Home > Politics > E-voting Failure!!!

E-voting Failure!!!

As I promised Brad Friedman a few months ago, I took half the day off in order to allow time to use the handicapped assistance device at my polling place. I had to give up after 3-1/2 hours and vote the normal way. Read on for details…

Where to begin?


In June of this year, after carefully reading a few of Brad Friedman’s posts on his attempts to use assistive devices in multiple elections, I decided I’d do the same, and  I posted this to my blog.  For full background, be sure to see that previous post – I’m not going to repeat it here!


My FIRST mistake was thinking it wasn’t going to be bad and NOT getting food before I headed to the polls when I left work at 11:00. Ultimately, it was the need for food that caused me to abandon my attempt after almost 4 hours!

Machine: ES&S AutoMark- only marking on machine, but per listing at California SecState website, it was the A100 variant – but there was no marking to that effect.

<<side note to Brad: Because the AutoMark only MARKS ballots, but doesn’t tabulate, I don’t think the “3 people must use the machine to preserve individual voting secrecy” idea applies. Unless the machine remembers the votes! The marked ballots are tabulated with the same equipment used by ALL voters in the precinct>>

11:20 AM I walk into my polling place in Contra Costa County, California. I’m not mentioning the precinct number here, but I will when I reformat this as a formal complaint to the County Registrar of Voters, the California Secretary of State, and the 1-866-OUR-VOTE hotline.  Staff friendly. Handed me ballots and had no problem with directing me to the AutoMark machine. One thing I DID notice – because Brad asked me to check – was that the machine was set up on a table, with no way to adjust the height. I immediately saw that voting would be a mechanical problem for people in some of the bulkier power wheelchairs. <<Fast Forward here: I later saw a diagram of the “clear space” guidelines for the poll workers controlling placement of the machine – and the placement was NOT in compliance!!>>

11:25 I feed my first (of 2) ballot cards into the machine for “scanning”. The machine grabbed the ballot, said it was scanning it, then gave a “scan error” message and returned the unmarked ballot, and directed I try again. THIS time, it took the ballot and gave a “Paper Jam” error – “See Election Official.

11:30 Election Official didn’t know how to clear the jam! It was quickly obvious that the stub (to be removed by the voter) had separated and was sticking 1/2 way out of the top of the machine, but held tightly. She then tried various things with her key that turns the machine Off/On/Test. In the process, when she went to test, a number of options came up on the screen – and she completely ignored them! She gave up after rebooting the machine and getting a “Scanner Error – Call for Assistance” Finally, at about 11:35, she called for a “Rover”, and was told the rover would be there “shortly”.

11:50 No rover yet, precinct worker called again – and it took her 15 minutes on hold before she got a real person! <<Fast Forward: It turns out that the number given to the precinct workers was the SAME NUMBER as is given to the general public!>> This time, she was promised a call back within 15 minutes – and the call back never came.

12:40 Still no rover. This time I tried calling the county number published on the web, and I gave up after 8 minutes on hold as my cell battery was getting low – and I wanted to maintain a reserve!

1:30 FINALLY! The rover arrives. After having to go back to his car to get his glasses, he apparently clears the paper jam in about 2 minutes without tools. And reboots the machine — only to get a “Error calibrating Scanner – TOP AND BOTTOM” message during the boot process. HE calls for technical assistance from an ES&S rep, talks to someone for a short time, then is promised a call back!

2:00 Call back comes. Technician suggests that there might still be a jamb, so the rover and I BOTH look over the machine again – and we discover that a ballot stub from the one person who used the machine before me had been torn off inside the machine and was ALSO jammed. You see, when we first cleared the jamb, it was obvious to remove both the jammed ballot – and the stub that was sticking out the top. It didn’t occur to us to look for anything more!

2:10 Machine now successfully reboots and I can start the process again. By now I’ve “spoiled” the ballot card that originally jammed and had a new one. It fed in just fine and after a few minutes, I was able to actually start voting. I initially went slowly and carefully, until I convinced myself that it was asking for the races in the exact order they appeared on the sample ballot – and the options in each race were presented in the same order. I also intentionally tried to overvote and it (correctly) didn’t let me.

<<digression>> There hasn’t been a good place to insert my distaste with the synthesized voice. I could hear it and understand it, as well as change the tone and tempo, but I’ve heard far better synthesized voices in other contexts and this one was truly annoying! <<end digression>>

My first REAL criticism of the actual voting process is that you are NOT prompted well if you want to UNMARK a previous mark. I was able to do it easily, but only because I could SEE it. Had I been relying only on the voice prompts (as a blind voter must), unmarking would have been very problematic.

After I’d voted 5-6 races, and in the process intentionally tried every user error I could imagine, I raced through the remaining races on the first ballot, reviewed my full ballot (A-OK) and hit submit. After about 2 minutes, the machine flashed “Printing Error – See Election Official” – and there was NO voice indication. Had I been blind, I don’t know how long I would have waited.

Well, this was ANOTHER error that the precinct workers didn’t know how to deal with. THIS time, they called the rover directly.

2:30 – Rover back. He consulted the “Troubleshooting Guide” which made no mention of this error, so he turned the machine off, then back on. Machine booted (very slowly) and, eventually, spit out my ballot – marked on only one side!

At this point, hunger got the better of me and I completed the ballots by hand, turned them in, departed the precinct at 3:00, went to the nearest food (Burger King – Blech!) ate, then went straight home to write this. The precinct workers did NOT try with a complete ballot – which (unless the law directs otherwise) they could easily have done by simply taking a ballot, going through the process, then ripping up the marked ballot. So when I left, the precinct workers had NOT assurance that the “printing” problem wouldn’t happen again!

After all was done, it was apparent that the ultimate source of the problem was mechanical: On the 2nd ballot cast by the voter BEFORE me, the machine retained the ballot stub which proceeded to gum up the works for me. This was compounded by bad decisions by the precinct workers AND by NO training in troubleshooting the AutoMark machine.

What was the bad decision? Now is the time to describe the physical ballots: 2 Large Ballots (approx 8″ x 16″) with races on both sides of both ballots. Each ballot had a stub, to be detached by the voter. The AutoMark is NOT a DRE machine – it simply MARKS the ballots and returns them to the voter to be inserted in the same tabulator used by all voters.  But get this: the precinct workers were tearing the stubs off PARTIALLY before handing the ballots to the voter! In chatting with the workers during the long hours waiting, they related that they did it because they were getting an initial high rate of “Ballot Stub Attached” errors when voters fed their ballots into the tabulator, so they started the partial detachment as a visual indicator — to the voter – to remove it completely after voting, but before insertion in the tabulator.

Thus the bad (and possibly illegal!) decision on the part of the precinct workers was to partially tear off the stub to avoid nuisance rejections at the tabulator – ones that would have been quite easy to avoid by simply leaving the stubs fully attached, then reminding voters to remove it before insertion!

Was the retained stub in the AutoMark one that had been partially detached by the precinct workers? Frankly, I forgot to ask. But I know that I didn’t completely detach it before I fed MY first ballot into the machine. The precinct workers tried to put the blame on ME for not detaching it saying (and I paraphrase) “Well, if you had been blind, we would have made sure you removed it” – but there was NO direction to either remove, or retain the stub before inserting the ballot into the AutoMark.

Here’s some additional observations:

  • NOBODY in the precinct was trained in troubleshooting problems with the AutoMark. The jambs were cleared by easily removing a couple of plastic parts – without tools – and without exposing any of the sealed portions of the machine – looking carefully and removing everything (and I note that both the Rover and I didn’t look carefully, so compounded the problem!).
  • Even the ROVER didn’t have a good troubleshooting guide, for either the “Calibration Failure” or the “Printing Problem”
  • I was frankly angered at the NON concern of the precinct workers displayed about the problems! The said “These Things Happen” and “Wouldn’t you know it, Murphy’s law again” (The worker who uttered this piped up about a minute later – “Just Kidding” as if she realized how stupid she sounded. And at the total disinterest in discovering if the “printing problem” would affect the NEXT person who tried to use the AutoMark.
  • Both the precinct captain and the rover were far too quick off the mark to try power cycling the machine, instead of tracing down the real problem or even reading the screen prompts!

As I said before, I’m going to reformat this slightly as a formal complaint and send it to interested parties.

I’ve omitted conversations I had throughout this on the phone, through Twitter, and through Email with Brad Friedman of The Brad Blog and Emily of the Velvet Revolution. Emily was in the process of getting a Video team to me to document this when hunger forced me to stop.

Will I do this again? YES. I’m particularly interested in how much more complex the process is in a primary election, where there are multiple partisan ballots! But next time I’ll eat first! And go with a spare cellphone battery!

  1. 03 Nov 2010 at 17:38

    I am troubled by “I was frankly angered at the NON concern of the precinct workers displayed about the problems! The said “These Things Happen” and “Wouldn’t you know it, Murphy’s law again”

    I have been working at the elections for a number of years. I have NEVER experienced a poll worker with that attitude. At yesterday’s (11/2/2010) election, I was the Chief Inspector at a polling place. If a worker had exhibited the kind of attitude you relate, he or she would have been taken aside and given “feedback”, the second time he would have been shown the door.

    Our county elections staff provide excellent training and an amazing concern for both the integrity of the process and the voter. Poll workers likewise. Remember that poll-workers are essentially volunteers–paid below the minimum wage as a token and few have the luxury of doing it repeatedly.

    As to the Accumark machine, your experience does not surprise me. The Automark machine was never actually used by a voter. However, it did fail in it’s boot-up and required the Rover to come to replace parts. In other elections I have only seen three voters use it. In one out of the three instances it did jam. Fortunately we were able to clear the jam, but 33% failure should be unacceptable.

    The AccuVOTE machine which does scan and count the ballots is also jam-prone.

    • Steve
      03 Nov 2010 at 17:52


      I’m guessing (from the Area Code listed for your counseling business) that you are in either San Francisco or Marin County. And I’m glad to hear that you (and others) are well trained and that you are a conscientious Chief Inspector who wouldn’t tolerate some of the stuff I experienced. I wish the same could be said for Contra Costa county – at least the single precinct example I experienced!

      One person, late last night, related to me that the Contra Costa registrar of voters MIGHT be someone with a long history of e-voting enthusiasm combined with a similarly long history of undertraining poll workers! I say “Might” because I’ve been provided with the name of a registrar who most certainly had both characteristics, I just don’t know if it’s the current Registrar! I plan more research this weekend on exactly this.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: