Yesterday was my fourth time volunteering in a polling place, and my second time as an Assistant Chief. Unsurprisingly, it was quite a day. Our precinct consists of 2,460 registered voters, and we processed over 700 of them (28%) in the first two hours. Those first couple of hours were just insane. Lines filled the 40 x 80 gymnasium, then stretched a quarter-mile outside (in a light rain).
About an hour into this, while constantly being pulled in a dozen different directions, I asked myself, “Is it going to be like this all day? What if we lost power right now?”, then threw up in my mouth a little.
The touch screen machines, which I despise (previously) held up fine, but didn’t help us move voters through as quickly as we would have liked. Some voters fly right through the electronic ballots, but all it takes to slow the line to a crawl is for a few confused voters to stand there gaping at the machines like tapeworms attempting neurosurgery. I don’t blame the voters, I blame the machines. They need to go, and happily, they are.
Another complication was the fact that we ran four registration tables this election (A-C, D-K, L-R and S-Z). D-K consistently ran double to triple the length of the other lines and, naturally, people in that line got pissed as they saw people who came in after them get through other lines and out the door before the D-K folks had even reached their registration table. We couldn’t break up the distribution differently for complicated reasons, so we just had to roll with it. If it is not immediately apparent, the reason that we do not simply run four A-Z lines is because we wish to prevent people from voting four separate times over the course of the day (I believe there is some sort of rule against this).
I had a chance to talk with a representative from the registrar’s office later in the day, and asked what method they used to choose the alphabet distribution. It turns out they go by the thickness of the poll books. Not the most precise method, but not a terrible one either. The problem is that even if we divided up the poll books perfectly, one cannot predict how many people with surnames starting with “B” versus those starting with “G” are going to show up on election day. Some voters complained that we should have gone with a “first come, first served” rule to be “fair”. I held back from sharing the observation (given the overall reduction in throughput that individual “fairness” would have caused) that if they were standing a quarter mile back, outside, in the rain, their perception of fairness might differ.
I don’t really blame people for being cranky, although I would love for them to get a taste of an election from my point of view. After all, they just wanted to cast their ballots, then get on with their lives. And lots of people went out of their way to thank us (17 of us, and four student pages) for volunteering, which is always swell. The volunteers and pages were indeed fantastic. I am grateful to and proud of each and every one of them.
A little after 8:00am the line died down. For the rest of the day we enjoyed a sane flow of voters, and closed the polls promptly at 7:00pm. All told (including absentee ballots) 85% of our voters turned out. And I’m proud as hell of them, too.
You realize, of course, that the new voting machines use the rfid built into your new, required, gov’t ID.
Seriously tho, what are they going to get this time? Paint for your thumb?
MrPikes reply on November 5th, 2008 6:05 pm:
Whatever works, brother.
Paper ballots (of the marking kind, not the punch) coupled with optical scanners, to my mind, satisfy the following five fundamental requirements of a meaningful election (mine overlap with and extend Schneier’s):
1) Accuracy – The method of capturing votes must reliably capture each voter’s intent.
2) Anonymity – A voters’ identities must not be connected with their ballot, allowing voters to express their intent without fear of retaliation.
3) Auditability – The outcome must be provably true. For instance, paper ballots make recounts possible. Purely electronic voting machines do not.
4) Scalability – Voting systems need to be flexible enough to accommodate a variety of scenarios, e.g., several different contests on a single election day, massive numbers of voters, etc.
5) Trust – A voting system must be intuitive to use and (within reason) understood by every citizens using it.
Bonuses (Boni?):
6) Speed – Optical scanners make it possible to report returns quickly. Were it not for our pitiful attention spans, this would not even be worth mentioning.
7) MrPikes’ Law – Use only as much technology as a problem requires. Anything beyond that increases complexity and invites unintended consequences. For instance, you may have heard that some Virginia precincts had a problem with the paper ballots yesterday… Voters (having been out in the rain) got their ballots wet and they gummed up the scanners. So what is MrPikes’ high-tech, multi-million dollar, super silver bullet countermeasure for next time?
Bring a fucking iron.
Special K reply on November 5th, 2008 9:38 pm:
But might not the marker run in the rain, the iron then sealing the fate of the poor, damp ballot? Maybe the real answer is manditory umbrellas at all precincts.
I gotta say that I like the touch screen, but I also understand the issues.
Hey, Mr. Pikes. I hear the registrar’s office is asking to use electronic pollbooks. I know you hate electricity when it comes to voting, but with epollbooks there will be no division in the alphabet. 3 epollbooks linked together where anyone can check-in. Let me know what you think about it.
MrPikes reply on November 6th, 2008 4:23 pm:
Thank you for writing. Virginia passed legislation in July 2007 stating that voting machines were prohibited from communicating wirelessly while the polls are open:
24.2-625.2. Prohibition of wireless communications on voting equipment:
I guess it depends on whether or not interpretation of the law considers the e-poll books as part of “voting machines”, but I think that they should be, since they are vulnerable to compromise and can influence the outcome of an election.
Avi Rubin (an electronic voting system expert) had direct, negative experiences with them in Maryland elections (search for “electronic poll books” in the linked post below:
http://avi-rubin.blogspot.com/2008/11/my-day-at-polls.html
My stance remains the same – use exactly as much technology as a problem requires.