The issue before the U.S. Supreme Court in the Crawford case (Crawford v. Marion County Election Bd. 2008) was whether a law (Indiana Senate Enrolled Act No. 483) passed by the Indiana legislature requiring most voters to show a photo ID in order to cast a ballot violates the First and Fourteenth Amendments. Plaintiffs argued that it works an unfair hardship on many people who do not have the government-issued documents that count as a legitimate ID (Indiana Democratic Party et al., 12–16). They argued that the law, in effect, constitutes a poll tax, inasmuch as there are costs to obtain the right kind of photo ID, costs that unduly burden many eligible citizens wanting to exercise their right to vote.

Given the long history of legally sanctioned disfranchisement of large and disparate groups of citizens, from the founding of the Republic to the recent past, the case raised important questions to scholars of voting rights. Indeed, Indiana's new law brought to mind events during the half-century following the Civil War,Footnote ¹ when the language of “progressive reform” cloaked the disfranchisement of blacks and poor whites in the South—those most likely to vote for Republican or Populist candidates. Actually adopted for partisan and racially discriminatory purposes, these laws were often presented as high-minded attacks on fraud—efforts to “purify” the electorate that would only inconvenience “vote sellers” or the ignorant and “shiftless” (Kousser Reference Kousser1974).

To be sure, unlike today, when proponents of voter identification must strain mightily to find the rarest examples of fraud, particularly in-person voter fraud at the polls, in the nineteenth century there was widespread and readily admitted fraud. However, this was often committed against African Americans and the Republican Party to which they then overwhelmingly adhered. Louisiana senator and former governor Samuel D. McEnery stated in 1898 that his state's 1882 election law “was intended to make it the duty of the governor to treat the law as a formality and count in the Democrats” (Kousser Reference Kousser1974, 46). A leader of the 1890 Mississippi constitutional convention admitted that “it is no secret that there has not been a full vote and a fair count in Mississippi since 1875,” which was the last year until 1967 in which blacks voted at all freely in the state (Wharton Reference Wharton1947, 206). Nonetheless, these same Democrats invoked the language of reform in calling for a wide range of restrictions on the suffrage: registration acts, poll taxes, literacy and property tests, “understanding” qualifications, and white primaries, among others.

Between 1889 and 1913, for example, nine states outside the South made the ability to read English a prerequisite for voting (Phillips Reference Phillips1906, 55–62; McCulloch Reference McCulloch1929, 54–58). Literacy tests were said to reduce the influence of immigrants or African Americans who supported “bosses” and “demagogues” (McCook Reference McCook1892, 171–76; Commons Reference Commons1920, 183, 195). Moreover, between 1890 and 1908, seven of the 11 ex-Confederate states adopted state constitutional amendments allowing only literate voters or those with a certain amount of property to vote. There were sometimes loopholes like “understanding” qualifications or “grandfather” clauses that allowed some whites to vote who could not meet literacy or property tests. Shortly after passage of these amendments, less than 10% of African Americans managed to register to vote in most states, and no more than 15% in any (Kousser Reference Kousser1974, 60–61).

The poll tax was one of the most notorious disfranchising mechanisms of its day. The current debate over the Indiana photo-ID requirement—as well as similar laws in other states—has led to claims that they are a “modern-day poll tax.” This implies that the new Indiana law, too, falls within the ignominious American tradition of disfranchising laws passed under the guise of “good government” reform.

Frederick Ogden, perhaps the foremost scholar of the poll tax, wrote in the 1950s: “While critics of legalized restrictions on Negro voting may find it hard to discover any high moral tone in such activities, these restrictions reflected a movement for purifying the electoral process in southern states” (Ogden Reference Ogden1958, 7).Ogden quotes the editor of the San Antonio Express writing in 1902: “By requiring a poll tax receipt, secured six months previous to an election, fraudulent elections can be prevented almost entirely” (Ogden Reference Ogden1958, 9–10).

Other essays in this symposium address the nature and extent of burdens imposed on various subsets of Indiana citizens by the photo-ID law, and I shall forgo discussing them. Suffice it to say that the most accessible photo-ID in Indiana consists either of the state's driver's license or a state-issued ID card. Obtaining one or the other has been shown to be a good deal more difficult for some people than it might seem at first glance. At least 43,000 persons of voting age in Indiana are estimated to have neither (Indiana Democratic Party et al., 12–17).

The demographic characteristics of persons lacking the requisite ID are suggested by a November 2006 telephone survey of 987 randomly selected voting-age American citizens by the independent Opinion Research Corporation conducted for the Brennan Center for Justice at NYU School of Law: 11% did not have valid government-issued photo ID, while 18% of citizens 65 years of age or older lacked it, as did 25% of African Americans (Brennan Center for Justice 2006, 3).Footnote ² The latter two demographic groups, the elderly and African Americans, are more likely to self-identify as Democrats—African-Americans disproportionately so.Footnote ³ Elderly African Americans, who are even more unlikely than members of their ethnic group in general to have a photo ID, would be strongly predisposed to vote Democratic. In close elections, the additional burdens placed on both the elderly and African Americans by the photo-ID law could help elect Republican candidates. There is no reason to believe this national pattern is much different from that in Indiana.

Nonetheless, it is often asserted that such barriers should not prevent a truly motivated citizen from voting. In a classic article by Kelley, Ayres, and Bowen (Reference Kelley, Ayres and Bowen1967, 359, 375) attempting to measure determinants of voter turnout, the authors make the following observation:

Their question, posed in terms of the burden of time alone, can also be posed with regard to money: particularly concerning the least well off, how large a monetary imposition should be placed on the right to vote before it becomes the functional equivalent of a poll tax? Regarding these twin burdens, two questions may be posed to help determine whether the Indiana photo-ID law should be interpreted according to the good-government language of its proponents: First, how will the application of the law help shape the Indiana electorate “to a size and composition deemed desirable by those in power,” as Kelley, Ayres, and Bowen put it? In other words, to what extent is the law motivated by partisan efforts to disfranchise voters who are undesirable to Republicans and thus increase their chances of winning elections? Second, did supporters of the law demonstrate a significant degree of fraud of the kind the law was fashioned to prevent? Let us consider each question in turn.

While it is impossible to know the motives of those lawmakers who favored the photo-ID bill under consideration by the Indiana legislature in April 2005, we can ascertain whether it was passed by a partisan vote. Significantly, Indiana's photo-ID bill was one of at least 10 bills introduced by Republicans in state legislatures between 2005 and 2007 requiring voters to show a photo ID at the polls. Two of these states' bills were initially enjoined, and a second bill was introduced in one state. (Besides Indiana, the other states included Georgia, Florida, Missouri, Kansas, New Hampshire, Pennsylvania, Texas, and Wisconsin.) If the House and Senate votes for all 10 proposals are combined, 95.3% of the 1,222 Republicans voting and 2.1% of the 796 Democrats voting supported the bills. Moreover, in the five cases in which both houses passed a bill and a Republican was governor, he signed it. In the three cases in which both houses passed a bill and a Democrat was governor, he vetoed it. (In two cases, only one house passed a bill.)Footnote ⁴ The Indiana vote was part of this pattern, although even more extreme. In the vote on Senate Bill 483, 85 Republicans voted for it and none against; 62 Democrats voted against it and none in favor. The Republican governor signed the bill into law (Indiana Democratic Party et al. 2007, 9).

Did supporters of the law demonstrate that there is a significant degree of fraud of the kind the law was fashioned to prevent? The debate over the extent and kind of vote fraud that exists in the United States today has been widespread and acerbic at least since the 2000 presidential election, and it shows no signs of abating. There are numerous kinds of vote fraud, and distinctions among them—which are necessary to determine the most effective means of their prevention—are often lost in popular debate. As critics of the Indiana law have asserted, the photo-ID requirement was implemented to prevent one type of fraud: voter impersonation at the polls on Election Day (Indiana Democratic Party et al. 2007, 42). Among the many kinds it does not prevent is that involving mail-in ballots, which some believe to be more common than impersonation at the polls (Indiana Democratic Party et al. 2007, 45–47). What makes the statute particularly suspect in the case of Indiana is the fact that there has not been a single prosecution for in-person vote fraud in the history of the state—a fact that Richard Posner, judge on the U.S. Court of Appeals for the Seventh Circuit and author of that court's split decision favoring Indiana, attributed to lax law enforcement (William Crawford v. Marion County Election Board 2007, 7).

Recent events in Texas are relevant in this regard. In both 2005 and 2007 Republicans in the legislature introduced photo-ID bills less restrictive than that in Indiana (Vertuno Reference Vertuno2005, 8A; Economist 2007). In 2007, according to a newspaper reporter: “Republicans like the voter ID bill because they believe it will weaken Democrats, but can argue that it is a reasonable requirement” because it would prevent vote fraud (Mack Reference Mack2007). Not all Republicans, however, shared the belief that it would curtail fraud. Royal Masset, former political director of the Texas Republican Party, was one. He told the reporter he agreed that among his fellow Republicans it was “an article of religious faith that voter fraud is causing us to lose elections.” He was not convinced. He did believe, however, that requiring photo IDs could cause enough of a drop-off in legitimate Democratic voting to add 3% to the Republican vote (Mack Reference Mack2007).

In January 2006, after his party's first failure to pass a photo-ID bill, Greg Abbott, the Republican attorney general of Texas, announced a “training initiative to identify, prosecute [and] prevent voter fraud” (Attorney General 2006). This was the most ambitious and costly effort in recent Texas history—perhaps ever—by the state's government to attack the alleged problem. “Vote fraud has been an epidemic in Texas for years, but it hasn't been treated like one,” Abbott said. “It's time for that to change” (Attorney General 2006). He promised that his newly created Special Investigations Unit (SIU) would “help police departments, sheriff's offices, and district and county attorneys successfully identify, investigate and prosecute various types of voter fraud offences.” Established with a $1.5 million grant from the governor's office, the SIU would have as one of its prime responsibilities investigating voter-fraud allegations, he said. Abbott targeted 44 counties containing 78% of registered voters in the state (Attorney General 2006). According to the Austin American-Statesman, “Complaints originate from voting officials, district attorneys or citizens and are sent to the secretary of state or the attorney general. Each complaint is evaluated by a professional employee to determine whether the complaint is legitimate and warrants further investigation” (Editorial Board 2006).

Such an initiative would seem to constitute a model of the aggressive, responsible, multi-level law-enforcement effort that Judge Posner seemed to believe had been lacking in Indiana. Moreover, given Republicans' desire to provide evidence of widespread voter fraud in order to justify new statutes criticized by Democrats and some media sources, one would expect Abbott to have conducted the effort with enthusiasm. What has been the result?

Texas is a large state, with thousands of elections occurring in a four-year period in its numerous governmental units. In 2006, there were 16.6 million persons of voting age, and of those, 13.1 million were registered to vote (Texas Secretary of State 2007). An anti-immigration organization estimated that 1.7 million Texas inhabitants resided there illegally in 2007 (FAIR 2007). Given these facts, one would expect an aggressive, centralized vote-fraud initiative by the state's highest law-enforcement officer to yield a sizable number of indictments during the more than 21 months of its existence if, in fact, vote fraud had reached “epidemic” proportions.

The data presented by Attorney General Abbot on his Web site told a different story. In the almost two years between the day the initiative was announced in late January 2006 and October 2007, 13 persons had either been indicted, found guilty, or sentenced for vote fraud, six on misdemeanor counts typically involving helping others with mail-in ballots. Of the 13, five were accused of having committed fraud before 2006, the year the initiative was announced, and the remaining eight in 2006. A total of 4.4 million Texans voted in the general elections for governor or U.S. senator that year, in addition to those who voted in primaries and local nonpartisan elections (Attorney General 2006). At that point, six of the 13 persons mentioned above had not yet been found guilty. This, then, is the extent of vote fraud in Texas that has been uncovered after the announcement and implementation of the $1.5 million vote-fraud initiative.Footnote ⁵ Moreover, of the seven found guilty and the six remaining under indictment, none of the types of fraud they had been charged with would have been prevented by the photo-ID requirement advocated by Republicans in the 2007 legislative session. That is to say, none involved voter impersonation at the polls. Most either involved political officials who were charged with engaging in illegal efforts to affect the election outcome, or persons who helped elderly or disabled friends with their mail-in ballots, apparently unaware of a law passed in 2003 requiring them to sign the envelope containing the friend's ballot before mailing it.

These data do not appear to be anomalous. A survey of the director or deputy director of all 88 Ohio boards of election in June 2005 found that a total of only four votes cast in the state's general elections in 2002 and 2004 (in which over nine million votes were cast) were judged ineligible and thus likely constituted actual voter fraud. (COHHIO/LWVO 2005) Interviews by New York Times reporters with election-law-enforcement officials and academic experts suggest that the pattern in Ohio is not anomalous. Professor Richard L. Hasen, an election law expert at Loyola Law School, summed up knowledgeable opinion about vote fraud to the reporters as follows: “what we see is isolated, small-scale activities that often have not shown any kind of criminal intent” (Lipton and Urbina Reference Lipton and Urbina2007).

While it is possible, as Judge Posner implied in his Seventh Circuit decision, that aggressive vote-fraud enforcement in Indiana might uncover its existence in the state, the Texas investigation suggests otherwise, and the burden of proof rests on those who allege that vote fraud there is widespread and of the kind that is deterred by a photo-ID requirement. Until that burden is responsibly shouldered by state authorities, the question of whether the Indiana voter-ID law has accomplished its ostensible purpose must be answered in the negative. When this conclusion is placed alongside our earlier findings that the legislative vote for the law was strictly along partisan lines and that the people most likely to be disfranchised by it are Democratic voters—particularly African Americans—Indiana's law appears to fit comfortably within the long and unsavory history of those in positions of power disfranchising blacks and less-well-off whites for partisan gain. Moreover, Indiana's attempt to justify its new law with claims of voter fraud is as dubious as those that justified the now unconstitutional poll tax.

A great deal of progress has been made over the past 50 years in combating racial discrimination in politics, thanks in part to such epochal events as passage of the Twenty-Fourth Amendment and the Voting Rights Act. However, race, class, and partisanship continue to be inextricably intertwined in the United States, just as they were from the end of Reconstruction to the Civil Rights Era. The 2005 Indiana voter-ID law, if the above analysis is correct, is an excellent example of this fact. ■