Can Science Help Save the Voting Rights Act?

August 25, 2021
Michael Fleshman/Flickr
Michael Latner
Senior Voting Rights Fellow

The House of Representatives has now passed the John Lewis Voting Rights Advancement Act (JLVRAA), named after the iconic civil rights leader, in hopes of restoring the strength of the original Voting Rights Act of 1965 and protecting voters of color from state intrusion upon their voting rights. This is a vital step to protect democracy and ensure that communities can advocate for their own health and safety.

The new bill relies on a large body of evidence demonstrating existing disparities, revises the data and methods used to determine whether a jurisdiction must get preclearance before changing election laws, codifies some scientific methods of identifying discrimination that have grown out of decades of case law, and provides guidance to courts on how to apply some of the tests that shape current voting rights litigation.  

Overall, the proposed restoration reflects many advances in the use of science to protect voting rights. But it is far from clear that a Supreme Court majority which seems to deny both the scientific basis of discrimination and the central vision of the Voting Rights Act, is prepared to uphold it. Further, there are additional, scientifically justified protections that should be included in this bill to protect voters’ most fundamental constitutional right. 

An extensive series of hearings and a House Administration Committee Report, “Ensuring Free and Fair Access to the Ballot”, provide the empirical and evidentiary basis for the continuing need to protect voters of color from state intrusion upon their voting rights. For example, Janai Nelson of the NAACP Legal Defense Fund noted that “there have been at least nine federal court decisions finding that states or localities enacted racially discriminatory voting laws or practices intentionally” since the 2013 Shelby v Holder decision did away with preclearance requirements for states with a history of discrimination. The Report identified numerous examples of current disparities: 

  • “voter list maintenance and voter purge processes can be, and are, wielded in a discriminatory manner and have a disproportionate impact on minority voters.” 
  • “minority voters are less likely than White voters to have [required voter identification] and are more likely to lack the documents required to obtain these IDs.” 
  • “significant gaps remain in adherence to the law and the provision of fair access to multi-lingual voting materials and assistance” 
  • “evidence [of the] the disproportionate, discriminatory effect polling place closures, consolidations, and relocations and under resourcing has on access to the ballot for minority voters.” 
  • “discriminatory redistricting, vote dilution, annexations, deannexations, drawing of jurisdictional boundaries, and changes to the method of election…” 

The existing body of research suggests that such racial disparities are a real and continuing threat to election integrity: analysis of cell-phone data and precinct analyses show that voters in Black neighborhoods wait 20-30% longer to vote in-person compared to voters in White neighborhoods; a new Brennan Center analysis by Peter Miller, Kevin Morris and Coryn Grange shows that even as turnout has increased in recent elections, the racial turnout gap has also increased in jurisdictions previously covered under the Voting Rights Act; and as my recently co-authored book Gerrymandering the States demonstrates, biased political representation can amplify the inequalities that affect people throughout their day to day lives in terms of environmental, public health, and other outcomes. 

The JLVRAA addresses these threats through a number of sound, science-based improvements. Most notably, the bill establishes several practice-based preclearance requirements for all jurisdictions seeking to require new identification requirements, adopt new districting plans, eliminate polling places, or implement new voter role purging policies. States are also specifically prohibited from using partisan advantage as a justification for vote denial. The well-established test (known as the Gingles test) for determining whether a geographically cohesive bloc of minority voters is dominated by a bloc of White voters through a racial gerrymander, is codified into the JLVRAA.  

A new rolling formula is specified which continually updates the data used to determine if a jurisdiction is “covered” for past discrimination and requires preclearance before changing election laws: if, in the past twenty-five years, (1) fifteen or more voting rights violations occurred in the State or (2) ten or more voting rights violations occurred and one of those was committed by the State (as opposed to a sub-state government). A county is covered if it commits three violations in the same period, and coverage lasts for ten years. 

Additionally, the bill explicitly rejects five factors proposed by Justice Alito in his recent Brnovich v DNC decision for determining unacceptable vote deprivation, replacing them with a more empirical and precedent-grounded two-part test that requires plaintiffs to establish that a practice 1) causes a racial disparity through 2) interaction with historical and social conditions of discrimination. This is an important correction, as Justice Alito’s factors weighed heavily in favor of states and against the rights of voters of color. For example, Justice Alito urged courts to look at the “practical” over the “statistical” significance of racial disparities in rejected ballots. He considers it a “statistical manipulation” to claim that a group with a 1% rejection rate suffers “twice” the rate of rejection of a group with a 0.5% rejection rate (which, of course, is accurate), when both groups have 99% or more of their ballots accepted. 

If, out of two groups of 200 voters, one group had a single ballot rejected, and another had two voters’ ballots rejected, one could see Justice Alito’s point, and some top legal scholars saw this argument coming, given the specifics of the Brnovich case. But even in the Brnovich case, the Court was not considering three, but thousands of rejected ballots, enough to determine the outcome of an election. If that isn’t practically significant, I don’t know what is. Thus, the re-assertion of the importance of statistically significant racial disparities, with a determination of causality through empirical examination of social and historical conditions of discrimination, is crucial. 

For all these improvements, several shortcomings remain in the bill. For one, as law professor Nick Stephanopoulos and others argued in testimony during the hearings, states should have to empirically justify challenged practices, and show that the practice, and its disparate impact, is causally linked to the harm they claim a compelling state interest in preventing. For example, such a test would prohibit states from claiming protection against voter fraud unless they can show that voter fraud is a significant problem that would be curtailed through their practice (the bill already explicitly rejects the “mere invocation” of voter fraud as a consideration).  

Additionally, while the JLVRAA explicitly endorses coalition districts –where a coalition of minority voters (Black and Hispanic voters, for example)—as satisfying the conditions for a Gingles test, it also specifies that the minority voters must constitute a majority in a single-member district. Whether this reflects a move to thwart a constitutional challenge or just the Democratic Party’s attraction to safe, single-seat districts, it is not scientifically justified: “crossover” districts with coalitions of minority and like-minded White voters have been shown to provide an effective remedy against racial vote dilution. Perhaps the Democratic Party leadership does not grasp the emergency situation that we find ourselves in.

Even the best science may be unable to restore the racial voting rights regime that courts have used to protect democracy and voters of color in the US for 50 years. The current Supreme Court majority has effectively turned the vision of the Voting Rights Act on its head, leading the voting rights scholar Guy-Uriel Charles to conclude that “the Court majority is more interested in protecting the electoral rules of the states from undue intrusion by voters of color.”  

On that point, another way of strengthening the JLVRAA would be, as Travis Crum has suggested, to federalize it and secure its protections from the federal government as well as state governments. Yet another is to press more strongly for universal, practice-based requirements directed at federal elections, such as those included in the For the People Act, and non-race-specific remedies for vote dilution such as those included in the Fair Representation Act. Color-blindness in remedies should be more attractive to a Court that chooses to ignore color in the causes or consequences of restrictive election laws.

The Senate will take up this great task soon. Let us all hope that they can put together some set of protections for voters of color while we still can.