Data Purge: Tech as a Partisan Tool

I first read about voting data purges in 2018. It is an effort to remove legally registered voters form the polls in efforts to, in theory, clean up records of people who have moved to different addresses. In practice, it prevents infrequent voters from voting and it attempts shape the electorate in order to ensure that those who would vote in favor of other candidates will have a difficult time voting, if they are able to vote at all.

Data purges are led by state officials, in the case of Ohio, the Secretary of State. The reason this administration has given for the undertaking is to remove names of people ineligible to vote. However, given that the process and the data system has had errors—up to 20% in 2019—, it is surprising that officials have continued to commit these actions, which hinder 14th and 15th Amendment rights.

As diverse as we are, American citizens have one thing in common—the right to vote. No elected official should impede that right.

The Columbus Dispatch reported in January 2020 that Ohio’s Secretary of State purged over 460,000 registrations, some in error. Many of those people found out when they got to the polls that they could not vote. At that point, it was too late. In August 2020, a list of over 100,000 soon to be removed registrants was published by the Secretary of State. In January of this year, a list was published of 98,000 additional voters scheduled to be purged. According to one report, registered Democrats were purged at disproportionately higher rates.

Ohio is not the only state where this has or is happening. Wisconsin is among many with elected officials that have taken steps to purge voters. Going further, state elected officials are enacting legislation that places further restrictions on voting freedoms, primarily because of the false narratives of voter fraud promoted by the previous president.

There is > 1 Way to Identify Ineligible Voters

There are many alternatives to deletion from a database. One method would involve a requirement to cross reference the state taxpayer database, and/or the IRS database before purging. Employment and tax address data is accurate and can easily be cross-referenced with SQL queries. But really, leaving voters’ data in place is the best option. With all of the voter ID requirements in place (state issued ID, utility bills, etc.), anyone misrepresenting an address will be caught. It calls into question the true motives.

Fight Back with an Old School Tool: Communication

Anyone who is online, knowledgeable and able to search for registration status will not be harmed by voter purging. The concern is instead for the populations who have no access to newspapers, transportation, computers or Internet, or are not able to view their status for any reason.

All of us in the tech community can help. First, register to vote if you have not done so already. Then, get online to verify that your Secretary of State lists you as an eligible voter, months ahead of any scheduled election. If you’re in Ohio, visit the Ohio SOS site.

While you’re checking your own status, verify the status of any relatives, neighbors and friends. Call them, or visit them if safe, and let them know you are checking for them. Teach them how to check for themselves.

If you are looking for a way to flex your tech muscles, volunteer! Check in with the local League of Women Voters, MoveOn.org and other voting advocacy groups to attend meetings or volunteer. Every bit of support will help. Let’s share ideas!


This post is political in nature and I normally do not wax political. However, the fact that elected officials have committed to data projects like this is disturbing. On a personal level, I will look to vote for candidates who work in everyone’s best interests, and who demonstrate that they want every eligible citizen voting age to be able to vote. The current officials have proven to me that every vote does indeed count—and they don’t seem to want it to.

Published by Be better.

Working on social and economic progress.

%d bloggers like this: