Jeremy Epstein on “My Day as a Pollworker”

[Image via truststc]

My friend and colleague Jeremy Epstein is well-recognized as a leading thinker on voting technology and security. But he’s also one of those electiongeeks who “walks the talk” and is a long-time pollworker in his home of Fairfax County, VA. He recently shared an account of his day at the polls, which has lots of fascinating detail on the issues that are likely to get attention as turnout spikes this fall:

I’m the “chief election officer” for my precinct (841 Popes Head in Fairfax County VA).  My team included an assistant chief election officer (required to be of the opposite party from the chief), and six election officers (I think there were two Ds, two Rs, and two Is).  The rules in Virginia are a bit different from some states – since there’s no party registration, when you sign up to be a pollworker, you state which party/parties you’re willing to represent; those parties have the right to accept or reject the application – although rejection is fairly rare.  I wish we had some high school students to help, but the decision to close schools on election day was only made a few weeks before the election, so too late to recruit them.

On to my report.  My precinct had 1711 registered voters, of whom about 50 had voted absentee in advance.  (Virginia does not have no-excuses absentee or early voting, but it has absentee by mail and in-person absentee, which is sort of like early voting except that it can only be used by voters with one of the approved reasons.)

As noted above, Virginia has no party registration, so voters state which primary they wish to vote in.  This is basically an open primary, but since there’s no party registration the term “open” is sort of redundant.  The precinct is significantly more R than D (I think it went for Romney by 60% over Obama), and is one of the most Republican precincts of the approximately 250 in the county.

There were about 750 in-person voters (I’ve forgotten the exact number) or about 43%, of whom slightly more than 2/3 chose to vote in the Republican primary.  This was consistent with other places in Fairfax County, where the participation in the Republican primary was higher than would be expected.  Whether this means more Republicans turned out, or more Democrats voted in the Republican primary is impossible to know for sure (again, since there’s no party registration), but exit polls I read indicated lots of Democrats were crossing over.

Our equipment was laptops set up as EPBs [electronic pollbooks] (I think they’re Lenovo, but forgot to look), ExpressVote for voters with disabilities, and ES&S DS 200 scanners.  We were set up with two scanners – one for the Republican primary and one for the Democratic primary.  Each was set up to count both ballot styles; the reason for two scanners was simply in case a recount was needed, only the one box would need to be opened.  Or at least in theory – despite large signs and a pollworker standing by the scanner, six ballots were put in the wrong scanner.  Interestingly, all six were Republican ballots put in the Democratic scanner; whether this was a meaningless error or a deeper sign (e.g., voters who thought of themselves as Democrats putting their Republican ballots in) I have no idea.  Or maybe Democrats are just better at following instructions 🙂

There were no significant issues.  One of the EPBs would occasionally hang up, but after a minute or two of thinking, recovered just fine.

Turnout was steady all day.  The checkin line was never more than five or six people, and nearly always no more than one or two.  We had set up 12 marking booths; given that there was just the one race on the ballot, many voters didn’t even bother to sit down.  I don’t think we ever had more than five booths in use.  The only time we ever had a line at the scanners is when one of our brand new pollworkers decided to give a detailed explanation to a voter of how to insert the ballot, instead of just “slide the ballot on the gray plastic; doesn’t matter what direction”.  (Various people … have noted the time it takes the DS 200 to scan a ballot; I did some quick timings, and it was MUCH faster than others have reported – no more than 10 seconds from when one voter started scanning until the next one started.  I strongly suspect that the time is proportional to the complexity of the ballot, which might be worth considering when localities are trying to decide the capacity of the system – it’s not a fixed number.)

We were concerned that voters would complain that there were candidates on the ballot who had dropped out (e.g., Christie, O’Malley).  The county provided us a bright yellow form to put on the sign-in table explaining the story (i.e., ballots were printed based on who submitted paperwork, and most of the drop-out candidates haven’t actually withdrawn but just suspended their campaigns).  There were a couple of questions, but no one made a fuss about it.  Interestingly, there were a total of about 10 votes for candidates who had dropped out.

We had LOTS of first-time voters, which I find personally very satisfying.  It was certainly a proud moment for me as a parent when I took my kids to cast their first votes!  The young people were very proud to be participating.

Another concern was about potential campaign observers – each candidate was entitled to two observers, and with 13 R and 3 D candidates on the ballot, we could have had 52 observers!  However, that was an unwarranted fear; we never had even a single observer.

There was some concern about a month or two ago that there might be protests or other issues relating to the election, and pollworkers had been told that police were on stand-by for any issues.  However, other than voters expressing disgust with the choices, there were no issues.

And speaking of disgust with choices, there were at least a half dozen voters who hadn’t decided which primary they wanted to vote in, even when they got to the checkin table!  One voter wanted two ballots, so he could look at them both and decide which to cast.  (Obviously, we didn’t allow that.)  There a few voters who wanted to know why we asked which party, and didn’t they get a secret ballot – we explained that who they vote for is secret, but which primary is public.

The most unusual item all day was a couple who had not moved, but whose address had changed!  A new street was built, and that changed where their driveway connected, which meant they got a new address.  The county election office had sent them a verification of address form, which the post office returned since the old address no longer “existed”, so they were marked as having a questionable status.  So it was a bit of a mind-bender to figure out how to interpret the rules on moving, since they hadn’t actually moved!  (And yes, they were permitted to vote normally, and I had them fill out a change of address form noting that it was the same house, just a new address.) [I’ve never heard of this happening before – DMCj.]

I had one provisional ballot, due to a pollworker error.  The pollworker appeared to have checked in the wrong person in a household; I hope the provisional ballot was counted, but don’t know.  Fairfax County is testing new iPad based EPBs in some precincts (but not mine); that would have reduced the risk because they can scan a Drivers Licenses and automatically pull up the correct voter. [They were in use at my precinct and the speed of check-in was remarkable; it should have a tremendous effect this fall – DMCj.]

And speaking of voter ID, there were a significant number of voters with military IDs, and a few with company IDs, but I had no issues with voters who didn’t have an ID.

No one used the ExpressVote.  There were several voters where I offered it, but they all declined.

We had two curbside voters, and only three spoiled ballots.  I guess with only one contest, it’s not hard to get it right.  No fleeing voters, but one voter cast a blank ballot.

Early in the day we had problems with one pollworker handing voters two or three ballots which stuck together; I emphasized being careful, and had no further problems.

Probably the biggest glitch of the whole day is that I had changed the layout of the room, to allow more room inside the building for lines (in case it was cold and lots of voters).  It turned out to be a warm day and no lines, so it didn’t make a difference – but LOTS of voters got confused, and ignored the “do not enter” sign I made out of tape on the floor, even when I stood by the door and said “follow the blue arrows”!  (It was still worth doing, because I determined that the revised layout did in fact leave more room for lines – which will be useful for November.  And speaking of which, in Fairfax County, each precinct chief is responsible for figuring out the layout for him/herself – I know in some states the election office makes those decisions).

Thanks to Jeremy for this report; that last point – that people often do what they remember rather than read and obey signage – is going to be a familiar story for many polling places this fall.

One person, one precinct, one report – and still, lots to look forward in Virginia and elsewhere in November. Stay tuned …

1 Comment on "Jeremy Epstein on “My Day as a Pollworker”"

  1. As one of the Fairfax County “rovers” who was responsible for supporting fourteen precincts, I found Jerry’s account of March 1 most interesting and dovetailed closely with my own impressions. The one thing I would add is that one of the primary reasons for having the two DS200s was simply that if one went down, there would still be an another machine operational. This happened in one of my precincts. Around 2:00 pm. one of the DS200s stopped accepting ballots. We were able to swap in a new machine in a way that no votes were lost but this took some time so for a while the other machine took both Rs and Ds. Incidentally when doing the canvass the day after it seemed that most precincts had cross over ballots (Rs in the D machine; Ds in the R machine). It was the rare exception where they were perfectly segregated. I think we sort of expected this and I don’t think any one believed that in an event of a recount only half the machines would have to be opened.

Leave a Reply to Bill Lewers Cancel reply

Your email address will not be published.