With voting, there's no place like home

By Associate Professor Justin Levitt

In a recent New York Times "Campaign Stops" piece, I mentioned concerns about citizens undermining the integrity of the election by attempting to impose and enforce the "law of their gut" rather than the law on the books.

One prominent example concerns the concept of domicile. Domicile shows up in different ways in the election process, but the most common misunderstanding has to do with where people are eligible to vote. Recent reports reveal that voters have been challenged in several states based on claims that they aren't eligible to vote where they're registered. These challenges often reflect a common gut instinct -- but more seldom reflect the law.

In several states, organizations "empowered" by a national organization called True the Vote have announced research finding registrations subject to challenge based on their addresses. This isn't the first time that people have been challenged based on unwarranted assumptions about their addresses -- voter caging schemes, for example, are a version of this.

The True the Vote approach apparently starts by combing large databases, flagging (for example) addresses where more than 6 people are registered. Other flags may include people registered at colleges or allegedly vacant lots.

One problem with quasi-automated challenges like this is the quality of the data: "vacant" lots are sometimes not actually vacant.

But even if the data are accurate, they don't necessarily show ineligibility. Let's start with the street addresses listing six or more individuals. Most households are smaller than this, yes. But 1.7 million households -- disproportionately African-American and even more disproportionately Hispanic -- have seven members or more. And that doesn't include more than 36 million people living in apartment buildings, with 5 or more units at the same street address, or the 8 million people living in "group quarters" like dorms or military barracks. Some Americans live with their 2.5 children in single-family homes; many Americans don't.

That's just the start of the disconnect. It's possible for more people to be validly registered at a street address than actually live there at any given moment. That is, it may be entirely lawful for seven people to be registered to vote at an address that sleeps four.

If that seems crazy, that's because the standard for voter registration isn't always intuitive. Like most election laws, the legal standard differs from state to state. In some states, you're properly registered to vote wherever you consider "home" at the moment. In other states, you should be registered at the last place you lived that you didn't consider temporary. Other states' rules are different still.

The proper legal standard usually turns on intent, which makes it particularly tricky to assess proper registration with software. Let's say that I move to 123 Main St., intending to settle in at my new address indefinitely. I may leave someday, but there's no date I have in mind, and until then, I've left my old address behind. In many states, 123 Main is now my domicile: where I can legally register to vote.

Now let's say that I leave for some reason. It may be the same day that I move in, without unpacking a single box. I may be called up for military duty, or called away for a business trip, or simply travelling to renaissance festivals across the country. As long as 123 Main St. was a valid place for me to register when I arrived, if I don't establish domicile somewhere else, 123 Main St. would still be the legal place for me to vote.

That is, 123 Main would be my lawful registration address even if I were only present there for a few hours. 123 Main would be lawful whether it was a single-family house, an apartment, a college dorm, or a homeless shelter. It would be lawful even if someone else were now living in my former room at 123 Main -- yielding more voters registered than people physically present. And it would be lawful even if the building had been knocked down while I've been away -- leaving a vacant lot.

States' legal definitions of domicile can lead to situations that may seem unusual to a casual onlooker. It's why college students may be able to claim valid domicile for voting purposes either at their campus address or a prior address, depending on the states in question and the students' intent when they get to school. It's why members of Congress may be legally registered to vote "back home," even when "home" was sold decades ago. It's why it's not a stretch to believe that Mitt Romney was lawfully registered to vote for a year at his son's house, even with multiple other permanent homes in his name.

It's even one of the reasons why there may lawfully be more registered voters in a county than the number of eligible citizens counted there by the Census. Some loudly decry such "over-registered" counties as breeding grounds for fraud. As a gut measure, it has appeal. But -- for example -- military personnel who are overseas on Census Day are registered at their domicile, but not counted there by the Census. The gut measure ignores valid reasons why one list might not match up with the other, including those serving overseas. The law serves us better.

To be sure, people are sometimes listed on voter rolls where they do not belong, often as the aftermath of a real residential move. Portable registration systems that actually follow people when they move, rather than relying on new registrations on the front end and purges on the back end, would go a long way toward solving the problem. At the moment, regular list maintenance, with notice to the voter and safeguards for error, does a moderately good job at cleaning the rolls over time, without unduly jeopardizing the rights of legitimate voters who have quirky residential situations.

By contrast, last-minute challenges founded on data screens, without personal knowledge of a voter's intent, only serve to capture situations that seem unusual -- without any realistic ability to know whether that unusual is also unlawful. In a diverse country, we should all be wary of efforts that sloppily conflate the two.

Justin Levitt is an associate professor of law at Loyola Law School, Los Angeles, where he teaches Law of the Political Process. He maintains the website and blog "All About Redistricting" at redistricting.lls.edu.