Election Deform: The Pursuit of Unwarranted Electoral Regulation

In 2008, for the first time in the nation's history, a candidate identifying (at least in part) as a member of a racial minority group was a major party's nominee for President of the United States. The overwhelming significance of this milestone has obscured, if only by comparison, the fact that the 2008 electorate was in some ways also quite distinctive. More than 130 million Americans, millions more than ever before, successfully cast a ballot.¹ In aggregate, turnout was about 63% of eligible electors—the highest turnout rate since 1960, and the third highest since 1920.²

Yet the volume of voters itself masks a substantial and noteworthy change in the character of those who were successfully motivated and able to exercise the franchise. Generally, white non-Hispanic turnout has exceeded minority turnout by a substantial margin: according to the U.S. Census Bureau's Current Population Survey, the gap has exceeded 10 points in every election since 2000—other than 2008.³ In 2008, the gap narrowed substantially, with non-Hispanic white turnout estimated at 66.1%, and minority turnout estimated (at the highest level in a decade) at 56.8%.

In 2010, as might be expected in most any midterm election year, the turnout wave receded, with estimates ranging from 42%⁴ to 46%⁵ of eligible voters casting ballots. The character of the voting public also reverted somewhat, with minority turnout lagging non-Hispanic white turnout by 11.3%: the gap remained closer than in the past midterm elections of the decade, but two points larger than the gap in 2008.⁶

Now, in 2011, America has witnessed a flurry of election legislation.⁷ Unfortunately, the new spate of state regulations include several highly controversial laws that appear to exact real burdens on real Americans, making it more difficult for citizens to exercise their rights to vote. Moreover, available data suggest that these burdens tend to fall unequally; while 2008 saw some steps in fostering a voting population that more equitably approximates the electorate as a whole, 2010 represented a step backward, and these new laws threaten to continue an unfortunate regressive pattern.⁸ Crucially, these burdens are not only real and inequitable but also unnecessary, which renders them suspect as a matter of constitutional law⁹ and fundamentally flawed as a matter of public policy.¹⁰ Not only do they make it more difficult for eligible Americans to vote, but they do so without any meaningful benefit. Indeed, in several circumstances, the new laws are directly counterproductive.

Although there are several new state laws and policies that deserve attention, this article focuses on three types of restrictions that together demonstrate the breadth of the concern. In Part I, I examine new limits on the ability to help citizens register to vote. In Part II, I examine new limits on citizens' ability to cast ballots before Election Day. And in Part III, I examine new limits on citizens' ability to establish their identity at the polls on Election Day itself, devoting disproportionate attention to the issue because of the disproportionate degree of misinformation it seems to have engendered.

Getting on the rolls, early voting, and voting on Election Day: all have been limited in new, and apparently unjustified, ways by the legislation of 2011.

I. Restrictions on Registration

First, 2011 brought renewed regulation restricting the ability of citizens to offer their colleagues assistance in registering to vote. These efforts are exemplified by provisions of Florida's HB 1355,¹¹ which was passed earlier this year; portions of the law remain subject to pending preclearance proceedings under section 5 of the Voting Rights Act, in the five Florida counties “covered” by section 5.¹²

HB 1355 is Florida's latest in a series of attempts to restrict voter registration over the past few years; its earlier efforts were also highly controversial, and challenged in court.¹³ In 2005, ostensibly concerned by organizations allegedly withholding and belatedly submitting or refusing to submit registration forms that they collected, the legislature imposed, inter alia, substantial restrictions on organizations conducting voter registration drives.¹⁴ These restrictions included substantial fines, with both individual and organizational liability, for each and every form delivered to elections officials more than 10 days after the form was completed.¹⁵ The fine structure was sufficiently severe to cause the nonpartisan League of Women Voters to stop its Florida voter registration activity for the very first time in the organization's 67-year history.¹⁶

In subsequent litigation, a federal court rightly recognized that voter registration drives entail core political speech, protected by the First Amendment and inextricably intertwined with efforts to “persuade others to vote, educate potential voters about upcoming political issues, communicate their political support for particular issues, and otherwise enlist like-minded citizens in promoting shared political, economic, and social positions.”¹⁷ And it rightly recognized that undue efforts to restrict registration drives impermissibly limit both political speech and association. The court explained that Florida had not “provided any evidence[,] much less an explanation,” supporting the need for its 2005 fine structure—and preliminarily enjoined the implementation of this portion of the law.¹⁸

One year later, the legislature enacted an amended law.¹⁹ The new law retained the 10-day deadline, but substantially reduced and capped total fines, and exempted organizations from fines due to situations beyond their control. It was challenged and upheld based on the more tailored regulatory structure—and based on the fact that the law did not place any direct preconditions on the protected activity of conducting a voter registration drive.²⁰

HB 1355 marks a severe step backward on both fronts flagged by the courts: it is no longer reasonably tailored to any existing problem, and imposes serious pragmatic obstacles to organizations as preconditions of conducting voter registration drives. It requires any person—any individual or group—to fill out an official state form before offering to help distribute, collect, and submit the registration form of anyone other than immediate family; this registration includes the name, address, and sworn declaration of every single individual soliciting or collecting registration forms, whether employee or casual volunteer.²¹ Groups may not offer to collect and turn in a single form until they have been issued a number by the state; individuals who are not working with organized groups are subject to the same requirements. The law requires that every individual and group account monthly for every registration form used by any volunteer, including blank forms simply printed off of public Web sites; county election officials have new daily reporting requirements.²² And without any indication that the ten-day deadline of prior law was insufficient to compel the prompt return of completed forms, the deadline has now become just 48 hours, with any waiver for circumstances beyond the organization's control now solely in the hands of the Secretary of State, a partisan elected official.²³

It is worth restating what the new law requires. Before offering to touch a voter registration form from anyone other than a family member, citizens volunteering their time must wait for permission from the government. In addition to tracking each and every registration form, blank or complete, a volunteer collecting a registration form must ensure that it is delivered to county officials within 48 hours, or face substantial fines issued or waived at the discretion of a partisan official.

These are stark limitations of, and penalties on, fundamental public engagement. They are the most restrictive provisions in the country, though recent legislation in Texas has some similar hallmarks.²⁴ They should trouble observers and policymakers across the political spectrum. Indeed, far less onerous regulations of political campaign spending in Florida were recently challenged as severe constitutional burdens by the Institute for Justice.²⁵

Given its burdens, the law will have some predictable effects—few of which increase the reliability of the registration system in any meaningful respect. Instead, the law has caused both Democracia USA, one of the larger civic engagement organizations in Florida dedicated to empowering the Latino electorate, and the League of Women Voters, a nonpartisan civic engagement enterprise of unparalleled lineage, to declare a halt to all voter registration activity within the state.²⁶

When voter registration drives are unable to offer their assistance, citizens lose one vital means to ensure that they are properly registered to vote—not merely new registrants, but also the 14% of Floridians who move within the state and need to re-register.²⁷ Moreover, the population impacted by such restrictions is not evenly distributed. According to the U.S. Census Bureau's Current Population Survey,²⁸ minority citizens disproportionately register and re-register through voter registration drives: while 6% of non-Hispanic white voters reported registering through a voter registration drive in 2008, twice as many—12% of Hispanic voters and 13% of non-Hispanic African-American voters²⁹—reported registering through a drive.³⁰ Statistics from non-presidential years are similarly lopsided. In 2006, 8% of non-Hispanic white voters reported registering through a voter registration drive, compared to 11% of Hispanic voters and 11% of non-Hispanic African-American voters; in 2010, 6% of non-Hispanic white voters reported registering through a voter registration drive, compared to 14% of Hispanic voters and 12% of non-Hispanic African-American voters.³¹

These new restrictions on civic participation are particularly troublesome in that they put Florida's League of Women Voters out of the voter registration business unnecessarily. That is, there is no compelling public policy need for such prerequisite burdens on informal voter registration drives on campuses, in houses of worship, at casual gatherings of friends and acquaintances, and in the many other circumstances in which individuals assist their fellow citizens without first creating a bureaucratic documentation, reporting, and tracking apparatus. Florida already had legal provisions requiring that voter registration forms be delivered in a timely fashion.³² Florida already had legal provisions vigorously defended in court as ensuring the accuracy of registration form information.³³ Florida already had legal provisions penalizing any intentional wrongdoing in the registration process.³⁴ The new regulations impose a burden out of proportion to their incremental benefit.

Indeed, the new regulations might well increase the expense to election officials. Only the most formally structured voter registration drives will, as a practical matter, be able to comply with the advance documentation requirement. Less formal citizen organizations will find it prohibitively impractical to ensure that volunteers at bake sales—and actual tea parties—have submitted sworn paperwork before they offer to help send in a neighbor's voter registration form. Many of these larger, more formally structured drives have historically conducted quality assurance, reviewing forms for errors or suggestions of impropriety, and flagging those forms for election officials to expedite processing. The 48-hour time limit on returning forms, however, is likely to constrain organizational ability to conduct centralized quality review. Instead, rational organizations seeking to forego liability will likely curtail centralized quality assurance in favor of speedy delivery, shifting processing and error-correction costs unnecessarily to the county supervisors.

II. Restrictions on Early Voting

Second, there have recently been efforts to limit opportunities for citizens to cast valid ballots in advance of Election Day. Here too, Florida's HB 1355 provides an example.

At least since 1998, Florida has allowed electors to vote ballots in-person before Election Day.³⁵ Such votes could originally be cast as soon as absentee ballots were available, on any day that the county supervisor's office was open. Beginning in 2004, the state limited its early voting period to two weeks, beginning on the 15th day before an election, and ending on the day before Election Day; the next year, the legislature eliminated early voting on the Monday before an election.³⁶ Jurisdictions were required to offer early voting for 8 hours per weekday, and 8 hours in the aggregate per weekend—96 early-voting hours total—but had discretion to apportion those weekend hours as they chose.³⁷

HB 1355 changed the early voting schedule again, restricting local authority. The new early voting period runs from Saturday (10 days before Election Day) to Saturday (3 days before Election Day), with 6–12 voting hours per day.³⁸ If county supervisors choose to offer the maximum permissible early vote schedule under HB 1355, voters would continue to have 96 total early-voting hours.

The allocation of these hours, however, represents a significant change for the worse. The most significant restriction is that jurisdictions no longer have the option to offer early voting on the Sunday before Election Day.³⁹ This was an option that several counties had offered in the past, as a service to their constituents, many of whom work long hours during the week, are more available on the weekend, and are most energized just before Election Day. The list of jurisdictions choosing to offer early voting on the Sunday before Election Day in the past includes the state's largest, most urban, and most diverse counties. In 2008, Bradford, Broward, Dixie, Duval, Jackson, Miami-Dade, Palm Beach, Pinellas, Sarasota, and Seminole counties offered early voting on the Sunday before Election Day; in 2010, Bradford, Charlotte, Clay, Duval, Manatee, Miami-Dade, Palm Beach, Pinellas, Sarasota, and Seminole counties offered early voting on that Sunday.⁴⁰ Under HB 1355, the state has precluded these counties from offering the service they wished to offer.

The change has a direct impact on a particularly notable form of mobilization in Florida: many houses of worship, particularly in minority communities, encourage their congregations in nonpartisan fashion to discharge their civic obligations after fulfilling their spiritual ones.⁴¹ After Sunday morning church services, many congregants would travel to the polls in the counties that offered Sunday voting. After HB 1355, this is no longer an option on the most important Sunday in the voting cycle, when most would-be voters are most engaged.

As with the restriction on registration drives, the elimination of early voting on the Sunday before the election does not fall evenly on the population as a whole. In the past, minority citizens disproportionately voted on the final Sunday before Election Day.⁴² In 2008, for example, African-Americans represented 13% of the total voters, and 22% of the early voters, but 31% of the total voters on the final Sunday; Hispanic citizens represented 11% of the total voters, and 11% of the early voters, but 22% of the total voters on the final Sunday. Notably, the pattern is similar in 2010: African-Americans represented 12% of the total voters, and 13% of the early voters, but 23% of the voters on the final Sunday; Hispanics represented 9% of the total voters, and 8% of the early voters, but 16% of the voters on the final Sunday.

Voters on the final Sunday before the election were also predictably newer voters. In 2008, first-time Florida voters were 12% of the electorate but 22% of the final Sunday voters. And they continued this pattern; in 2010, voters casting a ballot for the first or second time were 13% of the electorate, but 17% of the final Sunday voters.

As noted above, the costs of eliminating the final Sunday from early voting far exceed the potential benefits, because restricting county flexibility in this fashion has no appreciable upside. Before HB 1355, counties had the option to offer early voting on the final Sunday before Election Day if they wished. If county constituents used Sunday voting, if offering Sunday voting increased convenience, if Sunday voting offered a logistical means to ease excessive lines on Election Day, if Sunday voting were cost-effective, county supervisors were authorized to decide for themselves to open early voting stations. If Sunday voting were not cost-effective for the electorate of a particular county, the supervisors could opt to use the weekend time exclusively on Saturday instead. HB 1355 removes that flexibility, forcing the counties to shut their early-voting doors on Sunday whether they would prefer to do otherwise or not. For counties that had previously offered Sunday voting because they found it worthwhile, HB 1355 can only increase inconvenience, and in some circumstances, expense.

III. Restrictions at the Polls

Third, there have recently been renewed efforts to limit opportunities for citizens to cast a valid ballot at the polls, most notably in the form of new restrictions on how those citizens may demonstrate their identity. In 2011, four states—Kansas,⁴³ Tennessee,⁴⁴ Texas,⁴⁵ and Wisconsin⁴⁶—passed new restrictive laws requiring most citizens to show particular types of government-issued photo identification cards in order to cast a ballot at the polls that can be counted.⁴⁷ The considerable misinformation surrounding these laws impedes reliable cost-benefit analysis. Though it is beyond the scope of this piece to clarify all of the relevant misinformation, it is nevertheless worth addressing a few of the more substantial and oft-repeated myths.

A. The current identification landscape

Before 2011, only two states in the country—Indiana⁴⁸ and Georgia⁴⁹—required government-issued photo identification in order to cast a ballot at the polls that can be counted. The four additional states mentioned above represent disturbing additions, but they remain, together, only a small minority of jurisdictions.

Instead, the vast majority of states allow legitimate citizens a broader set of options to prove their identity, without sacrificing any appreciable measure of security. The alternatives range from signature comparisons, to sworn affidavits, to identification documents like utility bills, bank statements, employee IDs, and the like.⁵⁰ Some of these other states ask those without government-issued photo identification to vote provisional ballots, which can be further investigated if additional doubt arises about a voter's identity; these provisional ballots, however, can be counted without requiring the voter to provide the same photo identification card that she could not produce at the polls.⁵¹ And all of these identification provisions are layered atop the considerable security safeguards of the federal Help America Vote Act of 2002 (HAVA), which requires that each of a jurisdiction's first-time voters registering by mail have her identity confirmed—either by verifying her social security digits or driver's license number against reliable lists, or by presenting reliable documentation from a long and inclusive menu—before her ballot may be counted.⁵²

These other 44 states offer alternatives for a reason. They recognize that there are some legitimate, eligible American citizens who do not possess government-issued photo identification cards. And they do not wish to make it unduly difficult for these citizens to exercise the most fundamental right in our constitutional order.

B. Evaluating the burden of restrictive identification rules

There is no question that government-issued photo identification makes many common practices easier. Those who do not have such ID are likely to find it more difficult to take advantage of many of the privileges of modern society.

Still, the ability of Americans to function in modern society without government-issued photo identification is frequently exaggerated. Proponents of restrictive ID laws often assert that it is impossible to board an airplane, drink a beer, check in at a hotel, or enter a federal building without government-issued photo ID.⁵³ These claims are both untrue and irrelevant.

Personal experience proves the falsity of the above assertions. Recently, I was asked to testify at a hearing of the U.S. Senate Judiciary Committee's Subcommittee on the Constitution, Civil Rights, and Human Rights, pertaining to newly restrictive electoral laws, including restrictive identification rules.⁵⁴ When I arrived at the Los Angeles airport on September 7, 2011, my wallet contained no photo identification, government-issued or otherwise. Instead, it had two credit cards, a firing range card, a health insurance card, a blood donor card, a coffee shop frequent visitor card, and a few business cards, all without photos. I was also carrying a checkbook.

The TSA officer at the airport check-in station examined my boarding pass, and asked me to step aside for additional questions; another officer reviewed my non-photo documentation, and asked more questions. I was then asked to step through the (regular) security line, where my bags were screened, and a backscatter image was taken. I estimate that the procedure lasted approximately ten minutes longer than the normal procedure experienced by individuals in the same line who had photo identification on hand. This was policy, not parlor trick; as I have written before, the TSA and its predecessor agencies have consistently maintained a policy permitting those without particular government-issued photo identification to board a plane after additional screening.⁵⁵

After clearing security, I enjoyed a beer in the airport Chili's—without using photo ID. When I arrived in Washington, DC, I checked into my hotel—without using photo ID. I then made my way to the Dirksen Senate office building, and to the Committee's hearing room—without using photo ID. Commercial vendors and federal governments alike have demonstrated that when it is financially or politically important to extend access even to citizens without certain photo identification, such citizens can be accommodated with minimal disruption to normal business practices.

Government-issued photo identification, unquestionably, makes life easier for those of us who have it. But it is simply not true that one cannot board a plane—or otherwise navigate existence in America—without it. And even if government-issued photo identification were required in order to board a plane, purchase alcohol, check into a hotel, or (another frequent falsehood) purchase full-strength Sudafed,⁵⁶ access to those conveniences would be deeply beside the point. The franchise is referenced in two Articles and at least ten Amendments of the Constitution.⁵⁷ It is central to our constitutional order, a fundamental right protected for all eligible citizens once it is granted to some.⁵⁸ The protection of access to Sudafed is not quite so jealously guarded. No American ever gave her life to preserve a fundamental constitutional right to buy decongestants.

There is also no question that most eligible citizens have government-issued photo identification. It is likely that those reading this article have some form of government-issued photo identification. But the right of the franchise—and the responsibility to ensure its continued reasonable access—is not limited to the individuals perusing these pages, or even to the majority of the eligible American public. Voting is a fundamental right for more than just most of us.

It is, concededly, difficult to pin down the precise number of eligible American citizens who do not have the identification required by the most restrictive states above. But of the three methods that have frequently been used to estimate this population, two are substantially less reliable. First, some commentators have compared the number of records maintained on state Department of Motor Vehicles systems to the number of voting-age citizens reported by the Census Bureau.⁵⁹ Such comparisons are laden with error, including duplicate driver's licenses (commercial and non-commercial), expired licenses, and the impact of the 2% of Americans each year who move between states, often without canceling their motor vehicle record in the state that they have left.⁶⁰

Second, other commentators have attempted to assess the number of eligible citizens without the required identification by analyzing turnout: examining past voting patterns and trying to extrapolate the degree to which change in participation for any given election is due to the impact of particular identification laws.⁶¹ These studies' methods vary, and there are substantial differences in the results. But more fundamentally, the basic approach is flawed. It is for one thing conceptually incomplete: even if turnout provided an accurate assessment of the impact on past voters, it would cover only the impact on past voters, without any assessment at all of the impact on eligible Americans who have not yet participated but have every right to participate in the future.

Moreover, even if turnout provided the right measure of impact, we are not yet reliably able to mine the available data for useful information on the question presented. It is exceedingly difficult to interpret a few years of turnout data from a few jurisdictions to parse the impact of any given legal change. For example, proponents often cite the change in turnout—particularly minority turnout—in Georgia and Indiana before strict ID laws (in 2004) and after strict ID laws (in 2008), as evidence that ID laws did not impose any substantial impediment on the voters in those states, and on voters in general, including the minority electorate.⁶² But those proponents also often fail to note that both Georgia and Indiana were newly battleground states in 2008, with a minority candidate at the top of a major-party ticket for the very first time. Under those circumstances, any reasonable observer would have expected extraordinary increases in minority turnout, with or without ID laws. And under those circumstances, it is difficult to know whether a 19% increase in turnout⁶³ reflects a 15% increase because of the circumstances of the 2008 election with a 4% increase because of ID laws, or a 30% increase because of the 2008 election and an 11% decrease because of ID laws…or any other combination of causal responsibility.

This is a specific example of a general problem: in any given election, turnout may be affected by the competitiveness of high-profile races, candidate quality, fundraising and campaign spending, the media environment, the presence or absence of salient ballot measures, the efforts of mobilization groups on the ground, other legal restrictions or policies that facilitate access, and a host of other conditions, including the weather on Election Day. Without thousands of data points to account for all of the other factors that could instead be driving turnout up or down, it is unreliable to draw conclusions about the impact of identification rules by looking at how many people vote in a given election.⁶⁴

Instead of comparing motor vehicle records to census numbers, or reading the uncertain tea leaves of turnout, another methodology is available to determine how many eligible citizens have the sort of identification documents required by the most restrictive state laws: asking them. Several surveys have asked eligible Americans about the documentation they possess, with some varying conclusions.

Some of this research surveys either registered voters or actual voters—as explained in the discussion of turnout above,⁶⁵ such numbers understate the impact of strict identification requirements, because they do not include eligible Americans who may participate in the future. A 2008 survey found that 4.9% of registered voters responding nationwide did not have current government-issued photo identification; an additional 3.1% of respondents did not have current government-issued photo identification listing their full legal name (rather than, for example, a nickname or maiden name).⁶⁶ Another 2008 survey found that 5.7% of registered voters nationwide did not have a current valid driver's license or passport; an additional 1.1% of respondents had those documents, but not listing their full legal name.⁶⁷ (The same survey found that 4.7% of respondents had no valid driver's license or passport, but did have other government-issued photo identification; the survey did not inquire whether this latter ID was current.)⁶⁸ Still another 2008 survey found that 1.2% of registered voters in Indiana, Maryland, and Mississippi did not have any government-issued photo identification, but did not inquire whether the subjects' ID was current or reflected the same name on the registration rolls.⁶⁹ A 2007 survey found that 13.3% of registered voters in Indiana did not have a current government-issued photo identification card; an additional 3% of respondents did not have current identification listing their full legal name.⁷⁰ A 2006 survey found that 12% of actual midterm voters in California, New Mexico, and Washington did not have a valid state driver's license, but did not inquire whether the subjects had a non-driver's government-issued photo identification card.⁷¹

Other research surveys voting-age American citizens, whether currently registered or not. A 2007 survey found that 16.1% of voting-age citizens in Indiana did not have current government-issued photo identification; an additional 2.8% of respondents did not have current identification listing their full legal name.⁷² And a 2006 survey found that 11% of voting-age citizens nationwide did not have current government-issued photo identification.⁷³

Each of the above surveys appears reliable. Some of the variance can be explained by the difference in the questions asked (e.g., whether particular forms of identification are specified, or whether the identification is determined to be current) or the difference in the target population; some of the variance may simply reflect differences from state to state. Additional variance may simply reflect the natural variability inherent in surveys, which are estimates and subject to different weighting schemes and margins of uncertainty. Further reliable surveys, both national and state-specific, would be welcome.

But it is important to note that even choosing the most conservative estimate—a survey targeting registered voters in select states, rather than the electorate as a whole, and without assessing whether the ID in question was current or reflected the same name on the registration rolls—1.2% of registered voters do not have the identification required by the most restrictive states. Even this substantially conservative result amounts to an impact reaching more than two million registered voters if applied nationwide.⁷⁴ And the larger estimates show an impact reaching more than twenty-two million voting-age citizens.⁷⁵

Moreover, every study to have examined the issue has found that those without government-issued photo identification are not evenly spread across the electorate. Just as the surveys differ in their overall assessment of the magnitude of the problem, they differ in their assessment of magnitude of the disparate impact. But the available data clearly show that those without government-issued photo identification are more likely to be nonwhite, more likely to be either younger voters or seniors, more likely to be from low-income households, and more likely to have less formal education.⁷⁶ And while I am not aware of a reliable measurement of the incidence of government-issued photo identification among persons with disabilities, there is reason to be concerned that they, too, are less likely to have the identification required by the most restrictive states.

These impacts are both substantial and statistically significant. For example, one 2008 survey found that while 3.7% of responding white registered voters nationwide did not have current valid government-issued photo identification, 7.3% of Latino voters and 9.5% of African-American voters lacked this ID.⁷⁷ And among voting-age citizens rather than registered voters, a 2006 national survey found that 8% of white citizens but 16% of Latino voting-age citizens and 25% of African-American voting-age citizens do not have current, valid, government-issued photo identification.⁷⁸ While other studies differ in the precise magnitude of these (and other) differential ID rates, all show a substantial effect, with historically underrepresented groups much less likely to have current government-issued photo identification.

These statistics are not merely important for their reflection of the status quo, but for their reflection of significant impact into the future. It often takes ID to get ID. For example, Arkansas seems to require most native-born citizens to provide an official copy of a birth certificate to obtain a government-issued photo identification card,⁷⁹ and seems to require government-issued photo identification to obtain an official copy of a birth certificate.⁸⁰

Even without this sort of vicious loop, those without current government-issued photo identification often face some difficulty in procuring it. All states of which I am aware require documentation to procure state-issued identification. Even when the identification card itself is offered free of charge,⁸¹ an individual without identification must collect this documentation, which involves time and expense, and travel (without driving) to a government office open during limited (working) hours, which involves time and expense. Official copies of birth certificates cost between $7 and $30 depending on the state, with a median of $15; expedited processing will cost more.⁸² A passport costs at least $55, and a replacement naturalization certificate costs $345.⁸³

Moreover, some eligible citizens will simply not be able to procure the requisite underlying documentation, no matter how much they are able to spend or how much time they are able to take. In August 2011, South Carolina's Attorney General recognized, in a formal opinion interpreting the state's new identification law, that there are legitimate electors

These eligible Americans have names. Dr. Brenda Williams, of Sumter, South Carolina, has recently been attempting to assist some of her patients in getting the government-issued photo identification required in the normal course by South Carolina's new law. Her comments to the Department of Justice show that she has spent hundreds of dollars helping her patients attempt to get the necessary ID.⁸⁵ And still some have been stymied. For example, Dr. Williams wrote to the Department of Justice about Mrs. Naomi Gordon and her brother, Mr. Raymond Rutherford.⁸⁶ Mrs. Gordon's first name was apparently misspelled “Llnoie” by a midwife; a midwife also apparently misspelled Mr. Rutherford's first name “Ramon.” The misspellings appear on both of their birth certificates; Mr. Rutherford has the particular difficulty of possessing a birth certificate with an incorrect spelling and a Social Security card with a correct spelling. They have been told that they have to have their names changed through the courts before they will be able to get government-issued photo identification; neither has yet been able to procure the appropriate ID.

Nora Elze of Savannah, Georgia is 88 and has been married for 65 years.⁸⁷ But because the name on her birth certificate (her maiden name) and the name on her out-of-state ID (her married name) do not match, she has to produce a 65-year-old marriage license in order to obtain government-issued photo identification. At last report, she had not found the marriage license, and had not been able to acquire the necessary identification.

In 2008, at least ten retired nuns in South Bend, Indiana, all citizens in their 80s or 90s, were reportedly turned away from the polls because they did not have current government-issued photo identification. One of the nuns noted that “many others among the 137 retired sisters living at the Congregation of the Sisters of the Holy Cross convent were dissuaded from voting upon learning that several had been turned away.”⁸⁸

Royal Masset, former political director of the Texas Republican Party, discussed a personally relevant situation in the press:

Agnes Cowan and her husband lost many of their personal documents in a fire, including her husband's veterans' ID card.⁹⁰ At 81 in 2008, and confined to a wheelchair, Ms. Cowan said that it was virtually impossible for her to cobble together replacement documentation in order to get a government-issued photo ID before Georgia's 2008 primary election, making it the first major election that Ms. Cowan had missed in 63 years.

Among the Indiana citizens prevented from voting a valid ballot in 2007 was 61-year old Republican Valerie Williams. Ms. Williams brought her telephone bill, a Social Security letter, and an expired state driver's license to the polls—but she did not have the current government-issued photo ID that Indiana required. Her provisional ballot was never counted.⁹¹

In 2006, Eva Steele was an Arizona resident and her son was an Army reservist deployed in Iraq.⁹² Her disabilities left her in a wheelchair and unable to drive. “‘I don't have a driver's license,’ she said. ‘I don't get utility bills. I've never had a passport. I don't have property tax statements. All I did was raise my children and teach them to be good citizens and to vote. And now I'm the one who's on the outside looking in.’”⁹³

Mary Wayne Montgomery Eble was 92 and on oxygen in 2008, living on a family farm outside Rockport, Indiana.⁹⁴ She had no driver's license because she could not see well enough to drive; she did not know if she had a birth certificate, because she was born at home.

Chris Conley, a 50-year-old veteran of the Navy and Marines, tried to vote in Indiana's 2008 primary, but his Veterans Administration photo ID card did not have an expiration date, and therefore did not meet the state requirements.⁹⁵

Birdie Owen was displaced from Louisiana after Hurricane Katrina, where her birth certificate was lost in the storm. Without a birth certificate, she found herself unable to get a state-issued photo identification card in Missouri.⁹⁶

The stories above represent just a selection of the reports of individuals—real American citizens—without government-issued photo identification. The reliable statistics cited above indicate that there are many others.

C. The lack of justification for restrictive identification rules

As with the other restrictions discussed in Parts I and II of this article, the heavy costs on Mrs. Gordon and other eligible American citizens are not justified by any substantial benefit. Laws preventing citizens from proving their identity at the polls by anything other than certain government-issued photo identification cards are often ostensibly justified by the need to prevent election fraud. Here too, there appear to be pernicious misconceptions.

Requirements to present certain identification at the polls provide even theoretical protection against only one form of fraud: someone who arrives at the polls and pretends to be someone else. As explained in more detail below, and as I have described extensively in previous publications and official testimony,⁹⁷ all of the available evidence demonstrates that the incidence of any fraud that identification rules could prevent is extraordinarily rare. Though it does occur, there are only a handful of recent accounts, even fewer of which have been substantiated. During this same period, hundreds of millions of ballots have been cast. The most notable significance of the incidents that have surfaced is how rare they appear to be.⁹⁸

In order to assess the incidence of fraud that identification rules could possibly prevent, it is first necessary to cut through a large amount of noise. Some reports or allegations of fraud are simply mislabeled; the substance of a newspaper story simply does not support a headline claim of “fraud.”⁹⁹ Other reports claim fraud but instead reveal straightforward administrative errors, or administrative practices that concern some people, but are not errors at all.¹⁰⁰

Some of these reports actually do present worrisome evidence of fraud—but not any sort of fraud that identification rules could prevent. Instead, they allege schemes involving fraudulent absentee ballots;¹⁰¹ or absentee voters who have been coerced;¹⁰² or conspiracies to buy votes;¹⁰³ or efforts to tamper with ballots or machines or counting systems.¹⁰⁴ There are occasional reports of double-voting by individuals voting in their own names and without appropriating another's identity.¹⁰⁵ There are occasional schemes of insider complicity or forgery;¹⁰⁶ when pollworkers and officials are willing to break the law, or miscreants are willing to forge documents, additional requirements for pollworkers to review official documentation cannot prevent the wrongdoing. It is impossible to stop local bosses intent on breaking the law by giving them a new law to break.

There are also reports of fraudulent registration forms, though they most often involve rogue workers hoping to cheat nonprofit organizations out of an honest effort to register real citizens.¹⁰⁷ These forms are usually subject to the safeguards of HAVA, which flags potentially invalid registration forms for further security measures before a corresponding ballot can be cast.¹⁰⁸ I am aware of no recent substantiated case in which such registration fraud has resulted in a fraudulent vote.

The above forms of fraud do, sadly, exist. They are real, legitimate concerns, though fortunately not as common as media attention may make them appear. Fraud should be both prevented and punished when doing so does not exact an even greater cost to the system and to legitimate electors therein. But extreme limits on the ways in which individuals prove their identity at the polls do nothing to address them. Using restrictive identification rules to prevent this fraud is like amputating a foot to get rid of the flu.

In addition to the noise created by allegations of fraud that identification rules cannot possibly prevent, noise has been generated by sloppy science. Some reports purport to reveal evidence of fraud based on attempts to match registration rolls to other government sources, like registries of the deceased, but these reports often betray familiar, and significant, methodological flaws. One particularly common error is the seemingly straightforward assumption that individuals with the same name and date of birth are the same person. As Professor Michael McDonald and I have demonstrated, elementary statistics confirms that in any substantial pool, it is quite common to find two different individuals who share the same name and date of birth.¹⁰⁹ When comparing one list of millions of voters to another list of millions of ineligible individuals, it should not be surprising to find hundreds of perfect “matches” that actually represent different individuals, known to record-linkage experts as “false positives.” The incidence of such matches reveals statistics at work, not fraud.

In sum, my research confirms that there are hundreds of reports of alleged fraud, in thousands of elections, with millions of ballots cast. Yet after wading through the unreliable and irrelevant reports categorized above, only a handful of reports remain that even allege, much less substantiate, instances of fraud that increased identification requirements at the polls could prevent.¹¹⁰

Even fewer of these allegations stand up to real scrutiny. Indeed, careful investigation has more often than not debunked, not confirmed, allegations of impersonation fraud at the polls. One notorious and recurring example is a 2000 investigative report in the Atlanta Journal-Constitution, claiming that “the actual number of ballots cast by the dead” was “5,412 in the past 20 years.”¹¹¹ The article has been favorably cited by an Assistant U.S. Attorney General,¹¹² a governor,¹¹³ a state secretary of state,¹¹⁴ and several state attorneys general,¹¹⁵ among others.

This article did not, however, actually reveal 5,412 ballots cast by the dead, much less 5,412 instances of in-person impersonation fraud.¹¹⁶ Instead, it revealed 5,412 matches of Social Security death records to voting records. And it further revealed that these matches are flawed. The reporter acknowledged that death records contain errors, listing people as dead who are actually alive, but apparently did not investigate how many of the 5,412 identified ballots suffered from this error. The reporter also acknowledged that voter records contain errors, reflecting data entry mistakes and those who sign the wrong line of a pollbook, but apparently could not or did not investigate how many of the 5,412 identified ballots suffered from this error. The reporter neither acknowledged nor apparently accounted for the statistical likelihood that a record of John Smith dying and a record of John Smith voting might in fact reflect different “John Smith”s with the same date of birth.¹¹⁷ Finally, the reporter did not indicate how many of these 5,412 ballots were cast in person, rather than absentee.

Indeed, the article identified only one individual concretely alleged to have been the victim of in-person impersonation fraud. It cited the case of “[Alan Jay] Mandel, the tobacco shop owner, whose voter certificate was signed at the polls by someone after his death.”¹¹⁸ Repeated the reporter, “[S]omebody definitely signed his name on a voter certificate on Nov. 3, 1998.”¹¹⁹

This allegation, though amounting to only one concrete allegation of in-person impersonation fraud in approximately twenty million votes over 20 years,¹²⁰ would nevertheless be disturbing—if it were true. Further investigation, however, proved the allegation false. The signature (and voter certificate) in question belonged to Alan J. Mandell (with two “l”s), who was very much alive and eligible in 1998, but whose vote was mistakenly recorded in the name of Alan Jay Mandel (with one “l”).¹²¹

Investigation as thorough as the investigation into the vote of Mr. Mandel/Mandell is rare. Nevertheless, when researchers do expend the effort to follow through on initial allegations of in-person impersonation fraud, they often find those allegations to be unwarranted. A 2008 investigation of 48 purported “dead voters” in Dallas, for example, revealed only clerical error, voter mistake, and confusion; of all the cases investigated, “none involved a fraudulently cast vote.”¹²² A 2007 investigation of approximately 100 “dead voters” in Missouri revealed that every single purported case was properly attributed either to a matching error, a problem in the underlying data, or a clerical error by elections officials or voters.¹²³ Likewise, after compiling a list of potential “dead voters” in New York state, a Poughkeepsie journalist investigated seven local cases—and found that seven out of seven reflected clerical errors or other mistakes, not fraud.¹²⁴ An investigation in Hawaii in 1999, after reviewing precinct pollbooks and calling allegedly deceased citizens, similarly found that not one of 170 potential “dead voters” actually reflected fraud.¹²⁵

The most prominent recent examination of voter fraud—the evidence presented to the Supreme Court in Crawford v. Marion County Election Board¹²⁶—precisely fits the overall pattern described above. There were many claims of wrongdoing and irregularity, but few that even alleged the sort of fraud that in-person identification rules could possibly prevent, and a tiny portion, if any, that substantiated the allegations.

The Crawford case is often said to have validated laws requiring photo identification at the polls. It did no such thing. In Crawford, the fractured court rejected the plaintiffs' challenge to the law as overbroad, in light of the limited evidence in the record on the extent of the law's burdens.¹²⁷ That is, without solid proof of burden in the record, Indiana's asserted justifications were deemed legally sufficient to sustain the law against the particular facial challenge that was lodged. The Court did not issue a blanket statement declaring restrictive identification laws to be legal. And it certainly did not validate ID laws as a matter of good policy.¹²⁸

The policy decision, instead, starts with the rationale for such a law. Because the Supreme Court represented such a high-profile forum, it provided the most prominent focal event to date for supporters of an identification law to justify their support by showing their rationale to be real. Crawford was a national stage for those who believe in-person impersonation fraud to be a legitimate concern to present their proof. In the case, the lower courts cited several media accounts that, the courts claimed, reflected reports of in-person impersonation fraud.¹²⁹ In the Supreme Court, respondents and amici supporting respondents added citations to more than 250 reports, encompassing decades of elections.¹³⁰

I thoroughly examined each and every one of these citations.¹³¹ The evidence of in-person impersonation fraud was strikingly sparse. The vast majority of cited reports reflected either allegations that could not possibly be related to in-person impersonation fraud and which an identification law could not possibly fix (e.g., absentee ballot problems, vote-buying schemes, or ballot tampering), or allegations that did not mention whether the alleged wrongdoing was committed in-person or through more susceptible absentee ballots.¹³² Two reports involved single votes that were the product of official pollworker misconduct or forged documentation, which also could not be prevented by laws requiring pollworkers to examine documentation.¹³³ Two reports involved unsuccessful attempts to vote in the name of another.¹³⁴

That left, since 2000, nine allegations of votes that might have involved votes cast by individuals impersonating others, which identification rules might have prevented. There is also an alternative explanation for each of the nine votes: either pollworker error or voter confusion might have caused a different, legitimate elector to sign the wrong line of the pollbook, or a data entry error might have caused an elector's voter record to register a vote for an election when no corresponding voter ever signed in at the polls.¹³⁵ There are plentiful reports of similar mistakes, with fathers confused for sons, and vice versa.¹³⁶ Investigation of the pollbooks themselves could distinguish fraud from error, but in my research to date, I have not been able to find any evidence that the necessary investigation was undertaken.

This evidence is remarkable. There have been allegations of impersonation at the polls, but they are notable for their rarity. In the most prominent forum to date for collecting such allegations, proponents of these rules cited nine votes since 2000 that were caused either by fraud that in-person identification rules could possibly stop…or by innocent mistake. During the same period, 400 million votes were cast in general elections alone.¹³⁷ Even assuming that each of the nine votes were fraudulent, that amounts to a relevant fraud rate of 0.000002 percent. Americans are struck and killed by lightning more often.¹³⁸ And every year, there are far more reports of UFO sightings.¹³⁹

Some have claimed that the incidence of alleged in-person impersonation fraud is extremely low because this type of fraud is difficult to detect.¹⁴⁰ This is distinct from the issue of whether in-person impersonation fraud is difficult to prosecute: littering clearly exists, but is difficult to address through the criminal justice system, because the wrongdoer is not easily identified. Here, not only are there virtually no prosecutions of in-person impersonation fraud, but there are even strikingly few reports of potential impersonation. It is as if individuals were complaining about littering, but could find no garbage in the street. For those believing in impersonation at the polls, the answer is that this sort of fraud is simply difficult to detect.

In truth, there are multiple means to discover in-person impersonation fraud, all of which should yield many more reports of such fraud if it actually occurred with any frequency. An individual seeking to commit in-person impersonation fraud must, at a minimum, present himself at a polling place, sign a pollbook, and swear to his identity and eligibility. There will be eyewitnesses: pollworkers and members of the community, any one of whom may personally know the individual impersonated, and recognize that the would-be voter is someone else.¹⁴¹ There will be documentary evidence: the pollbook signature can be compared, either at the time of an election or after an election, to the signature of the real voter on a registration form, and the real voter can be contacted to confirm or disavow a signature in the event of a question.¹⁴² There may be a victim: if the voter impersonated is alive but later arrives to vote, the impersonator's attempt will be discovered by the voter. (If the voter impersonated is alive and has already voted, the impersonator's attempt will be discovered by the pollworker; if the voter impersonated is deceased, it will be possible to cross-reference death records with voting records, as described above, and review the actual pollbooks to distinguish error from foul play.) If the impersonation is conducted in an attempt to influence the results of an election, it will have to be orchestrated many times over, increasing the likelihood of detection.

As in all law enforcement, none of these detection mechanisms are perfect. Yet in hundreds of millions of ballots cast, they have yielded only a handful of potential instances of in-person impersonation fraud, precisely during a period when investigating voter fraud was expressly deemed a federal law enforcement priority,¹⁴³ and when private entities were equipped and highly motivated to seek, collect, and disseminate such reports.¹⁴⁴ The phone should have been ringing off the hook, but instead there was barely a whisper.

A more logical explanation for the extraordinary rarity of reported impersonation fraud at the polls is that such fraud is extraordinarily rare. It is an extremely inefficient means to influence an election. For each act of in-person impersonation fraud in a federal election, the perpetrator risks five years in prison and a $10,000 fine under federal law, in addition to penalties under state law.¹⁴⁵ In return, the perpetrator gains at most one incremental vote. It is understandable that few individuals believe such a trade-off worthwhile.

D. Balancing costs and benefits

In weighing the costs and benefits of restrictive identification rules, the limited incidence of any fraud that these rules could prevent is significant. Because most American citizens have the identification required, the number of eligible voters without ID is relatively small. But even the most conservative estimates of impact show that the “cure” of restrictive identification is—mathematically—half a million times worse than the ostensible disease. Even if only 1.2% of registered voters do not have the required identification, burdening 1.2% of the voters in order to address a 0.000002% fraud rate betrays an alarming devotion to disproportion. Put differently, burdening more than two million registered voters to address nine potential fraudulent votes seems a particularly poorly tailored means to “safeguard” our most precious constitutional right.

It is true that, as with any election regulation, a photo identification rule could prove outcome-determinative in a close election.¹⁴⁶ Indeed, in at least one race—a tie vote for an Indiana school board seat in 2010 with one provisional ballot cast by a voter without the requisite identification—it already has.¹⁴⁷ Given the known facts about the incidence of the burden of such laws and the incidence of the rate of alleged fraud, it is substantially more likely that the Indiana school board election was resolved by unnecessarily excluding an eligible voter, and not by preventing a would-be fraudulent ballot.¹⁴⁸ The most restrictive identification laws erect a real barrier to millions of real citizens in order to increase existing protections against an unlikely hypothetical. It is like amputating a foot in order to prevent a potential hangnail.

Indeed, preliminary evidence indicates that restrictive identification rules may have already prevented more individuals from voting than any incidence of fraud to justify the impact. The evidence submitted in Crawford cited nine potentially fraudulent votes—nationwide and over seven years—that strict identification rules might have prevented. The individual stories above¹⁴⁹ represent just some of the individual stories of citizens without government-issued photo identification, more than nine of whom have already been prevented from casting valid ballots due solely to restrictive identification laws. And there are many more. In just one Indiana county, in just one off-cycle limited-turnout election in 2007, 32 voters cast ballots that could not be counted because of Indiana's new restrictive identification law; fourteen of these voters had previously voted in at least ten elections.¹⁵⁰ In the 2008 presidential primary election, approximately 321 Indiana ballots seem to have been rejected because of the identification law.¹⁵¹ In the general election, 902 Indiana ballots seem to have been rejected because of the identification law.¹⁵² Similarly, in a 2007 off-cycle Georgia election, 33 voters' ballots were rejected because of that state's new, restrictive identification law,¹⁵³ and in the 2008 presidential primary, 254 Georgian ballots were rejected because of the new law.¹⁵⁴ It is impossible to know how many other voters without the proper identification came to the polls but did not cast provisional ballots (which would not have counted without identification), or how many declined to make the trip to the polls in the first instance (which would have been futile).¹⁵⁵ And though it is theoretically possible that each of the provisional ballots listed in this paragraph represented a fraudulent vote, there is no further evidence to support that conclusion.

Despite their demonstrated impact on many American citizens, some seek to justify overly restrictive identification laws by claiming that they will at least increase public confidence in the election process. On its own, this justification for a substantial burden is open to question: there is a significant danger to democratic participation in allowing the unfounded fears of the many to burden the fundamental constitutional rights of the few.¹⁵⁶ Moreover, even if public confidence alone were a legitimate reason to exact regulatory cost, it is not clear, even in theory, that restrictive identification laws would have a unidirectional impact: such laws might increase the confidence of those who believe that the system is presently endangered by fraud or potential fraud, but might also decrease the confidence of those who believe that the system is presently designed to unnecessarily burden or exclude eligible voters.

Available data suggest empirical uncertainty to complement the theoretical concerns. In a careful 2008 study, Professors Ansolabehere and Persily could not confirm the notion that requiring particular identification will increase voter confidence; the study found no statistically significant correlation between the rate at which citizens were asked to produce photo ID and their perception that either voter fraud generally, or voter impersonation in particular, exists.¹⁵⁷ Apparently, those who are inclined to believe that elections are, by and large, secure will continue to believe that they are secure—and those who are inclined to believe that elections are, by and large, insecure will continue to believe that they are insecure—no matter what the identification regime. Restrictive identification laws do not, in short, appear to make citizens feel measurably more secure about the integrity of their elections.

Finally, in addition to the negligible benefits of the most restrictive laws requiring government-issued photo identification, it is worth noting real costs of the policy, even beyond the cost to legitimate citizens who do not have the necessary identification. Indeed, as with the registration and early-vote policies reviewed above, the new laws may well be counterproductive. For example, Georgia has dramatically limited the identification that citizens may use to vote at the polls, but also offers no-excuse absentee voting without the dramatic ID restrictions.¹⁵⁸ While the comparative freedom of absentee voting may be seen by some to mitigate the burden on voters without government-issued photo identification,¹⁵⁹ it will also predictably drive more voters into the absentee system, where fraud and coercion have been documented to be real and legitimate concerns. That is, a law ostensibly designed to reduce the incidence of fraud is likely to increase the rate at which voters utilize a system known to succumb to fraud more frequently.

There is also a monetary cost associated with restrictive identification laws, and that cost can be substantial. As the Brennan Center has documented, courts approving restrictive identification requirements have required not only that the state offer free identification cards to eligible citizens who do not otherwise have the necessary ID, but also that the state prepare an education campaign sufficient to warn the electorate that their votes will not count absent the required identification.¹⁶⁰ These requirements amount to a real fiscal impact of millions of dollars. To produce just 168,000 identification cards in Indiana, the state estimated a $1.3 million dollar cost, with additional revenue loss of $2.2 million, which exceeds the Indiana Election Division's total budget for the 2009–2010 fiscal year—even before accounting for any education costs.¹⁶¹ And a more comprehensive fiscal note in Missouri estimated the costs of a photo ID law at $6 million for the first year, with about $4 million in recurring costs.¹⁶² Moreover, increasing any restrictions at the polls—identification or otherwise—will likely lead to an increase in the number of voters needing to cast provisional ballots. These ballots must be printed, collected, and processed, all of which leads to increased cost (and increased uncertainty in the event of a close election). In tight budgetary times, these costs also weigh heavily on the ledger against unnecessarily restrictive ID requirements.

IV. Conclusion

This article reviews several new state laws impacting the voting process before and on Election Day, and finds their assessment of costs and benefits wanting. There are other new policies of concern as well, beyond the scope of this piece—including repeals of election-day registration and repeals of practices easing the restoration of civil rights for those who have been convicted. As a theoretical matter, none of the above policies make it impossible to vote. Neither did the poll tax when it was in place. But in practice, these barriers increase the burdens to eligible citizens of exercising the franchise, and appear to do with unequal impact. More disturbing, the restrictions are unnecessary and unjustified, and in some cases, even potentially counterproductive.

Courts are rightly loathe to interfere with the duly enacted policies of democratically accountable bodies. When those bodies' work product increases the burdens of democratic participation, however, it renders public accountability less effective and impairs the vitality of our representative system. Such policy is only constitutionally justified when the provisions exacting burden are sufficiently tailored to achieving sufficiently weighty regulatory interests.¹⁶³ Nor is this a standard for the courts alone: litigation may be the forum of last resort, but the constitutional mandate applies to every governmental policymaker. Only by devoting more attention to hard-nosed assessments of the costs and benefits of election regulation—and to the rationale for these regulatory decisions—can we be sure that elections are conducted with the integrity necessary to command our allegiance.