SSN Key Findings

How Surveys Can Strengthen the Voting Rights Act

Policy field

Connect with the author

University of Colorado Boulder

In 2013 the U.S. Supreme Court gutted one of the most effective statutes ever enacted. The Voting Rights Act of 1965 had protected the voting rights of racial minorities in ways never before accomplished under the Fifteenth Amendment ratified in 1870. The Voting Rights Act did more than just allow victims to bring legal actions after they experienced voting discrimination. Instead, it worked to prevent discrimination before it happened. When drafting the act, Congress recognized that many jurisdictions, particularly in the Jim Crow South, were consistently losing voting rights cases in the courts, but then proceeded to devise and adopt more creative discriminatory election rules. The Voting Rights Act required these repeat offenders to get pre-clearance from the federal government before implementing any new voting rules that might deny or restrict the right to vote on account of race or color. Jurisdictions had to show that changes to their election laws would not make racial minority groups worse off. Many avoided pre-clearance altogether by not proposing changes.

In 2006, Congress reauthorized the Voting Rights Act and extended the pre-clearance provision for 25 more years. The evidence compiled by both the House and Senate showed that voting rights had improved for racial minorities in the South since 1965, but the covered jurisdictions still exhibited disproportionate amounts of voter intimidation and violence, discriminatory election practices, disparate treatment in voter registration, racial gerrymandering, and minority vote dilution. Nevertheless, in the 2013 case Shelby County v. Holder, the Supreme Court held that the process by which Congress identified who was covered by pre-clearance was unconstitutional, because it used a formula referring to election practices and Census data from 1965, 1970, and 1975. As Chief Justice Roberts noted, “history did not end in 1965.” But of course the relevant question was whether current conditions continued to justify pre-clearance for jurisdictions with a history of discrimination in voting.

Is the Hobbled Voting Rights Act Effective?

Not surprisingly, once the Court removed the pre-clearance requirement, previously covered jurisdictions responded immediately. For example, on the same day Shelby County was decided, Texas announced that it would implement its voter ID law that had previously been denied pre-clearance (a law has since been ruled unconstitutional via subsequent litigation). Two months later, North Carolina enacted a bill that eliminated same-day voter registration, shortened early voting periods, and imposed strict ID requirements in a way that the Fourth Circuit would later say targeted would-be African American voters “with almost surgical precision.” Local governments freed from pre-clearance also made some important changes. The city of Pasadena, Texas replaced two district council seats in predominately Latino neighborhoods with two at-large seats elected by the majority-white city; and Galveston County, Texas, eliminated virtually all district seats currently held by Latino and black incumbents.

Such developments, which continue, have heightened the debate about the current capacity of the Voting Rights Act to protect minority voting rights. Whether one advocates for pre-clearance or for other legal protections, the burden of evidence for those challenging voting restrictions has become higher since Shelby County. The processes of producing the evidence required in court are difficult and expensive, because they involve linking precinct-level vote tallies and demographics to social and historical conditions that encouraged discrimination, and making inferences about government motives.

But there is a simpler way to move forward – with recently improved analyses of survey data.

How Surveys Can Reinforce the Voting Rights Act

Because elections are administered at the state and local level, voting rights litigation also happens there. Until recently, however, there were no systematic survey measures of public opinion at the state and local level; very few people administer identical surveys to a representative sample in each of the 50 states, or in each of the 3,000 counties. Some analysts try to make inferences about state and local opinion by looking at the national survey answers of respondents who happen to live in a certain area, but the results can be unreliable because nationally representative surveys do not necessarily include representative samples within each state. Recently, though, a breakthrough in survey research called “multi-level regression and post-stratification” has improved the reliability of surveys that could be used in Voting Rights Act litigation by re-weighting pooled local responses.

So why does accurately measured public opinion matter in Voting Rights Act litigation?  Currently, Voting Rights Act litigation requires that minority voters alleging discrimination provide evidence of societal risk factors that could encourage government discrimination or evidence that racially polarized voting is connected to social and historical conditions. In both cases, surveys can provide telling evidence, at little expense. A large, nationally representative survey can be administered for a few hundred thousand dollars (or cheaper for online surveys) and can then serve as the basis for all voting rights litigation in the country for several years. Surveys can ask about policy preferences, generating data that can be compared between races to provide a cheap yet reliable measure of racial polarization. In addition, by using a national survey, judges would be able to evaluate the degree to which the racial polarization in any given state is greater or less than national polarization.

Because racial attitudes predict voting behavior and are a constitutionally valid measure of prejudice, survey measures of racial stereotyping could likewise be used by Congress to construct a pre-clearance coverage formula based on current conditions. For example, based on recent surveys, it looks like the pre-2013 Voting Rights Act actually did a remarkably good job of differentiating between states based on their constitutionally-prohibited proclivities to engage in discriminatory election practices. Thus a new coverage formula based on state-of-the-art survey measures would look almost exactly like the old coverage formula, but it would be solidly grounded in current conditions as the Supreme Court majority demanded in Shelby County.


Read more in Christopher S. Elmendorf & Douglas M. Spencer, "Administering Section 2 of the VRA after Shelby County," (Columbia Law Review, 2015) and Christopher S. Elmendorf & Douglas M. Spencer, "The Geography of Racial Stereotyping: Evidence and Implications for VRA Preclearance after Shelby County," (California Law Review, 2014).