Actually Good News About Voting for a Change

Colorado’s simple plan to increase voter registration is already working.

Row of "I voted" stickers.
Ben Sklar/Getty

About the author: David A. Graham is a staff writer at The Atlantic.

Updated at 8:29 p.m. ET on July 7, 2022

In 2020, with the coronavirus pandemic ravaging the country, many states altered their election systems to try to ease voting. Since then, some of those states, especially Republican-led ones, have aggressively reversed course, taking steps to make voting harder.

This sort of bad news has overshadowed one of the more interesting and encouraging changes in the country. Starting in May 2020, Colorado modified its registration system so that anyone who gets a new driver’s license and provides proof of citizenship is now automatically registered to vote. The state then sends people a postcard informing them of their registration and offering them a chance to opt out. That’s a small tweak from the prior system, in which anyone who came to the DMV would be offered the option to register, but a study produced earlier this year by two Stanford University political scientists shows that the new approach has made a significant difference: more than 200,000 new registered voters in the 16 months through September 2021, in a state where about 3.3 million votes were cast in the 2020 presidential election.

That’s the first real empirical indication of how effective this system, known as “back-end automatic voter registration,” can be, as compared with the more common “front end” system, in which people are offered a chance to opt out of registering when they visit the DMV. While democracy advocates and many in the Democratic Party fret over threats to voting, a shift to back-end AVR could be a simple and effective way to lower barriers to voting and expand access.

“We’re very proud of the results,” Jena Griswold, Colorado’s secretary of state, told me. “Registering to vote and voting itself should not be a burden. These are our constitutional rights. State governments should be seamlessly offering potential voters the option to register.”

Registering to vote in many states is annoyingly and unduly complicated. The idea of automatic voter registration dates back to the 1993 National Voter Registration Act, which required that states give people a chance to register when they get a license. But making that automatic was beyond the capacity of state systems at the time. Some advocates, especially those at the Brennan Center for Justice, pushed for states to more aggressively streamline their systems. AVR backers say the approach makes for a more representative electorate, because it tends to result in more registrations among young voters, low-income voters, and nonwhite voters. It also produces higher voter turnout.

Two states enacted AVR systems for the first time in 2015—a front-end system in California, and a back-end one in Oregon. Democrat Kate Brown, then Oregon’s secretary of state, had championed the idea as a bill, which became the first law she signed when she became governor. By the 2020 election, nearly 94 percent of the state’s eligible residents were registered.

“We knew we needed a blueprint for other states,” she told me recently. “You should not be required to register … However, we have voter registration, so why not make it as simple and accessible as possible?”

But in the past seven years, a majority of the 20 states, plus the District of Columbia, that have moved toward automatic voter registration have used a front-end approach. That included Colorado, which adopted front-end registration in 2017. Among AVR proponents, a robust debate has continued over which system is better. The Brennan Center, which is perhaps the most powerful voice for AVR, has strongly backed the front-end approach. Some front-end partisans worry that back-end AVR might end up inadvertently registering noncitizens, thus exposing them to prosecution for voter fraud. Others argue that if people don’t want to register, they shouldn’t have to.

Those who favor back-end systems say that there are safeguards against noncitizen voting. In Colorado, for example, only people who have already provided proof of citizenship are automatically registered. They also say that it’s a minimal cost on individuals who end up being registered against their preference, because they aren’t obliged to vote. Moreover, advocates argue that “nudge”-style systems will produce higher registration numbers, which is in the public interest. Americans famously loathe the DMV, and if they’re eager to get out as fast as possible, they might not take the invitation to register even if they would be happy to be on voter rolls.

“There’s a lot of evidence that defaults matter quite a bit. If the default is you’re going to be opted in, it’s a big pain to return the mailer,” says Justin Grimmer, one of the authors of the paper on Colorado’s new system. “What we’re talking about is a thing that is a relatively low burden on the individual.”

This view is intuitive—of course opt-out systems would produce more registrations—but voter behavior is not always intuitive, and there hasn’t been evidence to back it up, because states have mostly gone from non-AVR systems to either front- or back-end AVR. But Colorado’s decision to switch from front-end to back-end provided a natural experiment. The results suggest a significant increase, and at little price: Griswold’s office told me the shift cost just $86,000 total, most of that borne by the Colorado Department of State.

All of this might sound great, if you’re a Democrat or progressive and assume that this will juice turnout for candidates you support. Like Brown, Griswold is a Democrat. The Brennan Center is a leading proponent of liberal causes. Meanwhile, red states are busy trying to make voting harder, on the premise that greater turnout will aid Democrats. But the evidence suggests that isn’t necessarily the case. In Oregon, AVR increased registration in rural areas, lower-education areas, and poorer areas, which are typically more Republican than the state overall. Beyond that, political realignment means higher registration may help GOP candidates in general.

“There’s this notion that higher turnout favors Democrats and lower turnout favors Republicans,” Grimmer told me. “I think that notion is outdated. If you look at down-ballot races from 2020, a high-turnout election, Republicans did exceedingly well.”

Griswold also argues that having a clean, updated voting roll is actually a positive for election security. The political parties in Colorado have sometimes grumbled about AVR, because it does not automatically affiliate a voter with any party, but she shrugs that off as their problem. “As an elections official, my job is to get people registered, regardless of party,” she said.

That’s a long way from the view some of her counterparts are taking. “We see so many states trying to suppress the vote of low-income people, of people of color,” Griswold said. “I think that’s un-American; I think it’s undemocratic.” Unfortunately, trying to limit who can vote is squarely in line with much of American history, but it is certainly undemocratic. State officials and voters who want to guarantee a more democratic approach would be wise to emulate Colorado’s shift to opt-out automatic registration.


This piece originally mischaracterized front-end automatic-voter-registration systems like the approach recommended by the Brennan Center as opt-in rather than opt-out.