← Back to Posts

The voter is the customer: Ginny Badanes on designing an election around people

Sean Roberts•September 1, 2020•3 min read

Ginny Badanes and I have been talking about election technology for more than a year, and one project keeps coming up: Los Angeles County's Voting Solutions for All People, the system the county designed for itself and used countywide for the first time in the March primary. It was long-running and expensive, and it had problems on the day, which I have written about separately. But I asked Ginny what she thought was most innovative about it, and her answer was not a machine.

The ways in which you might draw some parallels to some of the efforts that we've been doing, though on a certainly much smaller scale, is really focusing on the voter. What are the needs of the voters? How do we bring them technology and solutions while still paying attention to the big things such as security and enfranchisement?

Enfranchisement here means not removing anything for people on the margins, including voters with disabilities. The equipment marks a paper ballot that the voter then feeds back in to be counted, so a voter who needs help reading or marking a ballot uses the same system as everyone else.

What impressed her most was how the county tested it. Before anyone voted, they held block parties and mock elections and did serious user testing, to make sure the system worked for the people who would actually use it. Then there are the languages. Ginny had seen a demonstration, and she had not appreciated until then how many languages a voter can get a ballot in. Printing paper ballots in every one of those languages, with enough copies for every voter at every location, would be, as she put it, quite a feat. Doing it electronically is one of the places where technology plainly makes the system better for voters.

The county also moved to vote centers. Instead of one assigned polling place, a voter can go to any vote center in the county during the days before the election. That choice was made for the voter's convenience, and it is also what made the system harder to run, because every center needs an up-to-date record of who has already voted.

Near the end of the conversation I tried to name what struck me. Most election technology is bought by the office that runs the election and designed around that office's needs. The county's project treats the voter, not the election official, as the customer, and that is a huge innovation. It shows in the public meetings the county has held and the time it has spent learning what voters need. I would like to see that product management experience spread to other teams and to new open source projects.

The system is not open source. Ginny pointed out that it is built with many open source components, and that the county has said publicly it would like to make it an open source project. That is worth following, because the most valuable thing in it may not be the code. It is the habit of testing with the people who will use the system before the day it has to work.

The other half of our conversation was about that day, and what went wrong at the vote centers in March: test it a million times and the poll book fails.

From Lincoln Shorts, the interview series I hosted for Lincoln Network, recorded 1 September 2020.

Share this post