Over the past year, states have redrawn their voting district boundaries for the first time since the Supreme Court struck down a significant portion of the 1965 Voting Rights Act (VRA) in its 2013 Shelby County v. Holder decision. Without that key federal protection, has anything changed in racial and ethnic minorities’ opportunities to elect representatives of their choice?
Here’s what we found when we looked at this question.
How the Voting Rights Act affected redistricting — and how the law changed
Section 5 of the VRA used to require states and counties with a history of discrimination — most in the South — to submit any election-related changes to the Department of Justice civil rights division for review. This Section 5 process, called “preclearance,” was designed to prevent discriminatory practices before they could be implemented. As applied to redistricting, preclearance sought to avoid “retrogression,” where minority representation was lower under the new plan than under the old one.
In 2013, a Supreme Court majority in Shelby County v. Holder struck down the formula that identified jurisdictions subject to preclearance. Voting rights advocates can still sue if they think a practice is discriminatory, but they have to live with that practice first. In other words, they are back to playing catch-up. Even this option might be at risk; this week, the Supreme Court heard oral arguments in Merrill v. Milligan, a case that might weaken that as well.
So how have states handled redistricting without preclearance?
How we did our research
In new research, we measured the Black and Latino shares of eligible voters in each state legislative and congressional district drawn around the country over the past two years. We focused on both U.S. House and state legislative districts in states where Black and Latino voters each constitute at least 10 percent of the voting-eligible population and there are at least four legislative seats.
With these numbers, we can calculate the gap between Black and Latino shares of a state’s overall eligible voter population on the one hand, and the share of districts where each group has “influence” on the other. Influence means a group is large enough that its support is critical or even essential for victory. We used 40 percent as our cutoff for influence, although you could make a reasonable argument for a higher or lower threshold.
All else equal, Black and Latino influence should be proportional: They should have influence in the same share of districts as their share of voting-eligible residents. With 20 percent of the voting-eligible population, they should influence 20 percent of the district outcomes; at 30 percent, it should be 30 percent, at 40, it should be 40, and so on. It’s not always this simple, as we discuss below. But it’s a reasonable place to start.
How did redistricting affect Black voters’ influence?
In the figures below, you can see the gaps in Black voters’ representation before and after the most recent redistricting. Negative gaps mean Black voters fall below proportionality (underrepresented), and positive ones mean they fall above it (overrepresented). The diagonal line in each plot marks the point where the old and new redistricting plans produce the same result.
In most states, we found that Black voters are underrepresented — both before and after the post-Section 5 redistricting plan. Some plans come closer to proportionality than others. For instance, in Illinois, Michigan and Missouri, congressional districts with Black influence either match the Black population share or exceed it.
While no states significantly increased the proportion of congressional districts with Black influence, Maryland, Michigan and Virginia did add more Black-influence districts to their state legislatures. In fact, we found that state legislative plans often come closer to proportionality, perhaps in part because they have more seats available to fine-tune representation. All the same, most congressional plans have enough seats to equitably represent Black voters, but do not.
Some states formerly covered by the VRA’s Section 5 did significantly cut Black representation. North Carolina did so in both its congressional and state legislative plans. Florida reduced Black-influence congressional districts, while Mississippi and South Carolina did so in their state legislatures.
How did redistricting affect Latino voters’ influence?
In the figures above, you can see that Latinos are commonly underrepresented as well. However, some states still shrank the gaps in their post-Section 5 district maps. California’s independent commission significantly increased the number of Latino-influence districts in the state’s congressional delegation, and somewhat increased that number in its state legislature. Florida also modestly increased Latino voters’ influence in congressional districts, as did New Jersey in its state legislative districts.
But in the Colorado, Florida, New York, Rhode Island and Texas state legislatures, Latino influence declined. Many of these declines — and many of those for Black voters as well — would have raised red flags under Section 5.
Would preclearance have changed these districts?
Such declines are not necessarily intentional. For one thing, if we were to define influence just a little differently — at 35 percent instead of 40 — Blacks and Latinos would look much better represented in some states. North Carolina’s decline in Black representation looks much smaller with such a threshold, as does the decline in Latino representation in the Texas legislature. At the same time, raising the threshold often makes representation look worse.
So why pick one level over another? The goal should be the level where a minority group is potentially able to support a winning candidate, either by itself or by joining with other groups. That threshold will be lower in an area where more voters from other groups are willing to join a multiracial coalition.
Nor does underrepresentation always warrant legal action. The VRA does not require minority representation if a group is small, dispersed or does not routinely choose different candidates than other groups. Even if states wanted to draw districts to support such a group, courts have limited how aggressively they can do so.
However, the VRA’s preclearance requirement would have required states that decreased minority representation to show that they were not discriminating before making the changes. Now, those opposed to these district maps are suing in six states — Alabama, Arkansas, Florida, Georgia, Louisiana and Texas — and all but one was at least partly covered by Section 5.
Eric McGhee is a senior fellow at the Public Policy Institute of California, where he studies elections and election reform.
Christopher Warshaw is an associate professor of political science at George Washington University and the co-author of “Dynamic Democracy: Public Opinion, Elections, and Policymaking in the American States.”
Michal Migurski is the executive director of PlanScore.