[See my update here.]
Today, the Boston (Massachusetts) Election Department emailed to poll-workers (of which I am one) an update about the upcoming municipal elections in September and November: we are going to be using electronic poll-books not just for early voting as we have in the past, but also on election day at every precinct.
The “poll-book” is the book that lists every voter in the precinct. Identical books are kept at the check-in and check-out tables, and every voter has to be marked off in both books on election day, at check-in before they are given their ballot and at check-out before they put their ballot into the tabulator.
Up until this year, on election day the poll-books at precincts have been literally printed paper books, while at early voting locations, iPads running custom software were used instead. These iPads have cellular modems in them and talk to a central database maintained by the Election Department (more likely a contractor on its behalf). Using electronic books for early voting is necessary to allow any voter to vote at any early-voting location while still preventing people from voting twice.
I’ve sometimes staffed early-voting locations, and I’ve enjoyed using the electronic poll-books. They are faster than the paper books, and they eliminate some of the drudgery of checking people in and out. I can assure all you voters out there that yes, voting will be easier and faster for you after the switch to electronic books, if they work properly.
That “if” is where things start to get a dicey. There are two really big differences between early voting and election day, and as a poll-worker, I am concerned that the Election Department may not have adequately anticipated and planned for them.
Early voting can fail; voting on election day can’t
If there is a tech issue with the electronic poll-books which prevents them from working properly during early voting, voters can be told to try again at a later early voting day or on election day itself. No early voting day is any voter’s last chance to vote in an election.
In contrast, obviously, if the electronic books stop working on election day, there is no do-over. The Election Department has until 8pm on election day to figure out what’s wrong, fix it, and allow everyone to vote, and the longer it takes for them to fix the problem, the more likely it is that people will be disenfranchised.
Switching back to paper poll-books held in reserve if the electronic books fail mid-day on election day is not an option, because no one who voted before the electronic books failed has been marked off in the paper books, which means it’s impossible to prevent people from double-voting.
There is, at least theoretically, one other option, which is to switch to issuing provisional ballots for every single voter, which can then be verified by the Election Department after election day. I say “theoretically” because issuing provisional ballots take much longer per voter than just checking people in and out in the poll-books, and they have to be counted downtown after the polls close rather than at the precincts. While I wouldn’t say all that is impossible, I’m not at all convinced it’s feasible, at least not if there is a widespread failure of the electronic poll-books, vs. a failure localized to a small number of precincts during a low-turnout election.
There are way more precincts than early voting locations
The most early voting locations the City of Boston typically opens on the same day has been around ten. In contrast, there are 275 precincts, so when we switch to electronic poll-books on election day, there will be 2,740% more poll-books being actively used than during the busiest early-voting day.
Now, I don’t know… Maybe the vendor that the City of Boston contracts with to provide the electronic poll-books and the centralized back-end infrastructure that they interact with has done robust capacity planning and ensured that there is sufficient capacity and redundancy built into the system to withstand 2,740% more traffic than it has ever had to withstand in the past.
Or maybe they haven’t.
Speaking as someone who has been working in tech for nearly 40 years, the latter seems to happen a lot more frequently than the former, especially when it comes to government contractors.
Only time will tell
Some time in the next month or so I will be attending mandatory training on how we will be conducting elections with the electronic poll-books. I intend to ask then for details about both what we will do if the poll-books fail, and what has been done to ensure that the system can handle the increased load of every precinct being online at once, if that information isn’t included in the training. Watch this blog to find out what I’ve learned.
If there are actual problems on election day (in either September or November) I’ll also be writing about that.
[See my update here.]
