As the November U.S. election approaches, anxiety is running high. There are fears of Russian meddling, breakdowns in the U.S. mail system, intense partisan rancor, voter suppression, fraud, and the logistics of enabling people to safely vote in the midst of a deadly pandemic.
So what are the seasoned election scientists’ biggest fears, and what are they doing about it? ScienceInsider asked political scientist Charles Stewart III at the Massachusetts Institute of Technology (MIT), one of the nation’s leading experts in the science surrounding election administration. He is co-director of the recently launched Stanford-MIT Healthy Elections Project, which is working with researchers and election administrators to “ensure that the 2020 election can proceed with integrity, safety, and equal access.” It has, for example, helped connect modelers with election officials trying to figure out how to best place ballot drop boxes and polling places. The effort, Stewart says, “is really research in the interest of action.”
The interview has been edited for brevity and clarity.
Q: What first got you involved in studying how to run elections?
A: I’m the director of the Caltech/MIT Voting Technology Project. We were put together in the heat of the recount in Florida in 2000, with the idea that we would design a perfect voting machine. But it was obvious that the problems of lost votes were broader than just voting machines and largely grew out of human systems, not the technology. I’ve had a long-term interest in trying to create an academic niche that some of us call election science, which applies social science and other scientific approaches to understanding how voters interact with the election administration system.
Q: What are you doing now?
A: The Stanford-MIT Healthy Elections Project got going back in March. We were very concerned that most of the energy around COVID was actually being channeled into litigation, into partisan politics.
We do research to try to understand what is going on in this election because of COVID and to explain that to the world, but also to encourage the meeting up of scientifically, engineering-oriented university people, civic tech people, and election administrators. At the University of Southern California, they have a web-based app to help election officials figure out how to optimally place polling places and drop boxes. We’re supporting a group of former election administrators who are parachuting in and helping local officials in terms of management changes that need to happen very rapidly.
I’ve been doing little bits of research to try to help estimate what voters are likely to do on the upcoming election in terms of seeking mail ballots, etc. I’m putting together surveys to track how voters are thinking about the COVID epidemic where they are and how that influences what they’re thinking about doing in November. [We want] to see how many people right now are going to be requesting mail ballots, how that unfolds during the next 2 months. Then, after the election, asking people what they actually did.
I have a survey that I put in the field immediately after every presidential election day going back to 2008 called the Survey of the Performance of American Elections. I survey 200 voters in every state, asking them about their experience on election day or in general on voting. There’s a series of questions about how long you waited to vote. Did you experience problems in the polling places? Did you experience problems in requesting your ballot? That will end up being the only source of objective data about the experience of voters in the month after the election.
This is going to be really interesting to see how voters have taken in the changing debate and discourse about the pandemic, the right way to vote, and how those two things work together.
Q: There’s a lot of anxiety about ways that the process of voting could be disrupted, whether it’s concerns about the pandemic, the Postal Service, or foreign actors interfering. What are you most worried about?
A: Two things: I’m worried about voters screwing up and I’m worried about not enough poll workers to work in-person polling places to meet demand for in-person voting.
With respect to voters screwing up, what I mean is voters screwing up voting by mail and voters, you know, requesting mail ballots or mailing them in too late, not having opportunities to return them in person. [Common errors on mail-in ballots include failing to sign the ballot and marking votes incorrectly so they don’t register.]
On the in-person side, I think we’re going to have a very high demand for in-person voting. In the primary season the real struggle with keeping polling places open has been with counties losing their poll workers. In 2016, 60% of poll workers were over the age of 60 and 25% are over the age of 70. So the prime poll working population is also the prime population that’s most at risk of COVID. That’s a big thing. Nate Persily [Stanford University law professor and Healthy Elections Project co-director] and I spend a lot of our time on phone calls about poll worker recruitment, which is not necessarily the highest and best use of our time. But there we are, making sure that the various groups, the various websites are actually going to be effective, rather than just simply a feel-good exercise.
Q: You hear people concerned about sort of high-tech hacking or things like that, but yours is very nuts and bolts.
A: This is a logistical issue. That’s how Nate and I approached it since March. This is about blocking and tackling, this is about doing it right. This goes back to the Voting Technology Project. What we saw back then is that, you know, just the little things of administration can trip you up.