Employment and Training Institute banner


Research Update

ETI Research on Disparate Racial Impacts of Using Driver's Licenses for Voter IDs

graph from ETI report A research report by the University of Wisconsin-Milwaukee Employment and Training Institute on The Driver License Status of the Voting Age Population in Wisconsin provides research assisting public officials and the courts assess the disparate impacts of state and local laws requiring photo IDs as a condition for voting and the Supreme Court examines challenges to the photo ID voter law.

In May 2014 federal Judge Lynn Adelman found Wisconsin's state photo ID law unconstitutional given its adverse impact on many Wisconsin citizens. The 90-page decision is posted online. In it, in note 32 Adelman cites the ETI research that only 47% of black adults and 43% of Hispanic adults compared to 73% of white adults in Milwaukee County hold valid driver's licenses as do 85% of white adults in the rest of Wisconsin compared to 53% of black adults and 52% of Hispanic adults. In October 2014 a three-judge federal appeals court panel found the law constitutional based on the Supreme Court Indiana decision. Here, in the court's decision Judge Frank Easterbrook referenced the 2005 ETI data but suggested that it was evidence that fewer nonwhites without licenses have registered to vote (putting aside the "felon-disenfranchisement" issue). Subsequently, the full 10-member panel deadlocked 5-5 on rehearing the case. On October 9, 2014, the U.S. Supreme Court voted 6-3 to block the law's implementation for the immediate Wisconsin elections scheduled in November.

The Employment and Training Institute study was the first research available that measured driver's license disparities by race and age. The ETI was able to measure possession of driver's licenses for subpopulations in Wisconsin, having reviewed the state license files for employment-related research, and particularly for lack of licenses among working age African Americans and Latinos in Milwaukee County. The ETI research

  • was cited by Circuit Court Judge David Flanagan as one basis for his temporary injunction against Wisconsin's new voter identification law. Flanagan quoted statistics from the ETI report showing the relatively low percentages of minorities possessing a driver's license: "In Wisconsin 45 percent of African American males and 51 percent of females possess a license. As to Hispanics, 54 percent of males and 41 percent of females have a Wisconsin license." He also noted that many elderly persons aged 65 and older do not possess a drivers license or DMV-issued photo ID. The injunction was issued on March 6, 2012, and based guarantees to the right to vote included in the Wisconsin state constitution. Flanagan later issued a permanent injunction against the law. A permanent injunction was also issued against Wisconsin's voter ID law by Circuit Judge Richard Niess on March 12. Niess's order maintains that the state legislature lacks the authority under Wisconsin's constitution to bar citizens from voting on the grounds that they lack a government-sanctioned photo ID. The lack of a driver's license remains one of the most serious barriers to employment for many workers in Wisconsin, and particularly for African Americans and Latinos in Milwaukee County.
  • was used by a New York Review of Books essay identifying possible disparate impacts on African Americans of voter ID policies relying on the driver's license as the primary government ID.
  • was cited by Supreme Court Justice David Souter to document the disproportionate impact of voter ID laws on racial minorities in his dissenting opinion in the Indiana voter ID case (Crawford v. Marion County Election Board), page 13, before the U.S. Supreme Court.
  • was used in the NAACP Milwaukee Branch and Milwaukee Teachers' Education Association amicus brief in support of dismissing the Wisconsin Attorney General's September 2008 mandamus petition to require driver's license matches of voter lists prior to the November election.
  • was at the heart of disagreements between career staff and political appointees in the U.S. Department of Justice regarding preclearance of Georgia's voter ID law.
  • was analyzed to estimate disenfranchisement impacts of proposed federal legislation to use the photo ID as a condition for voting.
  • was referenced by Governor Jim Doyle as the basis for vetoing photo ID legislation in Wisconsin.
  • was cited by the dissenting justice in the Michigan Supreme Court to estimate the number of Michigan citizens impacted by Michigan's photo ID law.
  • is recommended in law review articles as a standard for empirical research required to determine the constitutionality of state voter ID laws.
  • has been referenced by over 4,000 website discussions of the voter ID, including, for example, the AARP, university student newspapers, and the New Yorker, among others.
The ETI report on The Driver License Status of the Voting Age Population in Wisconsin issued in June 2005 examined the following:

  • Percentages of voting age Wisconsin adults without valid driver's licenses in 2002 compared to the 2000 Census for similar age cohorts. Statewide, 59% of Hispanic females, 55% of African American males, 49% of African American females, and 46% of Hispanic males, compared to 17% of white males and 17% of white females, were without a valid driver's license.

  • For young adults ages 18-24, 78% of African American males and 66% of African American females in Wisconsin were without a valid driver's license.

  • The percentage (93%) of college students living in Marquette University, University of Wisconsin-Madison, and University of Wisconsin-Milwaukee residence halls without a driver's license for those addresses.

  • The percentage (82%) of 18, 19, and 20 year olds without a driver's license at their ZIP code for college student-intensive neighborhoods in Eau Claire, LaCrosse, Madison, Milwaukee, Oshkosh, Platteville, River Falls, Stevens Point, Stout, and Whitewater, Wisconsin.

  • Residents aged 65 and older without a driver's license or state photo ID (estimated at 177,399) based on census counts of the population and unduplicated counts of licenses and state IDs. Another estimated 98,247 state residents ages 35 through 64 do not have a driver's license or state photo ID.

  • Census counts of voting age adults without a vehicle in their household by race and location.

  • Racial differences in mobility rates (i.e., those changing addresses within the last five years) among renters and homeowners.

ETI Research Used in Souter's Dissent in Supreme Court Decision on Indiana Voter ID Law

In the Crawford v. Marion County Election Board and Indiana Democratic Party v. Todd Rokita, decided by the U.S. Supreme Court on April 28, 2008, Justice David Souter wrote, "The upshot is this. Tens of thousands of voting-age residents lack the necessary photo identification. A large proportion of them are likely to be in bad shape economically...." Souter cited the ETI research in Wisconsin, noting "Studies in other States suggest that the burdens of an ID requirement may also fall disproportionately upon racial minorities" -- with Souter including the "June 2005 study by the Employment and Training Institute at the University of Wisconsin-Milwaukee, which found that while 17% of voting age whites lacked a valid driver's license, 55% of black males and 49% of black females were unlicensed, and 46% of Latino males and 59% of Latino females were similarly unlicensed." [Souter dissenting opinion at page 13].

The petitioners' brief to the Supreme Court had cited the Employment and Training Institute research on percentages of elderly in Wisconsin without a driver's license or photo ID (23%) and the low percentages of African American and Hispanic adults in Milwaukee with a driver's license (47% and 43%, respectively) to challenge the 7th Circuit Court ruling. The attorneys argue, "The court of appeals also erred in concluding that the challenged law has little effect on the electorate....This is not just a problem that is limited to the homeless and displaced. Citizens with lower incomes are less likely to possess photo identification, and minority populations disproportionately lack identification." (pp. 39-40)

In his majority opinion for the 7th Circuit Court, Judge Richard Posner had written that it is "exceedingly difficult to maneuver in today's America without a photo ID ... and as a consequence, the majority of adults have such identification." Posner noted that, "No doubt most people who don't have photo ID are low on the economic ladder," but assumed that only a small number would be affected, writing, "The fewer the people who will actually disfranchise themselves rather than go to the bother and, if they are not indigent and don't have their birth certificate and so must order a copy and pay a fee, the expense of obtaining a photo ID, the less of a showing the state need make to justify the law." In his dissenting opinion Judge Terence Evans noted that the group without driver's license or passport IDs "is mostly comprised of people who are poor, elderly, minorities, disabled, or some combination thereof." (page 13)

In a September 2008 New York Review of Books essay on Obama: The Price of Being Black, political scientist Andrew Hacker uses the ETI research to gauge the impact of the Indiana court decision on black Americans seeking to vote in the 2008 presidential election. Hacker estimates that 44,000 black residents in Indianapolis alone could need transportation to the DMV to complete a nondriver ID application allowing them to vote.

Career Staff in Justice Department Called for ETI-Type Analysis in Georgia

In their August 25, 2005 review of the proposed Georgia voter registration procedures, the career staff in the U.S. Department of Justice Civil Rights Division noted that the disparity ratio between whites and blacks lacking vehicle access in Georgia households was similar to that for Wisconsin and warned that the ETI data on disparities between blacks and whites on possession of a driver's license also likely applied. Staff noted, "This data suggests that complete records, or at least a more representative sample from Georgia would be expected to yield a stronger correlation between driver's license ownership and race. As this [ETI] study shows strong patterns of racial disparity among driver's license ownership in Wisconsin, it further underscores our concerns about the reliability of the Georgia DDS data, and suggests that predictions of driver's license ownership may be better analogized from vehicle access data ... As it is logical to infer that the relationship between owning a car and having a driver's license are similar in the two states, and the ratio of black to white households without vehicles are similar in Georgia and Wisconsin, an inference that a racial gap exists in driver's license ownership is appropriate." (p. 27)

The staff concluded: "While no single piece of data confirms that blacks will [be] disparately impacted compared to whites, the totality of the evidence points to that conclusion ... [I]t appears that neither the legislature nor the submitting authority conducted any analysis or presented any data regarding these racial disparities in access to various forms of photo identification. This leads us to conclude that the state has failed to meet its burden of demonstrating that the change is not retrogressive." (p.31)

The recommendations of the staff were rejected, and the following day Justice Department officials gave preclearance to the Georgia voter photo identification law as being in compliance with Section 5 of the Voting Rights Act. U.S. Assistant Attorney General William Moschella later informed the Senate that, "[A]ll individual data indicates that the State's African-American citizens are, if anything, slightly more likely than white citizens to possess one of the necessary forms of identification." He further stated, "The reality in Georgia is that the number of eligible voters who currently hold no photo identification from the Department of Driver Services ("DDS") is extremely small."

It appears from statements issued by Justice Department officials that the unmeasured disparate impact of driver's license possession on African Americans was considered to be offset by a second disparate impact on elderly residents of Georgia who, according to John Tanner, Voting Rights Section chief, are predominantly white. Tanner subsequently left this position for another post within the Justice Department.

Congressmen Cite ETI Research in Call for Ban on Photo ID Requirements for Voting

On November 1, 2007, Congressmen Keith Ellison of Minnesota, John Lewis of Georgia, and John Conyers of Michigan introduced legislation to ban photo ID requirements for voting. Ellison cited the ETI research as evidence of the possible disparate impact of photo IDs on African Americans, Latinos and elderly voters.

Michigan State Supreme Court Reviewed ETI Research

In July 2007 the Michigan state Supreme Court majority issued an advisory opinion upholding the constitutionality of the state voter ID law passed in 2005. The majority opinion dismissed the state attorney general's observation that the laws would limit access to the ballot for "poor, racial and ethnic minorities, elderly, and the disabled" by arguing that voters without photo IDs could use an affidavit process to vote. (p. 20) In his dissent Michael F. Cavanaugh cited the ETI research as the basis for his estimate that "hundreds of thousands of Michigan citizens will be affected by this legislation ... [indicating] that the requirement is a serious impediment on the fundamental right to vote for these citizens." (p.25)

Law Review Articles Call for More Empirical Work Like the ETI Research

In a law review article on "Voter Identification" in the Michigan Law Review (February 2007), Spencer Overton argues that, "Existing data [including the ETI report] suggest that the number of legitimate voters who would fail to bring photo identification to the polls is several times higher than the number of fraudulent voters, and that a photo-identification requirement would preclude political outcomes that are less reflective of the electorate as a whole." (p.1) After citing ETI research on percentages of African American men ages 18-24 without a driver's license, college students in dormitories without a license at that address, and racial differences in renters and homeowners who move, Overton called for "more detailed empirical work ... to determine the extent to which a photo-identification requirement will shape the electorate." (p. 662) Overton, who served on the Commission on Federal Election Reform, argues that judges "should demand statistical data to ensure that voter identification procedures are appropriately tailored to deter fraudulent voters rather than legitimate ones and do not disproportionately exclude protected classes of voters." (p.1)

In a second law review article entitled "If It's Broke, Fix It: Improving Voting Rights Act Preclearance" and published in the Howard Law Review (2006), Daniel P. Tokaji describes the ETI research as "the most detailed state-level study of who lacks a driver's license" and further observes: "Unfortunately, studies of driver's license possessions have not as yet been conducted for Georgia or other states." (p. 814) Tokaji's research assistant attempted unsuccessfully to obtain comparable data from the Ohio Registry of Motor Vehicles on the race of driver's license holders in that state. (Note 184 on p. 814) Tokaji recommends creation of a bipartisan Voting Rights Enforcement Commission as an alternative to the Department of Justice staff for determining preclearance for voting requirements in states subject to Section 5 of the Voting Rights Act.

Other law review articles citing the Pawasarat voter ID study include Gilda Daniels, Gilda Daniels, "A Vote Delayed is a Vote Denied: A Preemptive Approach to Eliminating Election Administration Legislation that Disenfranchises Unwanted Voters," in the Brandeis Law Journal (forthcoming); Michael J. Pitts, "Empirically Assessing the Impact of Photo Identification at the Polls Through an Examination of Provisional Balloting," in the Journal of Law and Politics (forthcoming); and Daniel P. Tokaji, "Leave It to the Lower Courts: On Judicial Intervention in Election Administration," in the Ohio State Law Journal, 2007.

Wisconsin Governor Vetoed Photo ID Legislation Based on ETI Data

In his veto of legislation to require photo IDs to vote in Wisconsin, Governor Jim Doyle cited the ETI research. "I am vetoing SB-42 for the same reasons that I vetoed AB 63: it unfairly restricts the right to vote at the expense of far too many of Wisconsin's law-abiding, elderly citizens. Two months ago, the University of Wisconsin-Milwaukee released a study showing that over 177,000 elderly persons in Wisconsin aged 65 and older do not possess a driver's license or state photo identification. Thus, under this proposal, nearly one-quarter of Wisconsin's elderly population could be disenfranchised. I cannot allow that to happen." Doyle introduced alternative legislation, without the photo ID requirement, to improve the election processes, but those proposals were not taken up by the Legislature.

In May 2011, legislation requiring voters to show photo identification at the polls was passed and signed by Governor Scott Walker, Doyle's successor.

ETI Research Is Debated by Proponents and Opponents of Photo ID Laws

David B. Muhlhausen and Keri Weber Sikich in a Heritage Foundation paper on "New Analysis Shows Voter Identification Laws Do Not Reduce Turnout" cite the ETI research findings of disparate rates of driver's license possession for African Americans and Hispanics compared to whites and findings that the elderly are less likely to have driver's licenses or other government-issued identification. The authors take the position that "Although these figures shed light on the types of people who are less likely to have driver's licenses ... the data still cannot tell us whether those individuals without driver's licenses have some other form of identification . . . [and] cannot tell us about future behavior." (p.5)

In an unfortunately shoddy attack on the ETI research, a July 2011 Heritage Foundation article by Hans A. von Spakovsky reported in a footnote that Pawasarat's study had been criticized in an article by the Center for Democracy & Election Management at American University for overstating "the percentage of residents without a driver's license by oversampling African-Americans and low-income people and failing to adjust their estimates." We were puzzled by this criticism since the Pawasarat study does not involving sampling but is a review of the entire driver's license population in the state using DMV records. It turns out that the CDEM report, entitled Voter IDs Are Not the Problem: A Survey of Three States, doesn't accuse Pawasarat of oversampling African Americans or low-income people or of failing to adjust estimates of these population. Instead, it reports that someone else criticized Pawasarat's research "for overstating the percentage of residents without a driver's license." The source of this "criticism" according to the CDEM, was a newspaper article that covered Pawasarat's just released study and included a comment from a faculty professor saying he thought the study overstated the number of people lacking an ID. The commenter made no reference to African Americans, low-income people, or oversampling. (Nor did the article indicate whether the commenter had actually read the newly released Pawasarat study.)

A policy brief by Project Vote on "Restrictive Voter Identification Requirements" raises concerns about Voter ID laws causing persons to be turned away from the polls for driver's licenses which do not contain the voter's current address. The policy paper references as concerns the ETI research on college students who do not have their current address on their photo ID and the disparities in driver's license possession between whites and African Americans.

A monograph by John R. Lott on "Evidence of Voter Fraud and the Impact that Regulation to Reduce Fraud Have on Voter Participation Rates" (cited in the Journal of the Federalist Society Practice Groups) criticizes the ETI and National Commission on Electoral Reform research as providing "only a very crude measure of whether photo ID requirements will prevent people from voting." Lott continues, "Some people without driver's licenses will not vote even when there are no photo ID requirements and others will go out to get a photo ID in order to vote. Just because they don't have a photo ID at some point in time (when they may not have any reason to have such an ID), doesn't imply that they won't get one when they have a good reason to do so." (p. 3)

A commentary in Mother Jones by Tova Andrea Wang and Jonah H. Goldman on "Photo ID Requirement Compromises Voter Rights" used the ETI study's statistics on racial disparities to bolster the position that, "The citizens who do not have photo identification are mostly poor, minority, elderly, disabled and young voters. This is largely because these voters cannot afford the requisite fees for obtaining photo IDs, and they are less likely than other voters to own a car or regularly drive. Case in point, a June 2005 University of Wisconsin study ..." The authors call for improvements in voter information, training of pool workers, location of polling places, security of voting systems, and investigation and prosecution of fraud rather than denying citizens without a photo ID access to vote.

The Politico website posted an article by Ben Adler on "Campus Voting Access Not Making the Grade," which raised concerns about Voter ID laws effect on young people. Concern was raised about the number of African Americans (74%) and Hispanics (66%) aged 18-24 without a valid driver's license and students without their current school address on their photo ID.

The University of Texas at Austin student government referenced the ETI report in a proposed resolution in opposition to photo requirements for voting.

The "ID Divide": The Economics of Government Identification Programs

A prescient February 1994 article in WIRED Magazine explored the use of the DMV (state department of motor vehicles) as a collection agency for unpaid parking tickets, municipal fines and civil forfeitures. State DMV officials in Massachusetts, Oregon and Wisconsin reported high rates of payments for drivers threatened with loss of their license. The article also explored the growing use of the driver's license as an control instrument, particularly for teenagers, citing examples of Kentucky's 1990 "high school dropout law" (which prohibited high school students from obtaining or keeping a driver's license if they had 9 or more unexcused absences or failed to pass 4 of their courses), Oregon's law preventing teens with underage drinking or illegal drug use violations from obtaining a license until age 17 (or age 18 if the violation occurred during the teen's seventeenth year), and Wisconsin's withholding of driving privileges for non-payment of any municipal fine (including violations unrelated to driving).

The current impact of such laws can be seen in Milwaukee County, where the experiences of teens prohibited from obtaining licenses were tracked for the Center for Driver's License Recovery & Employability. In 2000, the innercity areas of the City of Milwaukee showed the highest number of unlicensed teens with a "trap" on their future licenses due to suspensions and revocations (many due to failure to pay fines unrelated to driving, such as curfew violations, jaywalking, underage drinking) and the lowest rate of recovery from such traps. A total of 3,389 youth ages 15 through 17 had suspensions and revocations entered into the DMV system but no license, and only 16% of these teens subsequently obtained a license as young adults, according to the April 2008 DMV file. By contrast, the more affluent suburban areas of the county showed the lowest number of teens with unpaid fines resulting in suspensions and revocations. Even in the Milwaukee County suburbs, however, less than half of teens recovered from such pre-license traps and secured their license by 2008.

A 2008 study on The ID Divide: Addressing the Challenge of Identification and Authentication in American Society by Peter Swire and Cassandra Butts uses the ETI voter ID study to help identify groups that disproportionately lack IDs in today's society. The report warns that "those on the wrong side of the ID Divide are finding themselves squeezed out of many parts of daily life, including finding a job, opening a bank account, flying on an airplane, and even exercising their right to vote." (p.1)

Background on the ETI Driver's License Research

The University of Wisconsin-Milwaukee Employment and Training Institute has been researching and mapping driver's license issues for the last two decades as access to a valid driver's license is one of the most important employment issues facing central city Milwaukee residents. (See Removing Transportation Barriers to Employment: Assessing Driver's License and Vehicle Ownership Patterns of Low-Income Populations, 1998, and Removing Transportation Barriers to Employment: The Impact of Driver's License Suspension Policies on Milwaukee County Teens, 2000.) The economic importance of a driver's license cannot be overstated in Milwaukee's labor market, where three-fourths of area job openings are located in the Milwaukee County suburbs and suburban/exurban areas of Waukesha, Ozaukee and Washington counties -- areas not usually accessible by public transportation.

A number of Employment and Training Institute technical assistance projects have observed that possession of a valid driver's license and access to a vehicle are strong predictors of better success in the labor market. For Milwaukee women with children on public assistance, the driver's license has been shown to be more critical than even a high school diploma in predicting steady employment. A 2008 study of Milwaukee County clients assisted by the Center for Driver's License Recovery & Employability found early gains of $253 a month in earnings for women obtaining their driving privileges. In a recent study identifying employment barriers for Workforce Investment Act clients, program participants with a valid driver's license were 2 times more likely to show earnings above the family poverty level than those without a license. For WIA participants with less than 12 years of education, workers with a valid license were 4 times more likely to show post-program earnings above the poverty level than those without a license.

The Milwaukee Journal Sentinel editorial board has made a series of recommendations for improving the situation where thousands of Milwaukee County workers have suspended or revoked licenses.

The Employment and Training Institute driver's license/voter ID study, which examined individual driver's license records and state ID files from the State of Wisconsin, found large subpopulations of residents likely to face problems voting under various Voter ID laws proposed in the state. This included:

  • elderly persons aged 65 and older with neither a driver's license nor a state ID,
  • African Americans and Latinos without current driver's licenses,nse and residents using mass transit as their primary mode of transportation, and
  • students (living in 10 Wisconsin communities with large college populations) who do not have a driver's license with their current address on it.
Employment and Training Institute, University of Wisconsin-Milwaukee, 161 W. Wisconsin Avenue, Suite 6000, Milwaukee, WI 53203, email eti@uwm.edu.

ETI Home Page | Publications | Site Map

Employment and Training Institute
School of Continuing Education
University of Wisconsin-Milwaukee