Washblog

Democratic Councilmembers had been well briefed by King County Elections (REALS) staff, asking softball and nonsensical questions while mostly avoiding any uncomfortable areas. Assuming these huge changes right before a presidential election are too risky, there are precisely three questions relevant to the two business cases before the council, none of which were asked:

1. Is King County's current ballot tabulation system legally and technically capable of handling our county's plans for all mail balloting?

   Yes.

   In fact, we have many options available to us.

   
   
   • Upgrading the GEMS tabulation database.
   
   
   • Presorting ballots by precincts (which should also be done to protect voter privacy).
   
   
   • Splitting the election database into two.
   
   
   • Any combination of the above.
   
   

   None of these options were addressed yesterday.
   



2. Will the new system count the ballots faster, leading to faster election results on election day?

   No.

   Because most mail ballots are received election day and the following two days and you cannot count ballots you have not yet received.

   AJ Culver, former chair of the Citizens Election Oversight Committee and master of the obvious, has previously said "This idea that ballot tabulation is a bottleneck is a red herring". Which is completely substantiated by Dean Logan's Core Planning Team Report - Moving King County to Vote by Mail, delivered to the council in February 2006. (I know, we being those "black helicopter types", citing the county's own data and conclusions is unfair.)

   The King County Daily Return Statistics table on page 5 shows historical and projected daily mail ballot returns. There's a pretty graph on page 6. Day 14 is election day. It takes 3 days after receipt for a ballot to be tabulated. Also subtract all the ballots dropped off on election day, which is 23% in Oregon. At most, only 1/3rd of all mail ballots be available to tabulate on election day. That's 300,000 ballots. The Executive's reckless plan assumes 600,000 ballots on election day. Complete fiction.

   More to AJ Culver's point, the General Election Production Rates and Current Processing Capacity tables on page 7 show that ballot tabulation, actually feeding ballots into a scanner, is the fastest mail ballot processing step and has the highest daily capacity. That's with our current system!

   In fact, exactly no part of the Executive's reckless plan will yield faster election results. The three slowest steps in mail ballot processing are

   
   
   • waiting for the ballots to show up
   
   
   • sorting the ballots (1 day turn around)
   
   
   • opening the ballot envelopes (2 hours per batch)
   
   

   Counting ballots before election day, faster ballot scanners, electronically modifying ballots, and automatic signature verification do precisely nothing to address these actual bottlenecks. The Executive's reckless plan does, however, completely undermine election integrity and fleeces taxpayers. Lucky us.
   



3. Is anyone ever going to talk about the second business case, for ballot accountability and tracking?

   Probably not.

   At the CEOC meeting earlier this month, current chair Ellen Hansen asked for a show of hands, to see who's read the second business case. No one. Only we two activists had read the plan. (Mike Snyder had left the room; he probably read the plan. And presumably Bill Huennekens, who was in the room, had read the plan he authored.)

   Part of this is our (election integrity activists) fault. We've studied it. It's horrible. At least as bad as the effort to buy new Diebold. But with much greater costs and an even more tenuous claim to actual benefits. Alas, we've allowed ourselves to be distracted by Diebold.

   That was probably the plan all along.
   

---

Blow By Blow

Now that I've pointed out the pink elephant sitting in the living room, I'll point out the funny bits. I've broken it down by person, instead of by topic, because that's how it's organized in my head.

Nick Wagner, Council Staff

I have nothing funny to say about Nick Wagner. Sorry.

Wagner is council staff. He attends the CEOC meetings. He delivered the staff report, which was the actual agenda item. His report was okay. He explained our current system. He summarized the Executive's reckless plan, their various rationalizations (database limitations, speed of election results), and the alleged benefits (reduced manual ballot handling).

Wagner listed some issues staff identified.

The top being lack of certification. This caused some back and forth. Everyone seemed satisfied that federal certification would be enough. No one noted that at this late stage, our state would have to provisionally certify any new system. Meaning a wave of Sam Reed's magical certification pen. (Which is probably moot, because Paul Miller already told the CEOC that he wouldn't retest anything the feds had tested, as required.)

Next was Diebold's well deserved reputation as a bad actor and their horrible history on security. Diebold claims the stuff we're currently using is totally secure, completely addressing all the problems previously identified.

Wagner pointed out that the State of California is currently doing a complete review of our system (which is also used in their state). Their report is due next month. Wouldn't it be nifty to wait for their report? Then we could see if Diebold has actually fixed the security problems and changed their errant ways.

I thought it was a very reasonable suggestion.

Unfortunately, everyone got stuck on the security issue, deliberately missed the point. Mark Radke felt compelled to mention California isn't looking at the new system King County intends to buy (the Assure 1.2 product suite, including DRS scanners), inadvertently making my point that NO ONE has inspected or tested this new system.

Lastly, Wagner pointed out that the Executive's reckless plan was concocted by the Executive's staff, King County Elections, to justify the purchase of a new system and not the result of a rational, formal, standard practice request for proposal process. For instance, the Executive's criteria ruled out any system which did not count ballots before election day, under the mistaken assumption that such democracy endangering lunacy is required to achieve the mythical "early results".

Wagner suggested that given the rapid rate of development in elections products and technology, opening up the selection process with a formal RFP would likely generate solutions that King County Elections hadn't considered. Wagner is correct.

Julia Patterson

Patterson pointed out that closing the poll sites and switching to forced mail voting increases the number of mail ballots. Well, yea. So if you want earlier results, keep the poll sites open.

To her credit, Patterson did touch on two relevant issues.

First, she asked Sherril Huff if this reckless plan, on top of all the other changes, wasn't too risky.

She also pointed out elections officials are under tremendous pressure from campaigns, politicians, and the media for early results.

Notably absent from that list are us voters. We're not asking for early results. The polls show we want accurate results. And yet we voters are making the sacrifices, in both election integrity and treasure, to satisfy other people's obsessive self-interest. Same old, same old.

Larry Gossett

Gossert helpfully pointed out that the problems with the governor's race in 2004 wasn't with Diebold. But he stopped short of pointing out the obvious: The problems in 2004 were completely with mail ballot processing. So what should we do now? After 3 years of getting those problems fixed, the 13 successful elections Huff continues to point out, the answer is clear. We should throw away our current system and start from scratch.

Gossett asked if counting ballots before election day was okay. A question likely planted by KCE. Wagner answered no, but KCE says the new Diebold solution is okay. Because it only scans, not counts. Which is a complete fiction. The elections officials want us to believe that tabulation doesn't happen until the final report is run.

Wrong.

If a ballot's votes can be modified in a database, then they're counted. Diebold's Central Tally Systems (which sits between the DRS hardware and the GEMS database) are networked and every system has a copy of the entire election (for redundancy, in case a system breaks). So the risk of "cheat peeks" at early results remains.

And if you're using one of these anti-democratic products, you better be running daily reports. How else would you know if someone was tampering with the results?

Gossett then started to challenge the questioning of the Executive's selection criteria, stating that it would have been better if that feedback had been done earlier. Well, yea. But the council wasn't briefed earlier, so no feedback was possible. Gossett was quickly herded back onto the farm and that topic awkwardly dropped.

Dow Constantine

Constantine gets the award for non sequitur of the day. He delivered a goofy quotable about "criminal masterminds" hacking our elections. Though it had a good outcome. Mark Radke stated the Diebold systems were utterly secure and couldn't be hacked. A first in the history of computers. Nice.

Constantine did sponsor legislation to protect our secret ballots. Which we support.

Larry Phillips

Phillips made a great show of "pushing hard" against Diebold about security. He was more or less channeling the concerns of election integrity activists from the year 2005. I sincerely appreciate the effort. But we "black helicopter types" have moved on.

All of the computerized voting and counting systems are insecure. This isn't in dispute. King County has thus far successfully mitigated the problems with physical security procedures.

What we're talking about today, June 26th 2007, here, in King County WA, is completely cratering our elections by changing everything just in time for the presidential election. What's so hard to understand about that?

Phillips has been the best about following up on our concerns, even repeating our request to DRS for confirmation that their hardware can handle mail ballots. Which has been helpful.

Bob Ferguson

Ferguson gets the reward for most relevant question of the day. He asked Diebold, given their history of screwing our county, why we should reward them with more business. Ferguson, of course, was a lot more polite than I would have been. But his point is spot on. By being the first customer for this new system, we basically make their product successful. Diebold Election Services (DESI) is under pressure from their corporate overlords to stop tarnishing the company's good reputation in banking/finance and physical security systems.

So, if you look at it objectively, Diebold should PAY US to be their first customer, just to say THANK YOU for all our troubles and efforts. It'd only be fair.

Robert Chin, Diebold's director of screwing King County, replied that he was shocked, shocked at the low customer service rating Diebold got in the Executive's reckless plan.

Ferguson also brought up former Diebold executive O'Dell's notorious commitment to ensure that Ohio goes for Bush. Yea. Good meat for the base. Whatever. But that's not what we're talking about today.

Laird Hall, King County Elections

After enthusing about his work adding cryptographic digital signatures, Mission Impossible-style biometric authentication systems, and NSA-grade encryption to our elections systems, Laird Hall admitted that splitting the election database in two was beyond his technical ability. Apparently adding two values from two different spreadsheets was just too gosh darned tough.

Hall also reassured the council that only he and his buddies have the passwords to access the secured, encrypted, and quantum displaced election systems. So we shouldn't worry our pretty little heads about security. It's all taken care of.

Sherril Huff, King County Elections

I was most impressed with Huff's testimony. She gets the award for most selfless act of the day. When questioned by Patterson about the risks, Huff jumped on the grenade and took complete and full responsibility for the successful outcome of the Executive's reckless plan. Not a problem.

This is a huge change. In response to previous questioning at CEOC meetings, Huff stated that she was following orders, the risks were well known, but her department would do what they could.

Huff was also asked by Phillips why we can't split the election database. Huff said they wouldn't have enough room for all those scanners. So it's a lucky thing that we're getting a brand new huge warehouse just for elections!

Also, what's apparently too obvious to notice, splitting the election database doubles the maximum number of optical scanners from 40 to 80. Hello! That doubles the capacity for tabulation. Not that tabulation is the bottleneck...

Bill Huennekens, King County Elections

The good news is that some of what we're saying is getting through to acting superintendent of elections Huennekens. Now that he's acknowledged the DRS hardware he's buying has a "carriage printer" option for printing endorsement numbers on ballots, required for conducting manual recounts, it was safe for Constantine to ask the hard hitting question.

Although my reading of the statutes doesn't say that a "report" of ballots changed is legal. Imagine the procedures for manually counting ballots and constantly referring to these extra reports to see if the ballot before you has been changed during canvassing. So the next step is for Huennekens to figure out how to print out "electronically adjudicated" ballots. With the correct artwork. On original ballot stock. Should be interesting.

I figure if we can't stop the reckless plan, that we're at least somewhat responsible to make sure KCE doesn't blow off both feet.

As always, we remain eager for Huennekens answers to our other open questions:

1. What is the make and model of the electronic poll books to be used at the regional voting centers?

2. What are the procedures for fishing the ballot envelopes with challenged signatures out of the cage?

3. Can the DRS hardware handle King County's mail ballots? Why wasn't this tested during Diebold's demo to King County?

Mark Radke, Diebold

The importance of this transaction was made clear by Radke's presence. Radke is prominently featured in the HBO's documentary "Hacking Democracy". He told the council the same nauseating tall tales about security as is shown in the movie.

One wrinkle I find fascinating is the enthusiasm the voting equipment vendors have for mail balloting. First, they fleece us for $6b of federal and untold local money selling us unreliable, insecure computerized voting crap to comply with their HAVA legislation. Gee, oops, it looks like computerized voting is a bad idea. Our bad. How can we make it up to you? Hey! We've got just the solution! Mail balloting! Oh, and by the way, you'll need to replace all the crap you just bought with all new crap.

How convenient.

Kathy Rogers, Diebold

Diebold owns and operates the elections of the State of Georgia. Kathy Rogers, their former Director of Elections, made it happen. Rogers is the exemplar of the conflict of interest between vendors and elections officials. Diebold's enforcer was present to ensure King County knows what's what.

Robert Chin, Diebold

I have no idea who this guy is. Apparently, he's been haunting King County Elections since 1998. Who knew? Should I ever see him, I'll ask about Diebold misprinting our ballots last year, which screwed up our tabulators, slowed down election results, and stuck us with the bill for manually duplicating 15% (~62,000) of our mail ballots in Nov 2006.

Chin gets the award for most patently stupid statement of the day. He claimed that mail balloting is more secure than poll site balloting, because poll sites have all those pesky poll workers watching things with all their checks and balances, whereas mail balloting has centralized control entrusted to qualified elections officials.

Thereby demonstrating that Chin has precisely no idea what "security" means or how elections work. With centralized counting, just a few people can do evil. With poll sites and paper ballots, you need 100s in the conspiracy.

Republicans, King County

I'm new to this political thing, so I'm still easily amused. The moment the reckless plan came up, the Republicans disappeared. Making it clear the Democrats own this bad idea. That's brilliant. They're following the sage advice: When your enemy is digging himself into a hole, it'd be rude to interrupt.

So congratulations to the Democratic leadership. Another own goal. Bravo.

[Updated 2:42pm: Typos, clarity, and to tone down some of the sarcasm (per Noemie's comment below)]