"Use It or Lose It": The Problem of Purges from the Registration Rolls of Voters Who Don't Vote Regularly

With the 2020 elections looming, there are mounting concerns about the techniques some states may use to try to tailor the electorate and achieve their preferred political outcomes. One such technique is purging voters from the rolls for flimsy reasons. State election officials do, of course, have the obligation to try to keep voter registration records up to date by canceling registrations of people who have died, are imprisoned, have moved to another state, or become legally incompetent. But a minority of states go further and engage in a practice that ought to be seen as glaringly unconstitutional—purging people from the rolls solely because they have skipped voting in several consecutive elections and they have not responded to a letter asking them to confirm where they live. 

This practice results in the deletion of hundreds of thousands of registrants each year. Very often, those people get energized to vote in a given election but find when they show up at the polls that they are no longer registered and cannot cast a ballot.

Such a practice cannot be justified on the theory that voting rights are a “use it or lose it” proposition. It is well established, under the Constitution and federal law, that American voters have a right to choose not to vote and not to be penalized for doing so. Instead, the nine or so states that engage in this type of purge of registrants say it is justified because not voting in recent elections and not returning a mailed notice is a proxy for identifying people who have moved to a different jurisdiction.

The problem is that this proxy is both highly imperfect and entirely unnecessary as a mechanism for keeping voting rolls up to date. Most of the states have found they can do that job just fine by relying on indicators like the National Change of Address system maintained by the U.S. Postal Service and, in recent years, data generated by the Electronic Registration Information Center (ERIC). The latter is a cooperative program, with 30 states now joined as members, that identifies out-of-date voter records by comparing the voting rolls of the member states to each other and to each state’s motor vehicle records. Using this system, states can identify registrants who moved away and got a driver’s license and/or registered to vote in their new location.

Given these methods of cleansing the voting rolls, there simply is no justification for also using voter inactivity as an independent basis for eliminating registrations. But some states like Ohio and Georgia persist in canceling registrations of voters simply because they have not voted recently and failed to return a mailed notice. There is every reason to be concerned that this practice continues because it has a political skewing effect. Failure to vote regularly correlates with lower socioeconomic status and, at least in some places, with being a member of a racial minority.

Unfortunately, the courts have so far not provided the remedy that is needed. In 2016, voting rights advocates from Demos and the American Civil Liberties Union sued the State of Ohio on behalf of the A. Philip Randolph Institute, claiming that that state’s aggressive purge policy violated the National Voter Registration Act (NVRA). This was logical because the NVRA governs states’ obligations to purge ineligible names from the voting rolls while also providing that a person may not be purged “by reason of the person’s failure to vote.” Two years later, however, the Supreme Court ruled 5–4 in Husted v. A. Philip Randolph Inst. that the NVRA did not bar Ohio from continuing its policy of purging voters solely because they had failed to vote for two years, they failed to return a notice sent to them in the mail, and they failed to vote for four more years. The Court essentially held that, as long as a state mails the notice and waits four more years—two steps required by the NVRA as a protection to voters who are slated for purges—there is no statutory problem, regardless of how little justification the state may have had for identifying candidates for purges in the first place.

Whatever the merits of that statutory ruling, the Constitution ought to provide a remedy here. Under the Fourteenth Amendment, a state-law burden on voting is unconstitutional unless it produces some legitimate benefit that outweighs the burden. This so-called “Anderson-Burdick” balancing test stands as a bulwark against aspects of state election administration policies that do more harm than good. And that is what we have here. In one recent test of a constitutional argument, a federal court in Georgia, in a case brought by Fair Fight Action, denied a preliminary injunction seeking to restore the registration status of 98,000 Georgians who had had their registrations canceled solely because of a period of non-voting and non-return of confirmation notices. But the hope in that case and others is to persuade the courts that this kind of voter purge is outmoded and unjustified.

How do we know that? An assessment of the data supplied by the states themselves shows why purges based only on non-voting and non-return of a notice are guaranteed to dump a high percentage of people off the voting rolls who have not moved anywhere and otherwise remain eligible to vote in some future election. Those data are collected and published annually by the federal Election Assistance Commission. Take Georgia, for example. Even in the two very high-turnout elections of 2016 and 2018, well over a third of the state’s voters did not vote. Clearly, there are many people who have not moved away who choose not to participate in every election. How about the mailed notices that the NVRA requires? Unfortunately, those appear to be routinely ignored by most people. Georgia sent out almost half a million such “confirmation notices” to registered voters in 2017. Less than 10 percent were returned by recipients, about 15 percent were returned “addressee unknown,” and more than 75 percent elicited no response at all. So, the notices provide only limited information and do not seem to be effective as warnings that they are in the pipeline toward a registration purge.

Protecting the right to vote is one of the most important things our courts can do. Even those who rail against “judicial activism” and in favor of democratic decision-making ought to recognize that when the democracy is distorted by the actions of elected officials, only the courts can help. Stopping unnecessary and unjustified voter purges is a prime example.

Paul M. Smith is a professor at Georgetown Law and vice president for Litigation and Strategy at the Campaign Legal Center. He argued the Husted case in the Supreme Court and is one of the counsel for Fair Fight Action in the Georgia case.