With Republican politicians leading an assault on voting rights, the need for nationwide automatic voter registration (AVR) to safeguard democracy is clearer than ever.
Elections in the U.S. are noted for low rates of voter participation compared to other industrialized countries, and it’s not hard to imagine why: America’s byzantine “self-registration” process for would-be voters has long been an unnecessary barrier to civil participation. While congressional gridlock will likely prevent nationwide AVR from passing this year, states can and must continue to lead the way in the battle for AVR.
Automatic voter registration modernizes the National Voter Registration Act (NVRA) passed in 1993. The NVRA required people to be given an opportunity to register to vote when receiving government services. However, it was an opt-in, not an opt-out system, and many states did not fulfill its requirement. AVR changes the way NVRA is implemented to an opt-out, rather than an opt-in system. Behavior psychology studies show that opt-out dramatically increases participation.
ADVERTISEMENT
Since Oregon successfully adopted AVR in 2014, 23 states and Washington, D.C. have adopted their own AVR systems. Under these systems, eligible voters are added to state rolls through interacting with state offices such as the Department of Motor Vehicles (DMV), and are given a way to opt out if uninterested.
While any AVR system is preferable to the regressive self-registration system, it must still be stressed that not all AVR systems are created equal. Nineteen states operate under front-end AVR systems, where individuals who go to the DMV are given the direct option of whether to decline enrollment in the voter rolls while receiving service.
In contrast, under a back-end registration system, voters are automatically added to the voter rolls after going to the DMV, and are later given the option to opt out after being sent an official document asking them if they want to do so. It is unfortunate that the back-end system is only used by a small minority of states, since it has proven to be a superior system in terms of expanding the electorate, bureaucratic efficiency, and ensuring election security. In particular, activists should look to a recent effort to enact back-end AVR in New Mexico as an example to emulate.
While New Mexico has already established a front-end model of AVR statewide, the definitive evidence that back-end AVR is a more successful model led to a legislative push to adopt the latter this past legislative session through an omnibus voting rights package. New Mexico Secretary of State Maggie Toulouse Oliver, the state’s chief election officer, has herself acknowledged that “Back-end AVR is the nationally-recognized gold standard for voter registration and would improve what’s already a winning AVR policy here in New Mexico.”
Indeed, Toulouse Oliver’s argument is particularly convincing given the successful case of Colorado’s switch from a front-end system initially adopted in February 2017 to a back-end system over three years later. Beginning in May 2020, visitors at Colorado DMV offices who proved their citizenship were no longer asked to opt out of registering to vote during the transaction, and were provided the opportunity to decline registration through a post-transaction notice.
The results of this transition were an unequivocal success: In a historic study, Dr. Justin Grimmer and Dr. Jonathan Rodden of Stanford University found that Colorado’s transition from operating under a front-end system to a back-end system almost doubled the voter registration rate at DMV offices. As noted by Colorado Secretary of State Jena Griswold, “This research on Colorado’s Automatic Voter Registration shows how states can make dramatic improvements by streamlining the voter registration process at their motor vehicle agency.”
Rodden and Grimmer’s research helped identify a long-suspected flaw present in front-end systems: When would-be voters are “given a simple option to decline and procrastinate, a majority of prospective registrants at the DMV will take it.”
Whether justified or not, Americans are known to dread visits to local DMV offices, and being forced to endure an extra minute and a half during a DMV transaction (in the case of Colorado) will lead people to reflexively decline to register. "Not only does the back end system register more people to vote, it also decreases the time individuals spend at the DMV," Grimmer said in an email. While front-end advocates insist that having to directly choose during an agency transaction whether or not they intend to register will make individuals more likely to vote, there is no reason to believe this to be the case.
Indeed, while sound governance stands to increase public confidence in the state’s ability to deliver, having to navigate a frustrating bureaucracy tends to do the opposite.
In her book Fragmented Democracy: Medicaid, Federalism, and Unequal Politics, Jamila Michener highlights how the broken administration of Medicaid services has a negative impact on political participation: "There is a correlation between Medicaid enrollment and political participation and, as per my qualitative indications, that it is negative... Medicaid enrollment corresponds to a five-percentage-point decrease in the probability of voting, a five-percentage-point decrease in the probability of registering, and a six-percentage-point decrease in the probability of participation."
One of the main reasons that back-end AVR systems are more successful than front-end AVR systems is that they are simpler and, accordingly, easier to administer. Instead of a complex customer-facing interface for a front-end system, back-end systems rely on data already collected and verified by state agencies.
Take the launch of California’s front-end AVR program in 2018, widely considered a technical disaster, as an example: At the time, California’s DMV was equipped with a computer system mocked as a “40-year-old dinosaur” and staffed with DMV individuals untrained in assisting would-be registrants. Described by FiveThirtyEight as a “cautionary tale of how AVR can create serious problems when government agencies are incompetent or unprepared”, the front-end system’s chaotic rollout provided fodder to foes of automatic voter registration.
The voter self-registration system in the United States has deep roots in institutional racism, and continues to entrench white supremacy to this very day. As noted by the Southern Coalition for Social Justice, AVR is a crucial tool for enfranchising “communities of color, young people, the elderly, voters with disabilities, voters with limited English proficiency, and low-income communities.” Sound implementation of back-end AVR is one of the best tools for bridging the turnout gap that comes at the expense of voters from disadvantaged backgrounds.
In particular, it is crucial that AVR efforts center the cause of immigration justice. American immigration law is broken, and even minor mistakes can lead to longtime U.S. residents being precluded from ever attaining citizenship. As one immigration attorney noted, wrongly claiming U.S. citizenship on a voter registration form is a “kiss of death” for immigrants, “a permanent black mark that prevents a noncitizen from ever gaining status”.
As such, it is important to note the obvious risks to immigrant communities posed by front-end AVR systems. Under this arrangement, it is entirely possible that a non-citizen resident who is not fluent in English will accidentally respond in the affirmative if asked on a DMV screen if they are a citizen. This split-second decision could completely change the trajectory of a person’s life through no fault of their own. Employees at DMV offices and other source agencies do not necessarily have specific expertise in the realm of elections, and wrongful advice to an immigrant during a transaction could lead to this issue arising.
Under a back-end system, however, only people who provide proof of U.S. citizenship are channeled into the streamlined registration process, and people who provide proof of non-citizenship are affirmatively filtered out of any registration opportunity, protecting them from a mistake. Moreover, even if an error hypothetically occurs due to system error, the risk of such an error would fall on state officials, not an individual immigrant who has never made an erroneous claim of citizenship.
The chaotic launch of California’s front-end system reportedly led to the registration of 1,500 ineligible individuals, including non-citizens, being accidentally registered to vote due to a data error. While the error was thankfully resolved, a similar situation could arise in a front-end system that has real-world devastating consequences for an immigrant erroneously registered.
Back-end AVR is also a crucial tool towards increasing political engagement among formerly incarcerated persons, who continue to be targets of voter disenfranchisement. At least 13 states extended voting rights to individuals with felony records from 2016 to 2020, yet analysis from the Marshall Project found that the vast majority of those newly eligible did not register to vote. While many of these individuals surely chose not to vote for reasons shared with other non-voters, it is likely that confusion surrounding one’s eligibility deterred many from registering.
A 2018 statewide survey conducted by the Colorado Criminal Justice Reform Coalition revealed that misconceptions regarding eligibility to vote for those with criminal records were widespread. Under a front-end AVR system, it is easy to imagine cases where a formerly incarcerated person, unaware that they are eligible to vote, will choose to decline to register so as not to get in legal trouble. Under a back-end system, this will not be an issue, as eligible voters will be automatically added to official rolls after visiting a DMV.
The successful implementation of back-end AVR has been shown to increase turnout among young voters, an indispensable segment of the Democratic coalition.
After Oregon implemented back-end AVR in 2015, the state saw a sizable 4.1 percent increase in total turnout in 2016 compared to 2012, itself an impressive feat. In the 2016 election, 37 percent of voters registered through AVR were under the age of 30, while voters in that age demographic only made up 13 percent of voters registered not through AVR.
In their study of Colorado, Grimmer and Rodden note that “the adoption of the back-end system led to a substantial increase in pre-registrations among 16 and 17-year olds.”
It must be remembered that not all Americans drive, which means that DMV offices should not be the only agency where AVR transactions take place. In recognition of this problem, Colorado’s legislature has also directed the state’s Medicaid office to register eligible beneficiaries, which would help expand registration for many low-income individuals.
Medicaid’s expansive reach (over 83 million Americans are insured via Medicaid), robust eligibility verification mechanisms, and deep experience in data management make the agency especially well-suited to register voters. The fact that current research indicates that the National Voter Registration Act has failed to increase registration rates among Medicaid patients shows that it is time for states to take bolder action.
Back-end AVR remains the single best option for state policy-makers to safeguard the right to vote against Republican assaults. While New Mexico’s aforementioned voting rights push hit a recent legislative roadblock, officials across the country should account for the fact that state leaders know that changing from a front-end system to a back-end system would be a major upgrade. Automatic voter registration should be a seamless process, and automatically registering clearly eligible voters during agency transactions while giving them a chance to opt out at a later date is the best way to expand the electorate while respecting the wishes of voluntary non-voters.
