Election workers and voters rely on accurate voter registration data. Voter records missing from the poll books wreak havoc at the precinct on election day, as the following interview illustrates. If this sort of unreported chaos occurred at multiple precincts, it may have prevented thousands of voters from casting ballots in 2016. With over 6,000 polling places and a margin of victory of only 44,000 votes in the 2016 presidential race in Pennsylvania, problems at selected polling places could have changed the results of the election. What if 1200 precincts (20% of the total) were disorganized enough that 40 voters in each were too inconvenienced or overwhelmed to cast a ballot? That would be 48,000 voters technically disenfranchised by problems stemming from erroneous voter registration data.
How well do our counties and states safeguard our voter registration data? How easy would it be to change the results of an election by manipulation of this data? This article examines those questions.
Is there any evidence of unauthorized changes to the Pennsylvania voter registration data from around the time of the 2016 election? This article presents some disturbing data findings.
Based on this work, we have two strong recommendations.
- Every state should create a system to monitor changes in voter registration records and to alert each voter when the system detects any change to a record.
- Every state should create a system that allows poll workers and the public to report problems with the registration data at individual precincts.
Interview with a Pennsylvania Election Judge
What is your election-related job title?
Judge of elections. It’s an elected position.
How long have you held this position?
Since November 2008.
Does your county use paper poll books or e poll books?
Paper poll books. We have two poll books used to sign people in, plus a numbered list of voters that is more complete than the poll book. If someone isn’t in the poll book we first check the numbered list and if that person appears there we enter his or her name and voter ID in the poll book by hand. No ID is needed unless it’s that person’s first time voting. There is a note in the poll book if that is the case.
The “minority inspector”, in this case, a Republican, has a poll book. The “minority clerk” has the numbered list. There are so few Republicans in the precinct that we sometimes need to go to another precinct to find a minority inspector and a minority clerk.
Election Day Chaos: Missing Registrations
What was different in 2016?
2016 was very different from previous years.
Unlike other years I spent a very large amount of time on the phone (mainly on hold) with Voter Services. I actually pulled my phone records to see how many minutes I was on the line with them and how many calls I made. 184 minutes on the line with them, in the course of 18 phone calls. During my lunch break another poll worker took over and she must have made at least two or three calls as well. At least two to four people stopped waiting and left each time I was on the phone. I saved the phone numbers of several people who left and wanted to come back and vote once I got through. Later I was not able to reach all of them.
Compared to other years, Voter Services was very poorly prepared and/or trained.
For instance:
If a voter is not in the book and also not on the supplemental alphabetical list that includes people who registered or moved into the county too late to be included in the pollbook, I had to call Voter Services to see if they could retrieve the person’s ID. No one can vote without a voter ID for tracking that person.
Voter Services asks for first name, last name and birth date. Several times they could not find the voter given this information, but they did find the voter when I asked them to leave out the birth date. Most election officials would not know to do this.
A voter came in at 7:50, ten minutes before the polls closed. He wasn’t in the poll book so I called Voter Services. They didn’t pick up until after 8, and the said the person couldn’t vote because the polls were closed. Which of course is false because the person entered the polling place before 8:00.
There was a student going to school in Westchester – two hours away. He had never registered there, but his address listed him at a Westchester precinct. He didn’t vote because the distance was too far for him.
The system is so bad already, it’s always very chaotic. Any disturbance to the system causes people to walk away.
None of these problems were due to there being more voters. There was about same turnout as 2012.
How many people do you think left without voting?
That’s difficult to estimate. Because so many people were not listed in the poll books a long line formed of people waiting for me to get through to the county’s Voter Services. For almost the whole day there were at least three people in line, sometimes ten or more. Many people left instead of waiting. I would estimate that at least 30 people left instead of waiting in that line, more likely 60. That’s not counting people who left instead of waiting for help with provisional ballots, or simply saw the long lines and general chaos and left without even trying to vote.
(Note: Given that there are over 6000 precincts in the state, if an average of only 10 people per precinct left without voting due to problems with the poll books, that adds up to around 60,000 voters who were unable to cast ballots.)
Did it seem that more voters of one party than another were missing from the poll books?
It’s a heavily Democratic precinct – there are so few Republican voters that it’s hard to guess accurately whether the same proportion had trouble voting. Good guess that most people who left were Democrats.
Were the people who were missing from the poll books offered provisional ballots?
Yes, but provisional ballots are hard for people. There is a lot of information to fill out and a poll worker is required to finalize it. Poll workers were swamped, so there was a wait for that too. A lot of people decide not to vote instead of filling out a provisional ballot. They have already waited a long time, and some people say it doesn’t feel like they really voted unless they use the machine.
How confident are you that these ballots are counted?
I strongly suspect provisional ballots may not have been counted.
Other Problems
Did you notice any other irregularities besides the poll book issues?
Not really. There were several voting machine issues – about the normal number. If a machine malfunctions some of the time I can fix it, other times I need to make a phone call and get help figuring out how to fix the problem. When we downloaded the machine results there were very few votes for the Republican candidates, but that’s to be expected because the precinct is so heavily Democratic.
Were there any people hanging around looking suspicious?
The poll watchers were terrible this year – very suspicious of everything. Democratic poll watchers were claiming that we were favoring Republicans, which of course was not the case at all.
I actually had to oust someone this year. A Democratic poll watcher was being very belligerent and even kept trying to get behind the voting machines.
Did other poll workers you know report similar trouble with voters being missing from the poll books and trouble getting through to Voter Services?
Other precincts? I don’t know. I know other election judges from meetings we all attend, but don’t know them well. But if I wasn’t able to get through to Voter Services, no one else in Montgomery County could get through.
The Aftermath
Were you concerned about the results of this election? Did anything make you worry that the results were not accurate?
Like everyone I know, I was very surprised that Pennsylvania voted for Trump. We have been a blue state for several decades now. One thing I thought was interesting is that Josh Shapiro, a Democrat, was voted in as attorney general. It makes no sense that we would vote for him and also elect a Republican President.
Did you voice your concerns to anyone else?
I told [name redacted] – she runs the Democratic committee – a while afterward, maybe a few weeks. Didn’t officially report the problems we had to anyone.
Were your concerns dismissed or addressed?
She just listened. Probably not much she could do.