Mandatory Voter Registration: How Universal Registration Threatens Electoral Integrity


Mandatory Voter Registration: How Universal Registration Threatens Electoral Integrity

March 27, 2013 29 min read Download Report
Election Law Reform Initiative Manager, Senior Legal Fellow
Hans von Spakovsky is an authority on a wide range of issues—including civil rights, civil justice, the First Amendment, immigration.
It has been said that for every complex problem there is a solution that is clear, simple, and wrong. Washington soon may seek a complex solution—preemption of states’ responsibility; federal micromanagement of elections; eventual coercion of lackadaisical citizens—to the nonproblem of people choosing not to vote.

—George F. Will [1]

Mandatory voter registration (MVR), previously termed “universal” registration, could significantly damage the integrity of America’s voter registration system. The “voter registration modernization”[2] concept of automatically registering individuals through information contained in various existing government databases would throw the current system into chaos.

Specifically, voter registration modernization could result in the registration of large numbers of ineligible voters as well as multiple or duplicate registrations of the same individuals. When combined with the accompanying proposal that states allow any individuals who are not automatically registered to register and vote on Election Day, MVR presents a sure formula for registration and voter fraud that could damage the integrity of elections.

Automatically registering individuals to vote without their permission would also violate their basic right to choose whether they wish to participate in the U.S. political process. Indeed, this new scheme threatens one of American’s most cherished liberties: the freedom to be left alone by the government.

A “Solution” in Search of a Problem

Lack of registration is not the reason people do not vote. Ideological organizations such as FairVote and the Brennan Center for Justice are proposing that states automatically register all individuals to vote using existing government databases. Such proposals are based on the false premise that large numbers of Americans do not vote “for no other reason than they are not registered to vote.”[3] Yet after every federal election, the U.S. Census Bureau publishes reports on the levels of registration and voting, including surveys of individuals who do not vote, that disprove the claims that the major reason individuals do not vote is a lack of registration opportunities.[4]

For example, of the 146 million people who the Census Bureau reported were registered to vote in 2008, 15 million (10 percent) did not vote. Of those who did not vote, only 6 percent cited registration problems as the reason for not participating. Rather, the vast majority of these registered but nonvoters said they did not vote for reasons ranging from forgetting to vote to not liking the candidates or the campaign issues or simply not being interested.

With regard to those individuals who are not registered to vote, the Census Bureau’s 2008 report demonstrates that the major reason individuals failed to register was that they were not “interested in the election/not involved in politics.” That represented 46 percent of the individuals in the Census Bureau’s survey. Another 35 percent of individuals did not register for a variety of reasons such as not being eligible to vote, thinking their vote would not make a difference, not meeting residency requirements, or difficulty with English.

Thinking that their vote would not make a difference is quite true in some cases even if the rest of us enjoy and encourage civic participation for its own sake: “[E]ven a smart and hardworking person can rationally decide not to pay much attention to politics. No matter how well-informed a person is, his or her vote has only a tiny chance of affecting the outcome of an election.”[5]

Only 4 percent of individuals reported not registering to vote because they did “not know where or how to register.” This may be true, or it could be a convenient excuse for many who are too embarrassed to tell a pollster the truth given how easy it is to register by mail, at the many locations where registration is available such as libraries and numerous government offices and agencies, or (in many states) by using the Internet.

The Census Bureau’s 2010 report indicates similar results.[6] Only 3.3 percent of individuals reported not voting because of supposed registration difficulties. Given the tendency of many people not to take responsibility for their own failings or perceived failings, the actual number of people who did not vote because of registration difficulties may be even smaller. The overwhelming majority of those who did not vote said they were not interested (16 percent); were too busy (27 percent); forgot to vote (8 percent); did not like the candidates or the campaign issues (9 percent); or had various other reasons.[7]

Registration problems do not disproportionately affect minorities and low-income citizens. Among the tiny percentage of voters who said they did not vote because of “registration problems,” there was also almost no racial differential. For instance, the percentage of whites who claimed they did not vote because of a registration problem was 3.2 percent, compared to 3.3 percent of blacks and only 2.8 percent of Hispanics.

There is little evidence to support the oft-repeated assertion that “voter-initiated registration” has a “disproportionate impact on low-income citizens and those who are less educated.”[8] In fact, the Census surveys show otherwise. For example, in 2008, the percentage of registered voters who did not vote because of “registration problems” was 6 percent; among voters with a bachelor’s degree or more, the percentage was 7.4 percent compared to only 3.2 percent for those with an educational attainment of “less than high school graduate.” Furthermore, those attaining “high school graduate or GED” had a rate of 5.8 percent.

The Census survey, in other words, actually demonstrated that less-educated voters had fewer registration problems. The 2010 survey reported similar results for those who did not vote due to registration problems: less than high school, 2.5 percent; high school graduate, 2.6 percent; bachelor’s degree or more, 4.3 percent.

With regard to income, the 2010 Census survey demonstrated no discernible “disproportionate impact.” For example, the percentage of voters with a family income of $100,000 to $149,000 who did not vote because of purported registration problems was 3.5 percent; the percentage of those with an income of $15,000 to $19,999 who claimed registration problems was only 1.9 percent; and the percentage of voters with an income of $10,000 to $14,999 who supposedly had registration problems was 2.8 percent, just slightly more than the 2.6 percent reported by individuals making more than $150,000.

Thus, according to the federal government’s own surveys, the claim that “the single greatest cause of voting problems in the United States”[9] is the voter registration system is false. The greatest causes of individuals not registering and not voting are their lack of interest in politics and candidates and other reasons that have nothing whatsoever to do with registration or lack of registration.

Experience with the National Voter Registration Act of 1993 shows that voter registration is not a barrier to voting. The push to pass the National Voter Registration Act (NVRA) of 1993 was based on the same, similarly flawed premise: that voter registration is a barrier to voting. Before its implementation, “many researchers were optimistic about NVRA’s projected impact on voter turnout”; but while the act “did lead to millions of new registered voters,” it apparently made “no significant change in voter turnout.”[10] In other words, the NVRA only led to an increase in the number of registered voters who do not vote.

Other researchers point out that overall registration levels have not increased substantially since passage of the NVRA. The Census Bureau’s 2008 report shows that the reported voter registration rate in 1996—three years after the NVRA became law—was 70.9 percent. The reported registration rate in 2008 was 71 percent—an increase of only one-tenth of 1 percent after the NVRA had been in effect for 15 years.[11] In 2008, the highest level of turnout according to the Census Bureau was among non-Hispanic Whites (66 percent) and blacks (65 percent); turnout among Asians was 48 percent, and turnout among Hispanics was 50 percent.[12]

The experience with the NVRA shows the basic flaw in the underlying assumptions that led to its passage: that registration “barriers” were somehow the reason for the claimed decline in voter turnout. Research shows “that the motivation to vote is especially internal: people register because they plan to vote. Therefore people who are registered are very likely to vote. However, people who have no interest in voting do not register to vote.”[13]

One detailed study of nonvoters concluded that it is “[a]nother misconception about nonvoters…that they would vote if only the [registration] process was easier.”[14] The study concluded that the reason people do not vote is because for many of them, “voting is neither duty nor ritual.” They are not interested in politics, or are cynical about its outcomes, or do not believe their votes will make a difference (public choice scholarship confirms that such cynicism is often well-founded).

In other words, there are “competing strains of alienation and complacency” among the ranks of nonvoters.[15] Consequently, electoral reforms—“such as easing voter registration through motor-voter legislation, same-day registration, or uncoupling registration from jury duty—have had, at best, a negligible net effect on voter participation.”[16] Those with greater faith in government’s efficiency and efficacy may be more optimistic about its ability to have a positive impact on American’s lives. In the long run, however, that faith may do more to undermine civic virtue than a healthy cynicism about government bureaucracy.

MVR’s Numerous Practical Problems

Various recommendations made for a federally imposed, national mandate would require states and local governments to:

  • Use existing state and federal government databases to automatically (and permanently) register all citizens to vote.
  • Create an overriding policy to ensure that voters left off the rolls can register and vote on Election Day.
  • Require U.S. citizens to register to vote when completing taxes or actively opt out of the process.
  • Tie Post Office change-of-address forms to the voter registration database.
  • Require state or local governments to send every residence a notice of those registered at that location; residents could then make changes as needed and return the updated form.
  • Provide every U.S. citizen upon birth or naturalization a voter registration number similar to a Social Security number, to be used in all elections and activated when a voter turns 18.[17]

Some of the groundwork for these proposals and federalization of the voter registration process was laid at a Senate Rules Committee hearing by Senator Charles E. Schumer (D–NY) on March 11, 2009.[18] Senator Schumer advocated overhauling America’s voter registration system in favor of the “Voter Registration Modernization” proposal from the Brennan Center.[19] This proposal shifts the responsibility of voter registration from the individual to the government, leading to the erosion of distinctions between state and federal responsibilities in election management and the responsibility of individuals to take the steps required to participate in the election process.

The push for mandatory voter registration has accelerated recently. In December 2012, a month after the November election, the leaders of more than three dozen liberal advocacy groups met in Washington for an off-the-record meeting (though covered by Mother Jones in some detail) to plan strategy on election-related issues. One of the top three goals was mandating “voter registration modernization” and same-day voter registration; at the same time, one of the other goals agreed on was to oppose any efforts to improve election integrity through voter identification and proof-of-citizenship requirements.[20]

At a speech in Boston on December 11, 2012, Attorney General Eric Holder voiced the Obama Administration’s support for automatic registration.[21] The head of the Justice Department’s Civil Rights Division, Thomas Perez, said on November 16, 2012, that “all eligible citizens can and should be automatically registered to vote” based on compiling “from databases that already exist.” Perez also claimed that one of the “biggest barriers to voting in the country today is our antiquated registration system.”[22] The Brennan Center’s 2008 proposal was relaunched in January 2013 when the Brennan Center issued another report on “voter registration modernization,” and on January 23, 2013, Representative John Lewis (D–GA) introduced the Voter Empowerment Act (VEA).[23]

These mandates involve numerous practical difficulties. The most common proposal—for states to use existing government databases “to build”[24] their voter rolls—presents several immediate problems.

First, many government databases may lack a signature, which is required for voter registration and essential for verifying both petitions for candidates and ballot initiatives, as well as requests for absentee ballots and voted absentee ballots that are received by election officials.

Second, using government databases such as “motor vehicle departments, income tax authorities, and social service agencies,” as recommended by the Brennan Center, would fail to differentiate citizens from non-citizens. All states, for example, provide driver’s licenses to aliens who are legally in the United States, and several states provide driver’s licenses to illegal aliens. Many individuals who reside in the United States but are not citizens also file tax returns, which would allow individuals who filed with “income tax authorities” the ability to register to vote. It would also lead to duplicate and multiple registrations of individuals listed on different government databases, such as individuals who own property or pay taxes in more than one state.

Third, as an enormous unfunded mandate on the states, these proposals would prove costly: a diversion of limited government resources for little to no appreciable increase in voter participation rates.

In addition to DMV, social service, and income tax agencies, the VEA would require automatic registration of individuals from state agencies that provide benefits under Title III of the Social Security Act, that maintain records on students enrolled at secondary schools, that are responsible for administering criminal convictions, or that determine mental competence. Additionally, automatic registration would be required from the federal offices of the U.S. Immigration and Customs Enforcement Bureau, the Social Security Administration, the Federal Bureau of Prisons, the U.S. Probation Service, the Department of Veterans Affairs, the Defense Manpower Data Center of the Department of Defense, and the Indian Health Services and Centers for Medicare and Medicaid Services of the Department of Health and Human Services.

No transaction with any such agency could be completed “until the individual has indicated whether he or she wishes to register to vote.” Every time an individual applied for services or assistance, and “with each recertification, renewal, or change of address relating to such services or assistance,” the agency would have to ask the individual about registering to vote and could not provide any requested service or assistance until the registration issue had been addressed.[25]

Proponents of mandatory registration from government databases oppose even limited use of such databases to maintain accurate voter rolls. It is rather ironic that many of the organizations pushing for automatic registration of individuals based on government databases oppose states’ attempts to verify the citizenship, identity, and accuracy of the information provided by individuals registering to vote by comparing them to other government databases.[26] In 2007, for example, the Brennan Center, along with the National Association for the Advancement of Colored People (NAACP) and the Advancement Project, sued Florida for running database comparisons on registered voters’ information with “the state driver’s license database or the Social Security Administration’s database.”[27] In a related press release, the Brennan Center complained about “common database errors” and opposed matching as “an error-laden practice.”[28 ]

Furthermore, in 2006, the Brennan Center and other so-called civil rights organizations sued the state of Washington, claiming that attempting to match voter registration information with other government databases violated the Voting Rights Act and the U.S. Constitution and would disenfranchise voters.[29] In fact, the Brennan Center issued a report in 2006 complaining about the supposedly “wide variety of common database matching errors” caused by “data entry” mistakes.[30] Yet the Center now wants to use those same supposedly inaccurate databases to register voters automatically.

As Colorado Secretary of State Scott Gessler pointed out during a January 2013 discussion at The Heritage Foundation, there is no question that there are inaccuracies in state voter registration rolls. However, federal databases are also riddled with errors that may eclipse inconsistencies at the state level. It is important to note that state registration lists are transparent—such lists are available to candidates, political parties, and the public—but federal databases lack such transparency, and election officials and the public therefore cannot verify the accuracy of such lists.

Gessler has witnessed many inaccuracies in Social Security Administration information as well as the National Change of Address (NCOA) database used by the U.S. Postal Service. For example, the NCOA reports a move only if an individual informs the Postal Service of a move. Errors can also occur if the NCOA database classifies everyone at a particular address as having moved when only one person in the household has moved. Gessler believes these federal databases are valuable when they are being used by states to check the information contained in state voter registration lists, since any discrepancy can be researched and corrections made, but to use federal information to automatically register individuals to vote would be to court disaster.

The Brennan Center says that many of these government databases “already include all the information necessary to determine voter eligibility, and those that do not can easily be modified to include that information.”[31] However, as just one example, many of these databases do not contain citizenship information—a basic requirement for eligibility to vote. Organizations such as the Brennan Center have opposed states requiring proof of citizenship from registrants that would provide “that information.”

Even worse, in 2012, a number of civil rights organizations and the Department of Justice sued Florida in an unsuccessful attempt to stop the state’s verification of citizenship status through database comparisons.[32] Florida had to sue the federal government to get access to Department of Homeland Security (DHS) immigration databases to which it is entitled under federal immigration law to get citizenship information. DHS has also fought states through administrative measures, such as using bureaucratic red tape to prevent states from accessing its own databases—something Secretary Gessler experienced firsthand in Colorado.

As the trail of litigation makes clear, these organizations would fight any implementation of an automatic registration program that would allow states first to compare the information in one database with the information in other state and federal databases to ensure that the information is accurate and that only eligible individuals are being registered.

MVR makes maintenance of existing registration lists even more difficult. The VEA introduced by Representative Lewis would make it difficult—even more so than it already is—for states to maintain accurate voter registration lists. For example, the legislation would amend the NVRA to prevent states from requiring further documentation of new registrants—documentation, such as proof of citizenship, that might be needed to determine eligibility. Section 104 of the bill requires states to register anyone who has provided the state with a “valid voter registration form” that has been “completed” and “attested” by the applicant. The bill also prohibits the “transfer” of information from “the computerized Statewide voter registration list to any source agency.”[33] Election officials would not even be allowed to retain the “identity of the specific source agency through which an individual consented to register to vote” after the individual is added to the statewide voter registration list.[34]

Consequently, if election officials later determined that registration information was inaccurate or even fraudulent, they would be unable to notify whatever state or federal agency provided them with information on that registrant, making it impossible for the source agencies to investigate possible fraud in the state and federal programs they are responsible for administering. Lewis’s bill would even give noncitizens a get-out-of-jail-free card: It provides that any ineligible individual who becomes registered to vote “shall not be subject to any penalty” for registering “including the imposition of a fine or term of imprisonment, adverse treatment in any immigration or naturalization proceeding, or the denial of any status under immigration laws.”[35] In fact, government officials would be prohibited from using “the information received by” election officials “to attempt to determine the citizenship status of any individual for immigration enforcement.”[36 ]

The Lewis bill also prohibits comparison of voter registration information “with any existing commercial list or database” at the risk of imprisonment for not more than one year and subject to fines.[37] Many commercial databases are more accurate than government databases. There is no reason for such a prohibition—let alone such criminal penalties—other than to remove a valuable tool that could otherwise be used by state officials to deter fraud.

Supporters of a federal mandate for automatic and same-day registration rarely, if ever, mention that Canada has had such a system in place since 1997. This registration system is administered by Elections Canada, which is responsible for conducting all federal elections and referenda. The United States, for a number of good reasons, has no such equivalent federal agency, but one is particularly relevant to the current registration debate: America’s system of dual sovereignty is constitutionally guaranteed, and elections traditionally have been administered by the states. Canadians are automatically registered from a host of government databases similar to those proposed in the VEA, including the Canada Revenue Agency, Citizenship and Immigration Canada, National Defense, provincial and territorial driver’s license and vital statistics agencies, and provincial electoral agencies.[38] (Canadians can also still register and vote on Election Day.)

Yet Canada’s automatic registration system has had no effect in increasing turnout. Even before the implementation of Canada’s new system in 1997, Canadians voted in larger numbers than Americans, but Canada has still seen a steady decline in turnout since the 1970s.[39]

The reasons that Canadian voters who have been automatically registered by the government give for not voting are similar to justifications given by U.S. voters: 28 percent were not interested; 23 percent were too busy; and the rest said “they were out of town, ill or didn’t like any of the candidates.”[40] Automatic voter registration is no panacea for declining turnout or the unwillingness of individuals to participate in the voting process. Thus, it seems clear that Canada’s approach would cause considerable mischief in America’s state-administered election system while providing no benefit in terms of voter turnout.

MVR raises serious privacy concerns. Requiring individuals who would not register on their own to “‘opt-out’ from registration” if they want “to remain unregistered for whatever reason”[41] interferes with the basic right of individuals to decide whether—and to what extent—to participate in the political and democratic process. While society might hope that all citizens will vote, each and every American has the liberty not to do so for whatever reason. Americans who choose not to vote should not have to act every time they make a transaction with a government agency to avoid registration or to remove themselves from a government list that they had no interest in joining in the first place, particularly if it involves investigation of their citizenship, felon status, and other factors that are important to eligibility.

Even if individuals can ask to be removed from the registration list after the database information has been transferred to election officials, such automatic registration raises serious privacy concerns. Voter registration lists are public documents that are (and should be) accessible to journalists, candidates, political parties, and individual citizens. In fact, this transparency is an important component of our election process since these lists are often bought by candidates and political parties for the purposes of identifying voters for political campaigns and organizing get-out-the-vote programs for Election Day.

In contrast, not only are state governments obligated to keep the information in many types of other databases maintained by government agencies private, but information on individuals such as police officers, government officials, or victims of domestic violence must be kept confidential. Automatic voter registration could reveal information such as residential addresses, thereby violating the privacy of individuals who have registered for various other types of government benefits. The VEA does require that such information be kept confidential, but that may be very difficult for election officials to do when they are receiving large amounts of information on hundreds of thousands of individuals from other government databases. The source agencies, which may otherwise be required by law to keep all of their client information confidential, may not be aware that certain clients are police officers or victims of domestic violence—individuals with specific privacy requirements.

A Slippery Slope: Permanent Registration

The Brennan Center and others are also proposing that the federal government require states to institute statewide permanent registration. This requirement would mandate that “once a voter is on the rolls, she would be permanently registered within the state and able to vote without re-registering even if she moved within the state or changed her name.”[42]

Already, the National Voter Registration Act has curtailed states’ ability to clean up bloated voter registration rolls by removing ineligible voters who have moved or died. Making registrations permanent would exacerbate this problem. In fact, many states became so fearful of lawsuits by the Justice Department to enforce these NVRA restrictions that they simply stopped maintaining the integrity of their voter registration rolls.

Citizens have a responsibility to inform state election officials when they change their residence or become ineligible to vote for other reasons, such as being convicted of a felony. Notifying election officials of a change of address within a state is especially important because election officials estimate the number of ballots needed at a polling place based on the number of registered voters and past turnout. Allowing individuals who are registered elsewhere in a state but who failed to notify election officials of their move to vote at a new precinct would undermine election officials’ ability both to estimate how many ballots are needed and to ensure a smooth voting experience without long lines. Indeed, underestimating the number of ballots needed or the number of voters expected at a given precinct makes it more likely that some voters will be disenfranchised.

Furthermore, the proposal on permanency would require government agencies like state Departments of Motor Vehicles, the Social Security Administration, or the Post Office to provide updated address information to election officials in order to change the registration addresses of registered voters. Again, such a proposal smacks of hypocrisy: The U.S. Postal Service’s NCOA is supposedly so inaccurate that liberal civil rights organizations have objected to its use by private parties trying to investigate the validity of voter registrations.

“Vote Caging.” These groups even have coined a term—“vote caging”—to describe this practice. Specifically, they claim that private parties’ use of the U.S. Postal Service’s practice of returning non-forwardable mail to challenge the eligibility of voters constitutes voter suppression even if its records show that the individual no longer resides at the registered address.[43] Indeed, a number of bills have been sponsored in Congress that would make reliance on the U.S. Postal Service’s mail service in this manner a federal offense. Not surprisingly, Section 301 of Representative Lewis’s VEA would prohibit such “vote caging.” If the NCOA database is so inaccurate, why are some suggesting that it be used to pad the voter rolls?

The real problem with such a system is that without a unique identifier, it would be very difficult to match many of these records.[44] The only such unique identifier is a Social Security number. Only a handful of states require that an individual registering to vote provide a Social Security number, and these states, such as Virginia, are allowed to do so only because they were grandfathered into the federal Privacy Act of 1974, which restricts the use of Social Security numbers. Any states that did not require a Social Security number to register when that act was passed cannot implement such a requirement today.[45] When the Help America Vote Act of 2002 was being debated in Congress, a proposal to allow all states to require a full Social Security number from new voter registrants was defeated.

The proposal to provide every U.S. citizen upon birth or naturalization a voter registration number similar to a Social Security number, to be used in all elections and activated when a voter turns 18, would require the creation of a new federal bureaucracy. A more logical approach would be simply to amend federal law to allow all states to require that any individual registering to vote must provide his or her Social Security number. However, in the current political climate, such reform has little chance. Furthermore, political concerns aside, the use of Social Security numbers for voter registration raises valid privacy issues.

To the extent that state voter registration lists can be linked to state DMV records and other state databases, states should—and often do—conduct regular database matching to update registration information as individuals move, die, or become ineligible. But due to the inherent inaccuracies in all such databases, as well as the inability to keep up with all changes in the status of individual voters, states should not be prohibited from removing voters who do not vote in a certain number of federal elections—after they are sent notice of the impending removal. That failure to vote is one indication that a voter has moved or otherwise become ineligible without notifying election officials.

Election-Day or Same-Day Registration

Election-Day registration is highly vulnerable to organized election fraud. The proposal for a federally mandated “fail-safe” that would allow anyone to register and vote on Election Day raises constitutional concerns and is poor public policy.[46] Indeed, such policy is a prescription for fraud,

Allowing a voter to both register and vote on Election Day makes it nearly impossible to prevent duplicate votes in different areas or to verify the accuracy of any information provided by a voter. Election officials are unable to check the authenticity of a registration or the eligibility and qualifications of a registrant by comparing the registration information to other state and federal databases that provide information not just on identity, but also on citizenship status and whether the individual in question is a felon whose voting rights have been suspended. Since Election Day registrants cast a regular ballot, even if election officials determine that the registration was invalid after the election, they have no means of discounting the ballot.

For those states entering into cooperative agreements to compare their voter registration lists to identify individuals registered in more than one state, same-day registration would also eliminate that safeguard. In fact, many of the same organizations that are proposing this type of “fail-safe” have vigorously fought Wisconsin’s effort to begin providing some verification of Election Day registrants by requiring such individuals to show a photo ID. After a comprehensive investigation of voter fraud in the 2004 election, the Milwaukee Police Department concluded that the “one thing that could eliminate a large percentage of fraud or the appearance of fraudulent voting in any given Election is the elimination of the On-Site or Same-Day voter registration system.”[47]

In 1986, voters in Oregon got rid of same-day registration after the Rajneeshee cult tried to take over a local county by not only engineering a bioterrorist attack using salmonella to sicken hundreds of residents (and potential voters), but also planning to bring in large numbers of nonresidents (many of them homeless) on Election Day to flood the polls with ineligible voters. As Kansas Secretary of State Kris Kobach said at The Heritage Foundation in January 2013, double voting becomes almost impossible to stop with same-day registration. Voters can just make up names and false addresses and go from polling place to polling place to vote. Kobach was not aware of any state where the registration system is so automated that the temporary poll workers who staff precincts on Election Day could check the identities and residential addresses of instant voters against other state databases. Election Day registration invites fraud.

Election-Day registration is not likely to increase voter participation or turnout. Most important, however, is that what some call “convenience voting,” which includes “mail voting, no excuse absentee voting, early voting and even election-day registration,” may actually hurt turnout.[48] The general election voter turnout in 2008 was the highest in a presidential election since 1960. However, an American University study reported that of “the 12 states which had turnout declines in 2008 as compared to 2004, 10 had some form of convenience voting. Of the 13 states which had the greatest increases in turnout, seven had none of the forms of convenience voting.”[49] In fact, four of the eight states with Election Day registration reported lower turnout in 2008, when turnout generally went up in the rest of the country, than they had reported in 2004. The state with the largest decrease in turnout in 2008 was Maine (minus 3.6 percentage points), which also has Election Day registration.

Similarly, a study by the Maine Heritage Policy Center found that Election Day registration in Maine had “had no recognizable impact on voter turnout” since its implementation in 1973. In fact, the three election years in which Maine had its “lowest turnout years since 1960 occurred after EDR was implemented.”[50] Nationwide, turnout in the 2012 election was generally down from 2008, dropping a little over 5 percentage points, yet the turnout in Maine went down over 8 percentage points.[51]

Curtis Gans of the Center for the Study of the American Electorate has concluded that states that adopt “convenience voting” reforms “have a worse performance in the aggregate than those which do not.” The only temporary exception is for Election Day registration, which apparently helps turnout only “in its initial application and for a few elections thereafter.” In fact, in election years where turnout generally increases, “the increase in states with convenience voting” is smaller than the increase in those states that have not adopted such measures, while “in years of decrease, the decreases in the states [with convenience voting] are greater.”[52]

Election Day registration, particularly with its increased risk of ballot fraud, is not the answer to low turnout or registration.

Alternative Approaches to Registration Reform

States can help to ensure voting roll accuracy. There is no question that the U.S voter registration system could be improved. As the Pew Center on the States found, one of every eight registrations in the United States is “no longer valid or [is] significantly inaccurate.”[53] Over 1.8 million deceased voters remain registered, and almost 3 million people are registered in more than one state. However, the answer to these problems is not federal mandates or federal interference in election administration, which should be reserved to the states, consistent with America’s decentralized election administration system.

According to Kansas Secretary of State Kris Kobach, federal mandates would be “completely unworkable” and would “make a mess” of state voter registration databases. States have already begun to implement state-based, bipartisan remedies to voter registration problems that preserve the balance of power between states and the federal government while maximizing new registration technology in order to ease, rather than remove, an individual’s responsibility to register himself.

For example, Secretary Kobach has initiated the “Interstate Voter Registration Crosscheck Program” to “increase the number of eligible citizens who register to vote” while ensuring “that accurate and current voter registration rolls are maintained.”[54] As of January 10, 2013, 21 states are participating in this program,[55] comparing their voter registration lists to detect multiple registrations (and votes) by the same individual in different states. By the end of 2012, 15 states had compared over 45 million records, turning up hundreds of thousands of potentially duplicate registrations.

For those voters who registered in a new state because they moved but neglected to notify election officials in the state of their former residence, this program gives them an opportunity to correct their registration. For those who intentionally register in more than one state to commit fraud, it helps states to discover violations of the law that threaten the integrity of elections—violations that in the past have been almost impossible to detect. Prosecutions of individuals who were found to have voted in two different states under this program, according to Kobach, have already been initiated.

Similarly, the Pew Center on the States is working on a project with seven states—Colorado, Delaware, Maryland, Nevada, Utah, Virginia, and Washington—to improve the accuracy of voter registration lists and improve voter registration rates. This initiative consists of comparing registration lists with “other data sources to broaden the base of information used to update and verify voter rolls,” using the same proven data-matching techniques developed in private industry “to ensure accuracy and security,” and developing new ways for voters to submit registration information to “minimize manual data entry” errors.[56]

After a long struggle with the Obama Administration, states such as Florida and Colorado are also starting to gain access to the Department of Homeland Security’s records on aliens in order to check the citizenship status of registered voters. However, as Secretary Gessler noted while speaking at The Heritage Foundation, the DHS records are incomplete and contain errors. While access to the DHS database is needed, such access is no substitute for, or nearly as effective as, requiring individuals registering to vote or voting to provide proof of identity or citizenship as Georgia, Alabama, and Arizona have done.

Pouring huge amounts of information, much of it full of errors and mistakes, from federal databases into state voter registration databases would only make the current problems exponentially worse. States are solving the problems that exist in registration lists; additional federal bureaucracy will not help.

Moreover, the U.S. Election Assistance Commission, created by the Help America Vote Act of 2002, is one of the most dysfunctional agencies in the federal government and does not have the resources, personnel, or knowledge to direct states. These proposals that supposedly are intended to help states improve the accuracy and validity of state voter registration lists could instead sabotage the progress that states are already making.

States are improving the voter registration process. The National Voter Registration Act made voter registration easy: It requires voter registration at state DMV, welfare, and disability agencies and military recruitment offices, as well as mandating mail-in registration. Yet states have been initiating new measures to make registration even simpler. States like Colorado, Louisiana, and Georgia have implemented online registration that allows individuals who already have a state driver’s license to register to vote over the Internet. Colorado voters can register using the state’s online voter registration system through their computers, phones, or tablets. And Louisiana has implemented a smartphone application that allows voters to access information about their registration, polling location, voting district, and sample ballots.

In 2012, Colorado Secretary of State Gessler sent notices to 700,000 Coloradans who might be eligible to vote but were not yet registered to encourage and help them to register for the upcoming election.[57] By Election Day, Colorado voter registration reached a record level: 440,888 more voters registered than in 2008, a 13.7 percent increase. Colorado’s increase in turnout is even more notable when considering that most of the nation saw a decrease in turnout in 2012 compared to the 2008 election. Secretary Gessler attributes this increase to the deployment of “new technologies and systems such as multi-state data matching, electronic ballot delivery for military and overseas voters, and high-speed Ballot on Demand printers.”[58]


The federal government and Members of Congress should respect differences among states. America is not homogenous, and one size does not fit all, especially when it comes to issues like voter registration. Citizens in different states have different needs, desires, and values; therefore, it makes little sense for the federal government to micromanage state voter registration systems. Indeed, the federal government has almost no experience administering elections; states are the experts on voting and, as such, are already implementing new programs and systems to improve the accuracy, effectiveness, and ease of the voter registration process.

Requiring automatic registration from government databases risks the integrity of the election process and improperly shifts the responsibility for registering from the individual to the government. States are already using federalism and their unique responsibilities in the voting process as originally intended: to experiment in the laboratories of democracy. The improvements these states are implementing come at less cost—to our treasury, our Constitution, and the integrity of our elections—than mandatory universal registration.

—Hans A. von Spakovsky is a Senior Legal Fellow at The Heritage Foundation and a former Commissioner on the Federal Election Commission. He is the coauthor of Who’s Counting? How Fraudsters and Bureaucrats Put Your Vote at Risk (Encounter Books, 2012).

[1] George F. Will, Mandatory Voting: Is This the Obama Administration’s Goal? Investor’s Business Daily, Dec. 18, 2012, available at

[2] The Brennan Center’s first paper on this concept in 2008 was entitled “Universal Voter Registration.” In 2009, the Center issued an almost identical paper in which the title had been changed to “Voter Registration Modernization.” Apparently, “modernization” was believed to be a better term than “universal” for advocacy on this issue. The latest reissue of this paper, “The Case for Voter Registration Modernization,” appeared in 2013 and keeps the modernization language.

[3] FairVote, 7 Ways to Universal Voter Registration,

[4] U.S. Census Bureau, Voting and Registration in the Election of November 2008, Table 6 (2012), available at . Since this is a survey of registration and turnout as reported by voters, it may vary from actual registration and turnout reported by state election officials.

[5] Ilya Somin, Are American Voters Stupid? Maybe Not, South China Morning Post, Sept. 27, 2004, available at

[6] U.S. Census Bureau, Voting and Registration in the Election of November 2010—Detailed Tables (Oct. 2011),

[7] Id. at Table 10.

[8] Wendy Weiser et al., Brennan Ctr. for Justice, Voter Registration Modernization (2009), available at

[9] Id. at 1.

[10] Jason Marisam, Voter Turnout: From Cost to Cooperation, 21 St. Thomas L.R. 190, 202–03 (2008).

[11] U.S. Census Bureau, supra note 4, at Table 1.

[12] Id. at Table 2.

[13] Randall D. Lloyd, Motor Voter: A Dismal Failure, Nevada Journal (Feb. 1999), available at

[14] Jack C. Doppelt & Ellen Shearer, Nonvoters: America’s No-Shows 214 (1999).

[15] Id. at 220.

[16] Id. at 219.

[17] FairVote, supra note 3; see also Weiser, supra note 8.

[19] Wendy Weiser et al., Brennan Ctr. for Justice, Universal Voter Registration (2008), available at; Weiser, supra note 8.

[20] Andy Kroll, Revealed: The Massive New Liberal Plan to Remake American Politics, Mother Jones (Jan. 9, 2013),

[21] Scott Malone & David Ingram, U.S. Should Automatically Register Voters: Attorney General, Reuters (Dec. 12, 2012),

[22] Thomas Perez, Assistant Attorney General for Civil Rights, Address at the George Washington University Law School Symposium (Nov. 16, 2012), available at

[23] H.R. 12 is also cosponsored by Reps. Steny Hoyer (D–MD), James Clyburn (D–SC), John Conyers (D–MI), and Robert Brady (D–PA).

[24] Weiser, supra note 8 at 8.

[25] Voter Empowerment Act of 2013, H.R. 12, 113th Cong., § 111 (2013).

[26] See, e.g., Arcia v. Detzner, -- F.Supp.2d. --, 2012 WL 6212564 (S.D. Fla. 2012).

[27] Florida State Conference of NAACP v. Browning, 522 F.3d 1153, 1155 (11th Cir. 2008).

[28] Press Release, BerlinRosen Public Affairs, Voting Rights Advocates Challenge Florida Registration Law in Federal Court (Sept. 17, 2007).

[29] Washington Ass’n of Churches v. Reed, 492 F.Supp.2d 1264 (W.D. Wash. 2006).

[30] Justin Levitt et al., Brennan Ctr. for Justice, Making the List: Database Matching and Verification Processes for Voter Registration (May 24, 2006), available at

[31] Weiser, supra note 8.

[32] United States v. Florida, 870 F.Supp.2d 1346 (N.D. Fla. June 28, 2012); Arcia, 2012 WL 6212564.

[33] Voter Empowerment Act of 2013 § 112(b)(3).

[34] Id. § 112(d).

[35] Id. § 112(d) and §112(f)(1) (although this section does not “waive the liability of any individual who knowingly provides false information to any person regarding the individual’s eligibility”).

[36] Id. § 112(f)(2).

[37] Id. § 112(j) and (k).

[38] See Elections Canada,

[39] Conference Board of Canada, How Canada Performs: Voter Turnout,

[40] John Ibbitson, The Alarming Decline in Voter Turnout, The Globe & Mail (Oct. 14, 2011),

[41] Weiser supra note, 8 at 9.

[42] Weiser, supra note 8, at 10.

[43] Project Vote, Voter Caging,

[44] Additionally, a change of mailing address does not always mean that an individual has changed his or her residential address for residency and voting purposes.

[45] See 5 U.S.C. § 552(a) note (Disclosure of Social Security number); see also Schwier v. Cox, 340 F.3d 1284 (11th Cir. 2003).

[46] Voter Empowerment Act of 2013 § 121. Currently, eight states (Idaho, Iowa, Maine, Minnesota, Montana, New Hampshire, Wisconsin, and Wyoming) and the District of Columbia allow Election Day registration. Two additional states have passed Election Day legislation; Connecticut’s new law will take effect on July 1, 2013, and California’s law will take effect on January 1 of the year following the year the secretary of state certifies that the state has a statewide voter registration system that complies with the Help America Vote Act of 2002 (no sooner than January 2014). National Conference of State Legislatures, Same Day Registration,

[47] Milwaukee Police Dept., Report of the Investigation into the November 2, 2004, General Election in the City of Milwaukee 26 (2008), available at

[48] American University News, African–Americans, Anger, Fear and Youth Propel Turnout to Highest Level Since 1960, 14 ( Dec. 17, 2008), available at

[49] Id.

[50] Maine Heritage Policy Ctr., Protecting the Integrity of Maine’s Elections: Election-Day Registration in Maine 4 (2011), available at

[51] Bipartisan Policy Ctr., 2012 Election Turnout Dips Below 2008 and 2004 Levels: Number of Eligible Voters Increases By Eight Million, Five Million Fewer Votes Cast 2 (2012), available at

[52] American University News, supra note 48.

[53] Pew Ctr. on the States, Inaccurate, Costly, and Inefficient: Evidence That America’s Voter Registration System Needs an Upgrade 1 (2012), available at

[54] Kris Kobach, Kansas Secretary of State, Presentation at Meeting of Nat’l Ass’n of State Election Directors (Jan. 26, 2013).

[55] Arizona, Arkansas, Colorado, Illinois, Iowa, Kansas, Kentucky, Louisiana, Mississippi, Missouri, Nebraska, Oklahoma, South Dakota, and Tennessee.

[56] Pew Ctr. on the States, supra note 53.

[57] Press Release, Scott Gessler, Colorado Secretary of State, Colorado Registers Another Successful Election: Voters Exceed 2008 Turnout (Nov. 9, 2012).

[58] Colorado Secretary of State Scott Gessler, 2012 General Election Review: A Colorado Success Story (Feb. 7, 2013).


Hans von Spakovsky

Election Law Reform Initiative Manager, Senior Legal Fellow