Failure echoes 2008, as two different machines and at least six different failures stymie both poll workers and county officials at my own home precinct...
By Brad Friedman on 6/8/2010, 11:50pm PT  

-- Brad Friedman, The BRAD BLOG

Two years ago, in June of '08, the ES&S "InkaVote Plus" e-vote system in Los Angeles County misprinted 4 out of 12 of my own votes.

Today, as I tried to vote on the same system, the failure was even worse. Incredibly. And not just because I cover issues of Election Integrity for a "living."

I spent more than two and a half hours not casting a vote on the system before eventually I, the poll workers, and, apparently, the folks at the L.A. County Registrar's central help desk call center, simply gave up. A complete and total failure of the e-voting system for disabled voters in the nation's largest voting jurisdiction. Again. On a system the county spent millions to buy in order to comply with the federal Help America Vote Act (HAVA) boondoggle by allowing disabled voters to cast their votes independently. Complete and total failure.

I did my best to document the disaster as it happened, in real time, on my Twitter feed. But the cell signal was very bad at the polling place, so not everything got out. But here's what happened in detail, and with a few graphics tossed in...

As I did during the state primary in June of 2008, I requested to vote today on the ES&S InkaVote Plus audio balloting system. There is one for each of the two "precincts" in my polling place. In '08, the machine at my "precinct" wasn't working and the poll workers were afraid to touch it, due to exposed wires, so I was forced to vote provisionally on the machine at the other "precinct" inside the same polling place.

This time, the machine at my "precinct" was up and working, though I was the first to (try to) vote on it today, according to the poll workers.

Failure #1: My audio ballot started playing in some kind of Asian-sounding language. Instructions from the precinct chief Hal to push the arrows on the device until I was allowed to select English as my language didn't work. The audio simply kept repeating the same thing to me in Mandarin or Cantonese or Vietnamese, or whatever it was.

After Hal called the County Registrar's office to find out what to do, the system was reset, and the audio instructions finally began for me in English. All was well, it seems...before I noticed that the candidates listed by the audio in each ra1ce did not match the order that I had them on my sample ballot. I began to expect that once I printed out my ballot at the end of the process, I'd discover that it printed my choices onto differently numbered bubbles on the ballot than were supposed to be used in my precinct. By the time I got to the race for the U.S. House of Representatives, my fears were confirmed.

Failure #2: The audio ballot offered me the candidates for the 31st Congressional District instead of my own!

After Hal called the County Registrar's office back again, the system was reset, again, and I started my ballot over. Again, I found that it was offering me candidates in a different order than they appeared on my sample ballot, and then I was again offered the candidates in the CD-31 race. Wrong again.

More calls to the county resulted in my being given two choices by the poll worker at my precinct: Vote manually on a paper ballot, or wait for a poll technician to arrive with a new machine. There was no estimate available as to how long I might have to wait for a new machine to arrive.

I had already been there some 45 minutes by then, but was concerned whether disabled voters in the entire county would also be forced to vote in the CD-31 race instead of their own district --- or, more accurately, forced to vote for candidates which would be incorrectly inked into bubbles, for some other candidate, on the ballots for their own precinct. All of their votes would be recorded incorrectly. Since the majority of voters who use these assistive devices are vision-impaired, they would not likely be able to check which bubbles the computer had inked in on their paper ballot as it was printed (and then put into an op-scanner). They would, unknowingly, vote for different candidates than they actually wanted to --- in every single race.

I came up with a third option: I asked if I could vote provisionally again, on the machine at the other "precinct" in the same polling place. I'd hoped to see if the same problem was on every machine, or I had just gotten lucky on the machine in my own home "precinct."

The same fine gentleman and precinct supervisor who helped me vote back in '08, on that precinct's machine when it misprinted 4 of my votes, was eager to give it another go. So we did. Leading to...

Failure #3: This time, as the poll workers carefully tried to set up the computer to begin the audio balloting procedure (remembering what had happened in '08), all seemed well after I finally began to vote again about 15 minutes later. But when the options came up in the audio ballot for the very first set of candidates in the Gubernatorial race, it became clear I had been given an audio ballot for the Peace and Freedom Party's primary, as opposed to the primary I was supposed to be voting in!

The fine gentleman --- (Dale, who happens to play Thomas Jefferson in a one-man show --- see for more details!) --- made more calls to the County Registrar's office to figure out what was going on, leading to...

Failure #4: After trying again, and getting the Peace and Freedom Party ballot again, things get murky. One of the poll workers suggested the problem was that they weren't given the codes to specify which party the ballot should be for when loading up the audio ballot. Indeed, I was shown the reference sheet they had for "Operating the Audio Booth" and the "Audio Ballot Codes" for each party were left blank!...

The poll worker told me she had asked about that during the training, but had been told that the numbers would be there on Election Day. They weren't. But seemingly it wouldn't have mattered anyway, due to...

Failure #5: There was no place for the poll workers to specify the party selection when programming the e-voting system's audio ballot! So, presumably, even if they'd had the codes, there was no place for them to enter them into the system!

Here are the screens they were given --- from the beginning of the set-up for the audio ballot, to the end, when the ballot was made active so I could vote it. The bottom half of the screen, in the section titled "Audio Ballot," is what they use to program the audio ballot. The top is used for monitoring the paper ballots of other voters as they place them into the optical-scanner, part of the same computer, at the same time.

1) Hit the "Start" button...

2) Enter in the Precinct's ID number...

3) Select the audio language for the ballot...

4) Hit "Play Audio"...

5) And this is finally seen as the voter votes...

As you can see, the poll workers were never given a screen to input an option for a political party, even if they had been given the audio codes for it.

Failure #6: At this point, poll chief Dale was told by the person on the other end of the phone at the County Registrar's office that she would check with her superior and call us back. Thirty minutes later, the phone never rang. So, after more than two and a half hours trying to vote on the assistive e-voting device for voters in Los Angeles, I gave up, voted manually on a paper ballot, and ran it through the scanner. The ballot will either be counted accurately or not at this point. Unfortunately, there is no way to know how the computer op-scanner will record my votes. But at least I was able to select the candidates and ballot initiatives on the ballot, as I'd wanted. Whether those selections ever get counted accurately, or at all, is now out of my hands entirely, I'm sorry to say.

The ES&S InkaVote Plus system used in L.A. is a disaster. Today has proved that yet again. The County Registrar, Dean Logan, has been in the process of seeking out a new system for the county. I have been invited to participate in a number of public meetings with "stakeholders," as the county looks at various options for a new voting system.

I continue to hope that Logan will have the common sense to run a hand-marked paper ballot, hand-counting pilot program before spending millions on a new e-voting system which is likely to be as disastrous as this one --- if my years of covering virtually every available system, in virtually every jurisdiction in the country, is any indication. I have become convinced over the years, that hand-marked paper ballots, hand-counted at the precincts, in front of the citizenry and their video cameras, are likely the best, most efficient, most accurate way to both cast and count our ballots. Democracy's Gold Standard. It's certainly the most transparent way, and the only one which the citizenry --- who (theoretically) own these elections --- can fully oversee.

In any case, due dilligence would demand testing such a system before spending millions of dollars on a new e-voting system. Or, so one would think. At an event several months ago, while chatting privately with both Logan and CA Sec. of State Debra Bowen, the Secretary concurred that there is nothing in the state's election code that would disallow such a pilot project.

I can only hope Logan works with Election Integrity advocates to carry out such a pilot program --- finally --- in the near future.

Today, I had been fully expecting things to work just fine on the ES&S e-voting system, particularly given the diligent attention that Logan and the technicians at the Registrar's office had devoted trying to figure out what had gone wrong the last time I'd had a similar experience back in 2008.

Last time, the failure was ultimately attributed to "human error," as it was discovered the poll worker had entered an incorrect number into the audio ballot device. I had presumed they'd have re-programmed the systems --- two years later --- to remove the possibility for such "human error," since it should never have been allowable in the first place. Apparently, they didn't. Or maybe they did. In any case, we had a complete failure.

But the real "human error" here is by the humans who have mandated and/or chosen to use such ridiculously complicated and un-transparent devices in our democracy. Again, it's the voters who pay the price for such errors, while the private companies make millions of dollars bastardizing our once-public elections. Also paying the price: the candidates who might have received votes today but didn't, because the voters who used these devices had no idea that all of their votes may have been entirely miscast for candidates and initiatives those voters didn't actually support.

(Were you a candidate in a close race in Los Angeles today? You may want to check how many audio ballots were cast in your district to see if it might have made a difference in the results of your race!)

Last time this happened, I wondered aloud if the attention given to the problem I'd had, by both the Registrar's office and the office of SoS Bowen, would have been given to any voter --- even one who didn't happen to be a journalist with an expertise in issues of Election Integrity. It seems unlikely most voters would have even bothered to contact them with details of the problem. More disturbingly, most such voters using these machines are blind, so it's rather unlikely that they would even know there had been a problem at all.

In the meantime, though I've contacted both Logan and Bowen, I've yet to hear back tonight with comment from either office. [6/9/10: See UPDATE below for response from Logan.]

And the awful irony continues...

* * *

UPDATE 6/9/10 12:12pm PT: L.A. County's Registrar/Clerk Dean Logan emails a short response to my request for comment, explanation:

As is our practice, we will review the detailed account of your experience at the polls yesterday. We will perform due diligence in assessing what occurred, contacting those who were involved and determining what follow-up is necessary or appropriate.

Have yet to hear anything back from Sec. of State Debra Bowen's office. Will update when/if I do.

UPDATE 7/7/10: L.A. County's analysis of the problem I encountered this time around gets chalked up to "human error" (not mine) and "design weakness" in the ES&S InkaVote Plus system. Full details now here...

* * *

Please support The BRAD BLOG's fiercely independent, award-winning coverage of your electoral system, as available from no other media outlet in the nation, with a donation to help us keep going (Snail mail, more options here). If you like, we'll send you some great, award-winning election integrity documentary films in return! Details right here...