Think Our Voting System is Color Blind? Think Again.

Email a Friend
A new study from researchers at Harvard suggests that racial bias still plays a big role in the voting system.
From and

When Congress passed the 1965 Voting Rights Act, non-white voters in the South and urban areas across the country had been facing racial discrimination at the polls.

“Nearly 50 years later, things have changed dramatically,” Supreme Court Chief Justice John Roberts wrote in the 2013 The Shelby County v. Holder decision, which struck down the heart of the law. Has the “extraordinary problem” of voter discrimination really disappeared? 

New political science research suggests that our voting system has simply traded one form of racial discrimination for another. Ariel White and Noah Nathan, two co-authors of the first political science experiment to test implicit bias in the voting system in the modern era, conclude that local election officials exhibit racial bias in how they implement election regulations and distribute registration requirements to voters.

The spate of new voter ID laws passed before the 2012 presidential election prompted the study, which was conducted in the run up to the Shelby decision.

“We have a lot of relatively confusing laws being passed, often relatively near to the election or being implemented at the last minute,” White says. “That’s where we worry that people who contact their local election officials may have a hard time getting information about these laws, and may then go on to not fully know how to go vote.”

While the study doesn’t establish that voters are actively being suppressed, it does definitively show that, across the country, officials in charge of election law discriminate against voters based on ethnicity.

“Different types of voters have different types of hurdles to overcome to get accurate information about what’s actually going to be required,” Nathan says. “Just because a law might in theory apply equally to everyone, in practice, when people interact with the bureaucrats implementing election laws, they might have very different experiences, and that might trickle down into whether or not they have the right information they need when they go to the polls.”

Electoral discrimination brings to mind Ava DuVernay’s recent film "Selma," which opens with a sneering white registrar denying civil rights activist Annie Lee Cooper’s (Oprah Winfrey) application for voter registration on the basis of a twisted civic quiz—the kind of practice later outlawed by the Voting Rights Act.

While the Voting Rights Act effectively prevented this kind of bald-faced discrimination in targeted areas, the study uncovers implicit bias throughout the entire electoral system.

“What we’re showing in this experiment is not the county clerk in the first scene of 'Selma' turning away Oprah Winfrey with that explicit, overt racism,” Nathan says. “What we’re actually finding is a much more subtle and implicit bias that is similar to bias that’s been found in all sorts of other experiments similar to ours in other fields—with loan officers, HR workers, doctors, real estate agents—that show that when people make discretionary decisions about others, they can be implicitly biased against minorities in how helpful they’re going to be even if they don’t actually intend ill to those people.”

White, Nathan, and their co-author Julie Faller—all of whom are PhD candidates in government at Harvard—sent e-mails to over 7,000 local election officials in 48 states asking for basic information about voter registration and whether identification was needed on election day. They sent identical e-mails from names that sounded ostensibly white, and names that sounded ostensibly Latino.

While some officials did send correct information to both white and Latino voters, a disturbing pattern began to emerge.

“When we sent those e-mails—the same text of the same e-mail—from a Latino-sounding name we got fewer responses overall. We more often heard radio silence. And we got slightly less thorough responses,” says White.

“The aggregate effects of this in the population are that people from minority groups might be getting less informative information when they ask questions to these officials,” Nathan says.

With the patchwork of new voter ID laws in 31 states, this kind of unconscious bias shifts the discriminatory burden from the poll site on election day to the weeks and months leading up to an election, when voters get the majority of their information on voting requirements.

“With an experiment like this, you can’t explain this bias away based on anything else," Nathan says. "These are exactly the same e-mails sent to thousands of people. Everything about them is the same. The ethnicity of the person sending them is the really only thing that could plausibly be causing these differences.”

In the wake of the Shelby County v. Holder decision, White says that their study, far from settling these issues, is a call for more research on how implicit bias impacts voter behavior and election outcomes.

“We do need a critical mass of social science research that can address this question,” White says.

Surprisingly, the mostly Southern and urban areas covered by the Voting Rights Act actually treated voters more fairly than other areas of the country.

“In these areas the officials didn’t discriminate in their responsiveness or in the accuracy of their responses to our questions,” Nathan says. “It’s suggestive evidence that when you have people looking over your shoulder and making sure that you aren’t biased, you aren’t biased.”

In a post-Shelby County v. Holder America, whether this pattern will hold is still anyone’s guess.

“I think the best analogy here comes from Justice Ginsburg in the Shelby decision,” Nathan says referring to Ruth Bader Ginsburg’s well-known oral dissent from the bench. “She argued, ‘You’re not getting wet. Is that because it’s raining, or because you’re holding up an umbrella?’ And the question is, should you close the umbrella?”

“And the decision that they made was to close the umbrella,” Nathan adds.