The Hunt for Noncitizen Voters

Fatma Marouf *

Over the past year, states have shown increasing angst about noncitizens registering to vote. Three states—Tennessee, Kansas, and Alabama—have passed new laws requiring documentary proof of U.S. citizenship in order to register. 1 H.B. 56, 2011 Gen. Assemb., Reg. Sess. (Ala. 2011); H.B. 2067, 2011 Sess. (Kan. 2011); S.B. 352, 107th Gen. Assemb., 2011 Sess. (Tenn. 2011). In at least nine other states (Colorado, Connecticut, Maine, Massachusetts, New Hampshire, Nevada, Oregon, South Carolina, and Texas), similar legislation was proposed but did not pass. See Wendy R. Weiser & Lawrence Norden, Brennan Ctr. for Justice, Voting Law Changes in 2012, at 17 (2011), available at Georgia passed a law in 2009 requiring proof of citizenship to register to vote, but it is not yet in effect. Id. at 16. Arizona was the first state to pass such a requirement, but the Ninth Circuit struck it down in April 2012, finding it incompatible with the National Voter Registration Act. 2 Gonzalez v. Arizona, 677 F.3d 383 (9th Cir. 2012) (en banc), cert. granted sub nom., Arizona v. Inter Tribal Council, 81 U.S.L.W. 3048 (U.S. Oct. 15, 2012) (No. 12-71). Two other states—Florida and Colorado—have waged aggressive campaigns in recent months to purge noncitizens from voter registration lists. These efforts to weed out noncitizen voters follow on the heels of legislation targeting undocumented immigrants in a number of states. Yet citizens may be more harmed by the new laws than noncitizens, especially since the number of noncitizens registering to vote has turned out to be quite small. Wrongfully targeting naturalized or minority citizens in the search for noncitizens could also have negative ramifications for society as a whole, reinforcing unconscious bias about who is a “real” American and creating subclasses of citizens who must overcome additional hurdles to exercise the right to vote.

In September, Florida announced that it had identified 207 noncitizens registered to vote (.0018% of the electorate of nearly 11.5 million registered voters); 3 Brendan Farrington & Gary Fineout, Fla. Voter-Roll Screening Yields Few Non-Citizens, Yahoo! News, Sept. 12, 2012, Colorado identified 141 noncitizens (.004% of the electorate of about 3.5 million registered voters). 4 Ivan Moreno, Gessler: 141 Illegally Registered to Vote, CBS Denver, Sept. 7, 2012, The final numbers were far smaller than the initial lists identifying thousands or hundreds of thousands of potential noncitizens on the rolls. Florida and Colorado both whittled down their lists substantially in August and early September after gaining access to a database belonging to the Department of Homeland Security (DHS) called Systemic Alien Verification for Entitlements (SAVE). Florida sued DHS for access to this database in June while simultaneously fighting a lawsuit brought by the Department of Justice alleging that the campaign to identify noncitizens violated the ninety-day “quiet period” before an election mandated by the National Voter Registration Act. DHS agreed to a settlement just a few days after the lawsuit was filed and granted Florida access.

The SAVE database was designed to verify immigration status in order to determine eligibility for various public benefits. It does not provide a list of citizens or noncitizens. Rather, it compiles over 100 million records from at least twelve different databases about individuals who have interacted with the U.S. immigration system, such as noncitizens placed in removal proceedings, people with temporary visas, lawful permanent residents, naturalized citizens, and individuals born abroad who obtained certificates of citizenship by proving that they derived U.S. citizenship from their parents.

SAVE was designed to be accessed by providing either a naturalization or citizenship certificate number, an I-94 number given upon entry with a temporary visa, or an alien identification number, which is a unique nine digit number given to certain noncitizens. SAVE searches certain databases depending on the type of documentation submitted. 5 Office of Inspector Gen., Dep’t of Homeland Sec., OIG-12-125, U.S. Citizenship and Immigration Services Systematic Alien Verification for Entitlements Program Issues 4 (2012), available at Although SAVE can also be queried using names, dates of birth, and social security numbers, DHS has stressed that it is “alien-number-driven.” 6 U.S. Gov’t Accountability Office, GAO-05-478, Elections: Additional Data Could Help State and Local Election Officials Maintain Accurate Voter Registration Lists 47 (2005), available at The use of biographical information is considered a secondary means to verify immigration status, not the primary means. 7 See Office of Inspector Gen., supra note 5, at 4. In fiscal year 2010, SAVE required additional verification involving biographical information in about 6% of cases. Id. In order to access SAVE, Florida and Colorado drew on immigration-related information provided to the Department of Motor Vehicles (DMV). They compared the names of noncitizens in DMV records to the names of registered voters. 8 Using the Systematic Alien Verification for Entitlements (SAVE) Program for Voter Eligibility Verification, Immigr. Pol’y Center (Aug. 2, 2012), Election officials then used SAVE to try to eliminate false positives by determining whether the individuals on both lists became citizens after obtaining a driver’s license.

The process utilized by Florida and Colorado to identify registered noncitizens risks disenfranchising certain groups of citizens. One especially vulnerable group consists of people who naturalized after applying for a driver’s license. In 2011, over 694,000 people became naturalized U.S. citizens, of whom 87,309 resided in Florida. 9 Yearbook of Immigration Statistics: 2011, tbl.22, Dep’t Homeland Security, (last visited Oct. 13, 2012). If the SAVE database fails to show that someone who appears to be a noncitizen based on DMV records subsequently became a citizen, that person’s voter registration would likely be challenged. Errors could happen for various reasons, such as minor discrepancies in the spelling of someone’s name (especially when transliterated from another alphabet) or in the use of hyphens, prefixes, or suffixes, inconsistent use of middle names or initials, and different formats for entering dates. 10 U.S. Dep’t of Homeland Sec., Privacy Impact Assessment for the Systematic Alien Verification for Entitlements (SAVE) Program § 1.5 (2011) (stressing that the accuracy of SAVE’s information “depends on the accuracy of the data within the databases that share data with SAVE”), available at; see also U.S. Citizenship & Immigration Servs., U.S. Dep’t Homeland Sec., Information for Applicants: Verification of Immigration Status and How to Correct Your Record with USCIS (2009) (“[T]here are a number of reasons why the SAVE rogram may not be able to verify your immigration status.”), available at; Office of Inspector Gen., supra note 5, at 6 (“USCIS can do more to assist applicants in correcting their records.”). While status verifiers within DHS work to address such discrepancies, they are not always easily resolved. 11 U.S. Dep’t of Homeland Sec., supra note 10, at § 1.5 (“If possible, status verifiers will verify the applicant’s immigration status and request an update to the database.” (emphasis added)). An individual may even have to clarify discrepencies with the DHS in person. Id. If voters are challenged mere months, weeks, or days before the election—or at the polls—they may not have time to resolve these types of disputes in order to cast a vote. Moreover, the same errors may recur year after year due to lack of oversight, as the various agencies that own the databases searched by SAVE are not required to inform the Verification Division of U.S. Citizenship and Immigration Services’ (USCIS) that an error was corrected. 12 Office of Inspector Gen., supra note 5, at 6. USCIS refused to accept the Office of the Inspector General’s recommendation that it engage in more oversight to ensure that errors are corrected. Id. at 7-8.

Another group of citizens vulnerable to being challenged are those born in the United States who happen to have the same name and birthdate as a noncitizen. In large populations, the chances of this happening are surprisingly high. 13 See Myrna Pérez, Brennan Ctr. for Justice, Voter Purges 23 (2008), available at (explaining that in a group of 23 people, there is a more than 50% chance that two will have the same birthday, while, in a group of 180 people, there is a more than 50% chance that two will have the same exact birthdate, including the year); Michael P. McDonald & Justin Levitt, Seeing Double Voting: An Extension of the Birthday Problem, 7 Election L. J. 111, 112 (2008) (“In a sufficiently large population, two entries listing the same name and birthdate are likely to demonstrate statistical coincidence rather than fraud.”). For example, in 2000 and 2002, between 400 and 1,000 New Yorkers were alleged to have voted in both New York and Florida based on flawed attempts to match names and birthdates, but only two cases have been substantiated. See Justin Levitt, Brennan Ctr. for Justice, The Truth About Voter Fraud 13 (2007), available at In 2004, 4,397 New Jersey voters were alleged to have voted twice within the state, and 6,572 were alleged to have voted once in New Jersey and once in another state, but “[m]any of these alleged double votes were actually flawed matches of names and/or birthdates on voter rolls.” Id. States have failed to specify what criteria they will use to decide that someone registered to vote is the same person as someone who appears to be a noncitizen in the DMV and SAVE databases. Addresses are unreliable to “match” people, as DHS has acknowledged that the addresses in its system are often outdated and that its own Immigration and Customs Enforcement officers don’t use them. 14 U. S. Gov’t Accountability Office, supra note 6, at 45-46. Millions of voter registrations also have incorrect addresses, further undermining the value of addresses as a means of identifying fraudulently-registered voters. The Pew Ctr. on the States, Inaccurate, Costly and Inefficient: Evidence that America’s Voter Registration System Needs an Upgrade 4 (2012), available at The last four digits of the social security number (SSN4), collected by forty states for voter registration, likewise prove unreliable, as SSN4 mismatches affected thousands of eligible voters in 2008 and led to extensive litigation. 15 The Pew Ctr. on the States, Upgrading Democracy: Improving America’s Elections by Modernizing States’ Voter Registration Systems 12 (2010), available at Until states define what matching techniques they will use and how they will ensure that the matches are accurate, the potential for wrongful disenfranchisement looms large.

Once election officials challenge a registered voter’s citizenship, the question then becomes whether that person will be able to provide proof of citizenship. This question is also critical in the three states that have passed new laws requiring proof of citizenship to register. According to a survey sponsored by the Brennan Center for Justice and performed by the independent Opinion Research Corporation in 2005, 7% of U.S. citizens surveyed do not have ready access to citizenship documents, such as a U.S. passport, birth certificate, or naturalization certificate. Based on the 2010 census, this means that about sixteen million people lack proof of their citizenship.  Those who earn under $25,000 per year and women are especially likely to lack proof of citizenship. The survey found that only 48% of voting-age women with access to their U.S. birth certificates have their current legal names on their birth certificates, while only 66% of voting-age women with access to any type of citizenship document have their current legal name on that document. 16 Brennan Ctr. for Justice, Citizens Without Proof: A Survey of Americans’ Possession of Documentary Proof of Citizenship and Photo Identification 2 (2006), available at In other words, over forty million voting-age women have no document proving citizenship with their current legal names.

Naturalized citizens may also have trouble producing proof of citizenship in some circumstances. An individual who has lost her naturalization certificate must submit a form to USCIS that takes several months to process, pay a fee of $345, and possibly appear at an interview in order to obtain a replacement certificate. Someone who changed her name after naturalizing, perhaps due to marriage or a desire to assimilate, would also have to go through this process in order to obtain a naturalization certificate that reflects her current name. That person may have no reason to go through this ordeal before her voter registration is challenged, and, after being challenged, she may not have sufficient time to go through it and obtain a replacement certificate before the election.

In addition to women and naturalized citizens, Native Americans may be uniquely affected if their citizenship is challenged or if they are legally required to present proof of citizenship in order to register to vote. While tribal IDs can be used to verify citizenship, the process of obtaining a tribal ID is often contested. Each federally recognized tribe sets its own standard for enrollment, so whether a Native American with no other proof of citizenship may obtain a tribal ID will turn on whether or not a particular tribe decides to recognize that individual as its member. Many tribes require a one-fourth “blood quantum” for enrollment, while some require less or more. Disagreements therefore arise about an individual’s degree of Indian blood. Some disputes over enrollment have also been attributed to ulterior motives, such as desires to gain or limit access to Indian gaming revenues, since profits are distributed only to tribal members. These disputed determinations about tribal enrollment, which cannot be reviewed by any court, should not play a role in determining who has the right to vote.

While some citizens may not be able to provide proof of citizenship, others whose eligibility to vote is challenged may feel intimidated upon receiving letters demanding that they remove themselves from the rolls or provide documentation, and they may choose to withdraw their registrations simply to avoid any problems with the government. Those who recently became naturalized citizens, in particular, may feel apprehensive, especially if they came from countries where harassment or violence around elections was commonplace. Documented accounts of citizens, especially Mexican-Americans, being wrongfully denied entry to the United States, placed in immigration detention, and deported from this country demonstrate that fears of being treated like noncitizens are not baseless. 17 See, e.g., Jacqueline Stevens, U.S. Government Unlawfully Detaining and Deporting U.S. Citizens as Aliens, 18 Va. J. Soc. Pol’y & L. 606 (2011).

Although it is impossible to quantify the number of citizens that risk being disenfranchised, the categories delineated above suggest that this number may far exceed the small numbers of wrongfully registered noncitizens. On a deeper level, questioning the citizenship of certain individuals—who are likely to be naturalized citizens or ethnic minorities whose names match those of noncitizens—gives power to unconscious beliefs about who is a “real” American. Studies in implicit social cognition have shown a close association between American identity and Whiteness. In fact, studies demonstrate that White, Asian, African-American, and Latino subjects all unconsciously associate being American with being racially White, even if they outwardly espouse egalitarian beliefs. 18 See Thierry Devos & Mahzarin R. Banaji, American = White?, 88 J. of Personality & Soc. Psycol. 447, 453 (2005); Thierry Devos et al., Say “Adios” to the American Dream? The Interplay Between Ethnic and National Identity Among Latino and Caucasian Americans, 16 Cultural Diversity and Ethnic Minority Psychol. 37 (2010). Such studies should give us pause to consider how the campaigns to purge noncitizens from voter rolls will affect our society: will they effectively curb voter fraud or just deepen the racial divides that are intertwined with notions of American identity?

The potential for deepening racial divides becomes clearer when the hunt for noncitizen voters is viewed in the context of other restrictive voting laws. Since 2011, nineteen such laws and two executive actions have taken effect in states that will provide more than two-thirds of the electoral votes for the November 6 election. Some of the laws require voters to show government-issued photo IDs, which 11% of U.S. citizens do not have. 19 See Brennan Ctr. for Justice, supra note 16, at 3. Some have placed new burdens on voter registration drives, through which African-American and Hispanic voters are twice as likely to register as Whites. Others restrict early voting, specifically eliminating Sunday voting, which African-Americans and Hispanics also utilize more often than Whites. In two states, new laws rolled back reforms that had restored voting rights to citizens with felony convictions, who are disproportionately African-American. Each of these laws is a stepping-stone on the path to subsidiary citizenship. 20 For a detailed discussion of the laws mentioned in this section and their impact on different racial groups, see Wendy R. Weiser & Lawrence Norden, Brennan Ctr. for Justice, Voting Law Changes in 2012 (2011), available at Rather than creating new obstacles to democratic participation, we should focus our energy on ensuring that all eligible citizens are able to exercise the fundamental right to vote.