.

    Maybe this is the first time in the history of WA referenda that the count is this close and the arcana of the review process actually impact the outcome, but there’s a first time for everything. Anyway, those are my thoughts. I understand your position, but I predict you’ll get that court order.

  18. David, To say that you have no intention of double checking all signatures due to taxpayer expense when our lives and civil rights are on the line is insulting and minimizes the importance of this issue. Let me ask this – if the referendum sponsors reached exactly 120,577 under the current system would that trigger a rechecking of the accepted signatures? This is the equivalent to recounting only Governor Gregoire’s votes in her first election.

  19. to the last few commenters — my point about time and expense was simpling stating a practical reason to not double the entire process. the fact that campaigns on both sides need to know whether there is a November campaign or not is another. the fact that millions or ballots and Voters Pamphlets have to be printed with or without R-71 included is another. But the main reason, which i’ve stated over and over is that the Elections Division has a policy, grounded in the Constitution’s requirement to facilitate access to the I&R process and the ballotbox, that says we affirm a signature with one checker if it all lines up, the person is in the voter database and the signature on the petition and the voter registration on file is a clear match. If we’re going to deny a signature, then we want to make sure we used due diligence before rejecting that person’s voice. This policy has been used for as long as anyone can remember, and was not invented for the R-71 check. There is no intent by anyone to “make sure R-71 gets on the ballot” or, conversely, that signatures are rejected without good cause so that R-71 is denied a ballot spot. I know it must be hard watching from the outside, but that’s my perspective — no bias. Just using the best practices we have developed over many, many years.

    Re John Sutherland’s question. If indeed we accept 120,577 signatures as valid, we will certify it for the ballot. This is different than with candidate races, where there is an automatic recount if the race ends up with one half of a percentage point. As with election challenges, the I&R process is subject to judicial review.

  20. David, First I want to say that I am not questioning your and anyone else’s integrity in this process. But I am questioning the process – I don’t care how long it’s been in place. Technology has also changed. In the past the database info wasn’t as readily available so I’m sure it even took more time. I suspect that this was part of the original process consideration. This process should have changed when the technology did.

    Statistically, your hypothesis assumes that 100% of the rejected signatures are invalid. Conversely, the hypothesis is that 100% of the accepted signatures are valid. This is statistically impossible. It’s simple to ensure fairness – double check both sets of signatures. There is nothing else that will, and those of us who are frustrated by this will probably never be satisfied with anything less. The fact that there are observers in place is helpful, but on some level irrelevant. They are only there to ensure that the system currently in place is administered properly – they are not checking signatures. They are not there to challenge the system.

    You can imagine why some are freaking out over this. The process wasn’t as transparent to them before prior to this issue. And by posting the dailies, then removing them, it seems like there is something to hide, though I know from your perspective it is quite the opposite. My suggestion would be:

    1) Post the dailies pre-master checker in one column.
    2) The 2nd column would include only those signatures that have been finalized.

    Why not? Then we could at least watch the process unfold. The flaw wasn’t that you were posting the dailies, but the fact that the process wasn’td fully illustrated in the spreadsheet. Do this, and it will relieve some of the frustration. Please.

  21. John Sutherland– Paul Miller of the Elections Division got back to me on your question:

    “Thank you for your constructive critique of our reporting process. This is a new process to us – in the past no reports have been issued until everything was done – and we appreciate your feedback.

    “You might recall that the initial reports were daily summary totals from the database of the status of the work for that day, much of which was in mid-process. Now we are reporting the totals for each individual volume after the process of evaluating that volume is complete.

    “These reports come from two different systems and processes. The database the original reports came from is not designed to track pre-master check and post-master check statuses, so there is no simple way to comply with your request without introducing intensive manual record keeping into the process.

    “Providing data that makes this process transparent and understandable is important to us. We will take your suggestions under consideration for future development.”

Comments are closed.

Comments are closed.