In collaboration with Election Law Blog, Take Care is pleased to present a series of posts offering thoughts on legislation to reform the U.S. electoral process.
Things have changed in the South since Shelby County. And not for the better. After the Supreme Court struck down the Voting Rights Act’s coverage formula in June 2013, several previously covered jurisdictions passed discriminatory election laws. Perhaps most prominently, North Carolina enacted a voter-suppression law that the Fourth Circuit invalidated on intentional discrimination grounds. More recently, the 2018 midterm elections were marred by voter-suppression tactics in Georgia, Texas, and elsewhere.
In response to these problems, Democratic leaders have pledged to pass a revised VRA when the next House of Representatives convenes in January. As I have argued before, Congress should revise Section 3(c)’s bail-in provision to authorize courts to impose preclearance based on a violation of Section 2 of the VRA and any other federal law that prohibits discrimination in voting based on race, color, or language minority status. Currently, Section 3(c) authorizes bail-in only for violations of the Fourteenth or Fifteenth Amendments. It is well established that the Fourteenth Amendment prohibits intentional racial discrimination in voting. The Fifteenth Amendment standard is less settled, though a plurality of the Court concluded in 1980 that the Amendment requires a showing of discriminatory intent. If Congress amends Section 3(c) to authorize bail-in based on a statutory violation, courts could impose preclearance based on a finding of discriminatory effect—not just intent.
Why should Congress make this change? With an effects-test Section 3(c), it would be far easier, cheaper, and quicker to bail-in jurisdictions. Indeed, the small number of bail-ins is partly attributable to Section 3(c)’s constitutional trigger. Following Shelby County, civil rights groups and the Obama Administration’s Department of Justice requested Section 3(c) relief in several cases across the country. As I have written previously, the results of these suits have been mixed, though there is still the possibility of bailing-in Texas for its intentionally discriminatory post-2010 redistricting plans. A common thread running through these decisions is the difficulty of proving intentional discrimination and the time- and fact-intensive discovery necessary to make that showing. By contrast, it is often far easier for plaintiffs—and less intrusive on States and political subdivisions—to show that a law has a discriminatory effect. An effects-test Section 3(c) would therefore expedite litigation and increase plaintiffs’ leverage in Section 2 cases.
An effects-test bail-in mechanism also sidesteps the constitutional issues inherent in any new coverage formula, which will have to satisfy Shelby County’s equal-sovereignty standard. By its own terms, Shelby County applies only when Congress differentiates between the States—an intrinsic part of any coverage formula. Section 3(c), by contrast, establishes a nationwide rule that applies to all States equally. Thus, unlike a revised coverage formula, Section 3(c) can be defended under the far more deferential standard articulated in South Carolina v. Katzenbach for Congress’s Fifteenth Amendment enforcement authority.
To be sure, an effects-test Section 3(c) raises more constitutional questions than the current version. In decoupling bail-in from a constitutional violation, Congress would have to rely on its Reconstruction Amendment enforcement authority not only for the remedy but also for the underlying violation that triggers coverage. But that is true for all the proposed coverage formulas. Indeed, some proposals rely on persistently low minority voter turnout—a proxy for unconstitutional conduct that is not even a statutory violation. In addition, under Section 3(c), courts make the initial liability finding and subsequent bail-in determination and then fashion relief to the particular case. These distinctions are likely more persuasive to the Roberts Court than a reverse-engineered coverage formula passed by Congress.
All of the post-Shelby County bills to amend the VRA have incorporated an effects-test bail-in provision. But Congress should do more. Congress should also provide guidance on what type of voting rights violations require bail-in, the appropriate time period for bail-in, and the types of changes that ought to be precleared.
In providing this guidance, Congress should identify the most problematic voting rights violations and require the imposition of preclearance in certain circumstances. In my view, if a jurisdiction adopts a racially discriminatory redistricting plan, it should be automatically bailed-in for the next redistricting cycle. Redistricting is a problem that calls out for congressional action given its importance and many States’ histories as serial gerrymanderers.
Congress should further specify time limits for Section 3(c) relief. It is often assumed that a jurisdiction can be bailed-in for a maximum of ten years. Although this is a reasonable and constitutionally prudent time period, that outer limit is not found in the statute’s plain text, and some bail-ins have lasted longer than ten years or had no definitive end point. Setting an outer limit of ten to twelve years—i.e., long enough to capture the next redistricting cycle and its immediate aftermath—would make both practical and strategic sense.
In a similar vein, Congress should make explicit that courts are permitted to impose “targeted” preclearance, that is, require jurisdictions to preclear only certain types of voting changes rather than all voting changes. Courts have often relied on their equitable authority or consent decrees to fashion preclearance regimes that target the same type of changes that prompted the bail-in. This revision—as well as my time-limit suggestion—would further bolster Section 3(c)’s constitutionality.
There is a general consensus that an effects-test Section 3(c) will be included in a revised VRA; the coverage formula’s future is less certain.Voting rights advocates and members of Congress should be wary of investing precious time, resources, and political capital into resurrecting the coverage formula. Notwithstanding numerous warning signs in the 1990s and early 2000s that the coverage formula was constitutionally suspect, there was no agreed-upon alternative in 2006 and Congress simply re-enacted the same coverage formula that had been in place since 1975. Even five years after Shelby County, there is still not agreement.
This lack of consensus is apparent from pending bills in the current Congress: the House and Senate versions of the coverage formula contain disparate time horizons and different numbers of “voting rights violations” for triggering preclearance. For example, the House bill would cover a State that had five voting rights violations in the past fifteen years and one of those violations was committed by the State itself; the Senate bill would not. Moreover, the types of violations that count toward coverage differ. The House bill expressly excludes cases striking down voter ID laws on statutory grounds from the coverage formula. The Senate bill lacks this exemption. Furthermore, the House bill does not count consent decrees and settlements in the coverage formula; the Senate bill does.
Of course, these differences are the result of political deals in the lame-duck Congress. The incoming Democratic House’s coverage formula will probably look somewhat different. But the general point still stands that any revised coverage formula will be the product of compromise.
And here, in setting priorities, it is important to ask what these proposed coverage formulas will accomplish that cannot be done with an effects-test Section 3(c). Many of the proposed coverage formulas use specified “voting rights violations” as triggers for coverage: constitutional violations, statutory violations, denials of preclearance, and (in the Senate) consent decrees. Some proposals incorporate factors like persistently low minority voter turnout, but, as noted above, these proxies raise their own constitutional concerns. The proposed coverage formulas are also “rolling” in that they keep track of “voting rights violations” over time and trigger coverage if the number of violations reaches a certain point in a specified time period.
As I see things, a revised bail-in mechanism can accomplish most of this work—and without raising equal-sovereignty concerns. Section 3(c) already covers constitutional violations, and an effects-test Section 3(c) will encompass statutory violations. The vast majority of bail-ins have been accomplished via consent decrees, so that category is not mutually exclusive of bail-in. And although an effects-test Section 3(c) does not capture recent Section 2 cases or pre-Shelby County denials of preclearance, the rolling nature of the proposed coverage formulas means that the relevance of these violations will fade in the rearview mirror. Because it is unlikely that any revised coverage formula passed by the Democratic House would survive the Republican Senate and be signed into law by President Trump, the clock for a new coverage formula starts in 2021 at the earliest.
And looking ahead to a world where a revised coverage formula is actually enacted, there is the inevitable constitutional challenge. Although Chief Justice Roberts’ opinion in Shelby County left the door open for a new coverage formula, this dicta should be taken with a grain—nay, a tablespoon—of salt. Voting rights advocates and members of Congress must recognize that they are negotiating in the shadow of a hostile judiciary and should not trade away too much to pass a coverage formula that may not survive the Supreme Court.
In short, an effects-test Section 3(c) is a straightforward and pragmatic option for strengthening the VRA. It is also on firmer constitutional ground than any proposed coverage formula. And if it is robustly enforced, it can do much of the work of a resurrected coverage formula.