1. How Section 5 preclearance works in practice

Section 5 preclearance reverses the ordinary order of lawmaking. In most of American government, a legislature or a local board adopts a rule and the rule takes effect unless a court later strikes it down. In the jurisdictions covered by Section 5 of the Voting Rights Act, the sequence runs the other way: before any new voting rule can be enforced, the jurisdiction must first obtain federal permission.

The practical meaning of that reversal shows up in the routine decisions of local government. A county commission votes to move a polling place from a school gymnasium to a fire station across town. A city council redraws its council districts to reflect new census figures. A state legislature shortens the registration window or rewrites the qualifications for candidates. Outside the preclearance regime, each of these changes takes effect on its own schedule and remains in force until someone mounts a successful challenge. Inside a covered jurisdiction, none of them may be enforced until the jurisdiction has submitted the change and received clearance. The duty runs in one direction only: the jurisdiction must come forward, and the proposed rule waits in legal limbo while federal reviewers examine it. The broader statute in which this mechanism sits is surveyed in the complete guide to the Voting Rights Act of 1965, which places preclearance alongside the Act’s other enforcement tools.

The reversal was a response to a specific failure. Before 1965, voting discrimination was attacked almost entirely through case-by-case lawsuits, and Congress compiled a record showing that the strategy could not keep pace. Each time a court struck down a discriminatory device, jurisdictions devised new ones, and each new device required a fresh round of litigation that took years to conclude while elections went forward under the challenged rules. Preclearance was designed to break that cycle by shifting the burden of inertia: instead of the victims of discrimination having to sue to stop a change, the jurisdiction would have to obtain approval before the change could take effect.

The “whom” in this arrangement is defined with precision. Section 5 applies to the states and political subdivisions that fall under the coverage formula of Section 4(b), examined in the next section. Coverage may attach to an entire state, in which case every county, city, and school board within it must submit its voting changes, or to individual political subdivisions such as counties, in which case only those subdivisions carry the obligation while the rest of the state proceeds under ordinary law. The distinction matters enormously on the ground. A state covered in full submits everything from legislative redistricting to municipal annexations. A state covered only in part sees preclearance operate county by county, with neighboring counties subject to entirely different legal regimes. The statute uses political subdivision as a functional term: counties and parishes in most states, townships where those administer elections, and independent cities where applicable. Special-purpose districts, from water boards to utility districts, fall under the obligation when they conduct elections within covered territory.

Clearance travels down one of two paths. The path that carries the great bulk of submissions runs through the Attorney General. The chief legal officer or another appropriate official of the covered jurisdiction submits the proposed change to the Department of Justice, and the Attorney General then has sixty days to interpose an objection. The Attorney General may also give an early affirmative indication that no objection will be made. If the sixty days pass with no objection, the jurisdiction may enforce the change. If the Attorney General requests additional information, a new sixty-day period begins when that information arrives, and the published procedures treat only a complete submission as starting the clock. A jurisdiction facing an imminent election may ask for expedited consideration, and the procedures provide for it. The second path runs through the courts. The jurisdiction may instead bring an action in the United States District Court for the District of Columbia seeking a declaratory judgment that the change has neither a discriminatory purpose nor a discriminatory effect, heard by a three-judge panel with direct appeal to the Supreme Court. In either forum the burden of proof rests on the jurisdiction. It must show that the change clears the statutory standard; the reviewers need not prove that it fails. A jurisdiction that draws an objection from the Attorney General is not without recourse: it may ask for reconsideration, it may modify the proposal to answer the objection and submit the revised version, or it may take the change to the District Court for the District of Columbia, where the court considers the question anew. In practice the court route sees far less traffic. It is slower and more expensive than administrative review, and jurisdictions tend to reserve it for changes of exceptional importance or for cases in which an objection has already issued. One further feature deserves emphasis. The statute provides that neither the Attorney General’s failure to object nor a declaratory judgment bars a later lawsuit to enjoin the change. Clearance is permission to proceed, not a permanent shield.

A submission is more than a cover letter. The jurisdiction describes the change in detail, explains the reasons for its adoption, identifies the prior practice it replaces, and supplies the demographic and electoral data needed to assess the change’s impact on minority voters. Reviewers examine whether the jurisdiction has carried its burden under the purpose-or-effect standard, and an objection, when it issues, states its reasons in writing. The process is administrative in character but consequential in operation: an objection blocks the change unless and until the jurisdiction prevails in court. The published procedures also invite comment from interested members of the public while a submission is pending, so the review need not rest solely on the jurisdiction’s own account of its change.

The review itself is comparative. The jurisdiction’s submission is measured against the benchmark, the voting practice in force at the time of submission, and the question is whether the proposed change was adopted with a discriminatory purpose or would have a discriminatory effect relative to that benchmark. The benchmark is simply whatever is then in effect, whether or not that practice was itself precleared when adopted. The purpose inquiry looks to the reasons behind the change; the effect inquiry looks to its practical consequences for minority voters. Because the burden rests on the submitting jurisdiction, an incomplete or unpersuasive showing is itself grounds for an objection. The jurisdiction need not prove that the change is ideal or even wise, only that it clears the statutory bar, but a submission that leaves the question in doubt does not earn clearance.

The set of changes that must run this gauntlet is deliberately broad. Section 5 reaches any voting qualification or prerequisite to voting, and any standard, practice, or procedure with respect to voting, that differs from what was in force when the jurisdiction became covered. That baseline matters. The statute measures every proposed change against the voting rules in force or effect on the jurisdiction’s coverage date, which is why the trigger dates of the coverage formula do double duty: they determine who is covered and they fix the starting point against which later changes are judged. Courts have read the operative language to cover far more than the obvious cases of new registration rules. Redistricting plans must be submitted, as must polling place relocations. So must annexations and deannexations that alter the composition of the electorate, shifts from district-based to at-large methods of election, changes in the form of municipal government that affect how officials are chosen, new requirements for candidates seeking office, and alterations to the times, places, and manner of registration. The Supreme Court confirmed the breadth of this reach in Allen v. State Board of Elections (1969), holding that changes such as a move to at-large voting fall within Section 5 even though they do not directly regulate who may register. The operating principle is functional rather than formal: if a governmental action touches the mechanics of voting in a covered place, the jurisdiction should assume the submission requirement applies. And the requirement follows the change into every election the jurisdiction conducts, whether federal, state, or local. School board contests, municipal bond referenda, and special district elections are all within it; the trigger is the act of voting, not the importance of the office.

The obligation has teeth. A covered jurisdiction that enforces a voting change without preclearance acts unlawfully, and the change may be enjoined before it affects an election. The Attorney General may bring an enforcement action, and the Supreme Court held in Allen that private citizens may also sue to compel compliance with the submission requirement. In enforcement suits, courts have treated the absence of preclearance as decisive: a change subject to Section 5 that has not been cleared cannot be given effect, and elections conducted under such a change may be enjoined before they occur. Courts hearing such cases have halted elections conducted under uncleared changes and ordered jurisdictions back into the preclearance process. The remedy is preventive by design: it stops the change before it can affect an election rather than attempting to unscramble the results afterward.

It helps to keep Section 5 distinct from the Act’s permanent, nationwide prohibition. Section 2 forbids voting practices that deny or abridge the right to vote on account of race, and it applies everywhere in the country without any need for coverage. Section 5 is the extraordinary preventive remedy layered on top of that prohibition for the covered jurisdictions. Where Section 2 addresses discrimination after the fact, Section 5 requires permission before the fact. Its question is narrower and forward-looking: not whether discrimination has occurred, but whether a particular change may go into effect.

The provisions at the center of this article are sections 4(b) and 5 of the Voting Rights Act of 1965, Pub. L. 89-110, 79 Stat. 437, codified at 52 U.S.C. 10303 and 10304. Readers working with older commentary will encounter different numbers for the same provisions; the guide to reading federal statutory citations explains how to follow a federal provision across the compilations of the code in which it appears.

Preclearance is operative federal law rather than a historical artifact. Congress reauthorized the Act’s temporary provisions, including the coverage formula and the preclearance requirement, in 2006 for an additional twenty-five years. The reauthorization left the trigger dates and the formula’s structure unchanged, extending the existing coverage rather than rewriting it. Election officials in covered jurisdictions continue to submit changes for review as a matter of routine administration, and the machinery described here continues to turn.

2. The coverage formula in Section 4(b)

Congress faced a drafting problem in 1965. It wanted the extraordinary remedy of preclearance to apply where the need was greatest, but it also wanted the statute to operate as a general law rather than a measure singling out states by name. The solution was Section 4(b): a formula that describes, in objective terms, the jurisdictions to which Section 5 applies. The formula names no state. It states two conditions, and any state or political subdivision meeting both conditions falls under the preclearance obligation. The elegance of the device lies in its neutrality of expression; its precision lies in the way the two conditions were calibrated to capture the jurisdictions Congress had in mind. The calibration was not guesswork. Congress built the formula on an extensive evidentiary record: hearings documenting the devices registrars used to exclude Black voters, Justice Department litigation files showing the same jurisdictions appearing again and again, and findings on registration disparities compiled by the Civil Rights Commission. The two prongs distilled that record into administrable tests. The test-or-device prong identified the mechanism of exclusion; the participation prong measured its apparent consequence. Together they marked, without naming, the places where federal supervision of voting changes was thought necessary.

What two facts made a jurisdiction covered under Section 4(b)?

A jurisdiction became covered under Section 4(b) when two facts coincided: the Attorney General determined it used a test or device as a condition of voting on the relevant November 1 trigger date, and the Census Director determined that registration or presidential-election turnout there fell below fifty percent of voting-age residents on that same date.

The first prong looks to the Attorney General’s determination that the jurisdiction maintained a test or device as a prerequisite to voting as of November 1 of the trigger year. As enacted in 1965, the trigger date was November 1, 1964. The 1970 amendments added November 1, 1968 as a second trigger date, and the 1975 amendments added November 1, 1972 as a third. The word “maintained” carries weight. The question was not whether a test or device existed somewhere in the jurisdiction’s statute books but whether it was in use as a condition of voting on that date. The determination was administrative rather than judicial, made by the Attorney General and published for the record.

Each trigger date arrived as part of a broader reauthorization. The 1970 amendments extended the preclearance provisions for five years and added the 1968 trigger, sweeping in the jurisdictions that met the formula’s two prongs under the new date. The 1975 amendments extended the provisions for seven more years, added the 1972 trigger, and, crucially, rewrote the definition of test or device to reach English-only election administration. The structure never changed: each amendment kept the existing coverage intact and layered new coverage on top, so the map of covered jurisdictions only grew.

The statute defines “test or device” with considerable specificity. The term covers any requirement that, as a prerequisite for voting or registration, a person demonstrate the ability to read, write, understand, or interpret any matter; demonstrate educational achievement or knowledge of any particular subject; possess good moral character; or prove qualifications by the voucher of registered voters or members of some other class. In the covered jurisdictions this meant literacy tests and the constellation of kindred devices that registrars had used to screen applicants, from examinations in constitutional interpretation and demands for proof of schooling to character attestations and requirements that an established voter vouch for a newcomer. The 1975 amendments widened the definition through Section 4(f)(3), which treats as a test or device the practice of providing registration or election materials, including ballots, only in English in any jurisdiction where more than five percent of the citizens of voting age belong to a single language-minority group. The statute understands language-minority groups to include persons of Spanish heritage and those of Asian American, American Indian, or Alaskan Native descent. The expansion reflected Congress’s judgment that English-only election administration could function, in the covered circumstances, as a barrier equivalent in practical effect to the older written tests. Applying the expanded definition required two administrative findings working in tandem. The Attorney General determined whether the jurisdiction provided its election materials only in English; the Census Bureau’s data established whether more than five percent of the voting-age citizenry belonged to a single language-minority group. Where both findings coincided with depressed participation under the second prong, coverage attached exactly as it had under the original literacy-test prong. The mechanism was new, but the logic was the formula’s original logic: a barrier to the ballot combined with low participation marked the jurisdictions where preclearance would apply.

The second prong looks to the Director of the Census, who was to determine whether less than fifty percent of the persons of voting age residing in the jurisdiction were registered to vote on the November 1 trigger date, or whether less than fifty percent of such persons voted in the presidential election of that trigger year. Either shortfall satisfies the prong; the statute joins the two measures with “or.” Registration rolls and presidential turnout thus served as the formula’s quantitative signal. The measurements tracked the voting age then in force: at the time of the 1964 and 1968 triggers, that age was twenty-one in most states, and the Census Bureau’s figures reflected it. A jurisdiction that combined a test or device with depressed participation, measured as of the same date, was treated as a place where the ordinary protections of the political process could not be relied upon to correct discriminatory administration.

The making of these determinations was itself a substantial administrative undertaking. The Attorney General surveyed the election laws and practices of the states to identify where tests or devices were maintained in fact, while the Census Bureau compiled registration figures and presidential election returns to measure participation against the fifty percent line. The resulting determinations were published in the Federal Register, giving covered jurisdictions formal notice of their status. And the statute declares that the determinations are not reviewable in any court. Congress thus insulated the factual predicates of coverage from collateral attack. A jurisdiction could not relitigate, in a Section 5 proceeding or anywhere else, whether the Attorney General had been right about its literacy test or whether the Census Director had counted correctly. Once both determinations pointed at a jurisdiction, coverage attached by operation of law, carrying with it the full preclearance obligation described in the preceding section. Because the determinations were not reviewable, the Federal Register notice was effectively the last word on a jurisdiction’s status.

The formula’s design survived early constitutional attack. In South Carolina v. Katzenbach (1966), the Supreme Court sustained Sections 4 and 5, holding that Congress had acted within its enforcement authority in reaching the covered jurisdictions through the two-pronged test rather than by naming them. The decision treated the combination of tests or devices with depressed participation as a rational basis for identifying where the preclearance remedy was needed. Congress was not required to prove discrimination jurisdiction by jurisdiction; the formula’s proxies were enough. South Carolina itself, the lead plaintiff, remained covered under the formula the Court sustained.

When the two determinations were first made under the 1964 trigger, they captured seven states in full: Alabama, Alaska, Georgia, Louisiana, Mississippi, South Carolina, and Virginia. Alaska’s presence on that list surprises readers who associate the formula exclusively with the South, but the state maintained a literacy test and its participation figures fell below the statutory threshold, so the neutral formula reached it as written. The determinations also captured counties in four additional states: Arizona, Hawaii, Idaho, and North Carolina. Coverage in part was a feature of the design rather than an afterthought. The formula operates at the level of the political subdivision as well as the state, so a county that met both prongs fell under Section 5 even when its state as a whole did not. The seven-state list reflected where both prongs coincided; states that used tests but maintained higher participation, or that showed low participation without tests or devices, stayed outside. Partial coverage also created an administrative split that election officials had to navigate carefully. A statewide change adopted by the legislature of a partially covered state, such as a congressional redistricting plan, had to be precleared for its application to the covered counties even though the state as a whole was not covered. The submission obligation followed the covered territory, not the level of government acting.

The 1968 and 1972 trigger updates extended the formula’s reach without altering its structure. Each new trigger date simply re-ran the same two-pronged test against fresh facts: did the jurisdiction maintain a test or device on the new November 1 date, and did registration or presidential turnout fall below half of the voting-age population? The updates brought new ground under coverage, including Texas in its entirety, counties in California, Florida, Michigan, New York, and South Dakota, and, through the language-minority expansion of the definition of test or device, jurisdictions in Texas and Arizona where English-only election materials coincided with concentrated language-minority populations. The pattern shows how the formula compounded over time. Each amendment preserved the existing coverage while adding the jurisdictions that the new trigger dates and the broadened definition swept in. The language-minority expansion supplied the formula’s second great wave of coverage, reaching jurisdictions in the Southwest and elsewhere that the original literacy-test prong had never touched.

Coverage, once attached, was not irrevocable. Section 4(a) provides a mechanism known as bailout by which a covered jurisdiction may seek release from the preclearance obligation. In 2009 the Supreme Court decided Northwest Austin Municipal Utility District No. 1 v. Holder, a case that turned on the availability of that mechanism while leaving the coverage formula itself undisturbed. Broader constitutional questions about the formula have been raised in litigation as well, and the later court challenge to the coverage formula is examined in the companion article on that case.

3. The two preclearance routes

Section 5 preclearance works through two distinct channels, and a covered jurisdiction may choose either one when it seeks approval for a proposed voting change. The first is administrative review by the Attorney General. The second is a declaratory judgment action in federal court. Both routes carry the same substantive requirement, and both place the burden of persuasion on the jurisdiction rather than on the government, but they differ markedly in procedure, timing, cost, and institutional posture. In practice, the administrative route handled the overwhelming share of submissions, while the judicial route functioned as a slower and more expensive alternative that jurisdictions invoked only occasionally, often when a proposed change raised novel or especially sensitive questions.

The Administrative Route: Review by the Attorney General

The administrative route begins when a covered jurisdiction prepares a submission. The submission must identify the proposed voting change, describe it in enough detail that a reviewer can understand what the jurisdiction intends to do, and supply supporting information relevant to the central question the statute poses: whether the change has either a discriminatory purpose or a discriminatory effect. That means the submission typically includes the text or a full description of the new rule or practice, a statement of what the rule or practice replaces, and whatever demographic, geographic, or electoral data the jurisdiction believes bears on the change’s consequences. A submission concerning a redistricting plan, for example, would be expected to include maps, population figures by relevant categories, and an account of how district boundaries shift minority voting strength. A submission concerning a polling place change would describe the locations involved and the voters affected. The precise informational burden rises and falls with the complexity of the change, but the principle is constant: the Attorney General cannot evaluate the submission on bare assertions, so the jurisdiction must give the reviewer enough to work with.

Once the submission arrives, the clock starts. The Attorney General has 60 days to act. The statute gives the Department of Justice a defined window rather than an open ended review period, and that window matters a great deal to the jurisdictions waiting on approval. If the Attorney General interposes an objection within the 60 day period, preclearance is denied, and the jurisdiction may not implement the change. If the 60 days expire with no objection, the change is deemed precleared by operation of law. Silence grants permission. This feature of the design is one of the most important practical facts about Section 5 preclearance: the administrative route does not end with an affirmative certificate of approval in every case. Many changes enter into force simply because the Department raised no objection in time.

The 60 day period is not entirely rigid. The statute permits the Attorney General to request additional information, and such a request can extend the review period under specified conditions, effectively restarting or lengthening the window while the Department waits for a complete record. The extension mechanism exists so that the Department is not forced to decide on a thin file: when a submission arrives incomplete or raises questions the supporting materials do not answer, the Department can ask for more and take the additional time the request triggers. For the jurisdiction, this means the 60 day clock is a conditional promise. A well prepared submission moves faster; a thin or ambiguous one invites follow up that delays the outcome.

The Department’s handling of submissions is governed by published procedures that set out the form a submission should take, the categories of information expected, and the mechanics of the review period. These procedures give jurisdictions a known framework: they can see in advance what the Department regards as a complete file and how objections, withdrawals, and requests for reconsideration are processed. The existence of a published framework matters because it makes the administrative route predictable enough for routine use. A city clerk contemplating a polling place move or a county planning an annexation can consult the procedures, assemble the file, and submit with a reasonable sense of what comes next, rather than guessing at an opaque process.

An objection, when it comes, is a formal interposition by the Attorney General stating that the jurisdiction has failed to carry its burden. The statute frames the inquiry as the jurisdiction’s to satisfy, and the objection letter reflects that framing: it concludes that the change has not been shown to be free of discriminatory purpose or discriminatory effect. Objections are uncommon relative to the volume of submissions, but they are the enforcement backbone of the administrative route, and an objection freezes the proposed change in place. The jurisdiction then faces a choice. It may withdraw or modify the proposal and resubmit. It may seek reconsideration of the objection within the Department. Or it may turn to the second route, filing a declaratory judgment action and asking a court to grant the preclearance the Attorney General withheld.

The Department’s procedures also provide a route back from an objection. A jurisdiction that believes the objection rests on an incomplete record or a mistaken premise may request reconsideration, supplying the additional information or analysis it thinks the Department missed. Reconsideration does not always succeed, and it consumes time, but it offers a cheaper remedy than litigation for objections that turn on facts rather than on legal principle. Between reconsideration, withdrawal and resubmission, and the judicial route, the administrative system thus gives jurisdictions a sequence of escalating options: fix the file, fix the change, or take the dispute to court.

The administrative route also permits withdrawal before a decision. A jurisdiction that realizes its submission is weak or that anticipates an objection can pull the change back, revise it, and start again. This safety valve is frequently used, and it means the count of objections understates the practical bite of the review: some changes never see the light of day because the jurisdiction, knowing the standard, decides the proposal cannot survive scrutiny.

The Judicial Route: Declaratory Judgment in the District of Columbia

The second route runs through the courts. A covered jurisdiction may file suit seeking a declaratory judgment that its proposed voting change satisfies the statutory standard. The suit is brought before a three judge court in the United States District Court for the District of Columbia, with appeal lying directly to the Supreme Court. This is a deliberate structural choice. Concentrating the judicial route in a single federal district, and one located in the nation’s capital, channels the cases to a bench with accumulated experience in voting matters, and direct appeal to the Supreme Court accelerates final resolution of questions that would otherwise climb the ordinary appellate ladder.

The judicial route is, in formal terms, a de novo proceeding. The court does not defer to the Attorney General’s prior handling of the change, and indeed the route is available whether or not the jurisdiction has first sought administrative review. A jurisdiction may go straight to the District of Columbia court without ever submitting to the Department of Justice. In practice, however, the judicial route was used far less often than the administrative one, and the reasons are straightforward. Litigation before a three judge court is slow. It is expensive. It requires the jurisdiction to assemble counsel, develop an evidentiary record, and litigate against the United States as defendant. For the routine run of voting changes, polling place relocations, annexations of small parcels, adjustments to registration hours, the administrative route offered a faster and cheaper path, with the added advantage that silence at the end of the review period operated as approval. The judicial route was reserved, as a practical matter, for changes where the stakes or the legal questions justified the investment: major redistricting plans in the wake of a census, controversial changes that had drawn or were likely to draw an objection, and proposals raising questions of first impression about the reach of the standard.

The proceeding itself resembles ordinary civil litigation in its evidentiary demands. The jurisdiction presents its case through documents, demographic analysis, and testimony about the change’s design and expected consequences; the United States responds with its own evidence and argument. Because the burden rests on the jurisdiction throughout, the United States need not prove discrimination to prevail. It need only show that the jurisdiction has failed to carry its affirmative burden. This allocation shapes litigation strategy on both sides: the jurisdiction’s case is built to demonstrate compliance, while the government’s case is built to expose gaps in that demonstration.

The jurisdiction, as plaintiff, must prove to the court that the proposed change has neither a discriminatory purpose nor a discriminatory effect. The United States, represented by the Attorney General, participates as the defendant and may present evidence that the change fails the standard. The three judge court issues findings of fact and conclusions of law, and either party may appeal directly to the Supreme Court, which must consider the case under its direct appeal jurisdiction.

One procedural consequence of the burden placement deserves emphasis, because it reverses the usual posture of civil rights litigation. In ordinary discrimination cases, the plaintiff alleges discrimination and must prove it. Under Section 5 preclearance, the jurisdiction steps forward as the moving party and must affirmatively demonstrate the absence of discrimination. The Supreme Court upheld this arrangement in South Carolina v. Katzenbach, 383 U.S. 301 (1966), as a reasonable response to the record of evasion that had prompted the preclearance mechanism in the first place: jurisdictions that had repeatedly defeated case by case litigation could fairly be required to show their work before implementing new rules.

4. What counts as a voting change, and the standard applied

The reach of Section 5 preclearance turns on a deceptively simple phrase: the statute covers changes in voting qualifications, prerequisites, standards, practices, or procedures. For a jurisdiction trying to decide whether a proposed action needs preclearance, everything rides on the breadth of that phrase. If an action is a voting change, it must be submitted before it can take effect, and implementing it without preclearance exposes the jurisdiction to suit by the Attorney General or by private parties. If it is not a voting change, preclearance is unnecessary and the action stands on its own. The boundary between the two categories generated decades of litigation, because the consequences of the classification were so stark.

The Broad Reading: Allen v. State Board of Elections

In 1969, the Supreme Court addressed the scope question directly in Allen v. State Board of Elections, 393 U.S. 544. The case arose from challenges to several Mississippi and Virginia practices, including changes to election procedures that the states had not submitted for preclearance. The states argued, in essence, that Section 5 preclearance reached only the kinds of rules the coverage formula had originally targeted, the literacy tests and similar voter registration devices whose use had triggered coverage. On that reading, the preclearance obligation would have been confined to the machinery of voter registration and ballot access.

The Court rejected the narrow reading. It held that Congress had intended the preclearance requirement to reach any change in voting qualifications, prerequisites, standards, practices, or procedures, read broadly. The opinion reasoned that the evil Congress sought to address was not limited to the specific devices that had triggered coverage; it was the inventiveness of jurisdictions determined to preserve minority disenfranchisement by shifting tactics whenever an old device was struck down. A preclearance regime limited to registration rules would have left untouched the very ingenuity it was designed to police. Under the broad reading announced in Allen, Section 5 preclearance reaches annexations that alter the composition of an electorate, redistricting plans that redraw the lines of representation, shifts from district based to at large election schemes, changes in polling places and their locations, and alterations to registration procedures, all of them changes that affect voting in ways far removed from the literacy tests that had put the jurisdictions under coverage.

The Court soon applied the broad reading to specific categories of change. In Georgia v. United States, decided in 1973, it held that a city’s annexation of surrounding territory was a voting change subject to preclearance, confirming that alterations to the composition of the electorate fall within the statute even when no rule about voting procedures is rewritten. In City of Rome v. United States, decided in 1980, the Court sustained the requirement that changes satisfy both the purpose and the effect inquiries, rejecting the argument that Congress could only reach changes adopted with a discriminatory intent. Together these decisions locked in the two pillars of the doctrine: the reach is broad, and the standard is demanding.

Allen also rejected the idea of a de minimis exception. The Court held that the preclearance obligation applies to changes however small, reasoning that a regime designed to catch evasive ingenuity could not function if jurisdictions could self certify that their changes were too minor to matter. The practical consequence was a steady stream of submissions covering adjustments that, viewed in isolation, might seem trivial: a moved polling place, a shifted precinct boundary, a modest change in registration hours. Each had to be submitted, because the statute trusted the review process, not the jurisdiction’s own assessment, to decide whether a change was consequential.

The practical effect of Allen was to convert Section 5 from a checkpoint on registration rules into a comprehensive review of electoral structure. A covered jurisdiction that annexed a neighboring area, for example, changed who could vote in its elections even though no rule about voter eligibility had been rewritten; after Allen, that annexation was a voting change subject to preclearance. A city that moved from electing its council by wards to electing it at large changed how votes were translated into representation even though every registered voter could still cast a ballot; after Allen, that too required preclearance. The decision made the scope question the threshold question in nearly every Section 5 dispute, because once an action qualified as a voting change, the burden shifted to the jurisdiction to defend it.

Why did most preclearance disputes turn on whether something was a voting change?

Because the classification was dispositive. If a practice was a voting change, the jurisdiction had to preclear it before enforcement and bore the burden of proving it nondiscriminatory; if it was not, Section 5 did not apply at all, and challengers had to proceed under ordinary litigation where they carried the burden of proof.

That asymmetry explains why so much of the case law orbits the definitional question. For a covered jurisdiction, the incentive was to characterize an action as something other than a voting change: a routine administrative decision, an internal personnel matter, a budgetary adjustment, a change in the form of government. For the Department of Justice and for private plaintiffs, the incentive ran the opposite way: to show that the action fell within the broad ambit Allen had established. Courts deciding these disputes had to determine, on the facts of each case, whether the jurisdiction’s characterization held or whether the action, whatever its label, altered the effective exercise of the franchise.

The disputes took on a recognizable pattern. A jurisdiction would implement a change without submitting it, taking the position that preclearance was unnecessary. The Department or a private plaintiff would sue to enforce the preclearance requirement, arguing that the change fell within Section 5’s reach. The threshold litigation then consumed the question of coverage: was the action a change in a voting standard, practice, or procedure? Only if the answer was yes did the court reach the substantive question of whether the change was discriminatory. This sequencing meant that many Section 5 cases were won or lost before anyone discussed discrimination on the merits. A jurisdiction that persuaded the court its action was not a voting change escaped the preclearance regime entirely; a jurisdiction that lost the threshold fight then had to satisfy the retrogression standard with the burden on its own shoulders.

The Supreme Court continued to refine the boundary in the decades after Allen, and the refinements cut in both directions. Some decisions confirmed that seemingly peripheral actions fell within the statute’s reach, reinforcing the breadth of the Allen reading. Others held that certain changes, such as alterations in the structure of government that did not themselves touch the electoral process, fell outside it. The through line was functional: the question was not what the jurisdiction called the change, but what the change did to the electoral process in practice.

The Retrogression Standard: Beer v. United States

Once an action qualifies as a voting change, the governing question is whether the change satisfies the statutory standard, and the Supreme Court’s definitive statement of the effect prong came in Beer v. United States, 425 U.S. 130 (1976). The case concerned a redistricting plan for the city of New Orleans. The question before the Court was what the statute’s prohibition on discriminatory effect actually meant: did it require that a change affirmatively improve the position of minority voters, did it require only that the change not discriminate in some absolute sense, or did it set some other benchmark?

The Court adopted the retrogression standard. Under Beer, a voting change has a discriminatory effect if it leaves minority voters worse off than they were under the practice the change replaces. The comparison is to the status quo ante, the benchmark practice in force before the proposed change. A change that preserves or improves minority voting strength is not retrogressive. A change that diminishes it is. The standard does not demand that every new plan maximize minority representation, and it does not ask whether the existing practice was itself fair. It asks a narrower, comparative question: does the new practice make things worse?

This comparative structure shapes how both the jurisdiction and the reviewer approach the analysis. The first step is to identify the benchmark. The benchmark is the voting practice in force and properly precleared, the practice against which the proposed change will be measured. Getting the benchmark right is essential, because the wrong benchmark corrupts the entire comparison. If a jurisdiction measures its proposal against a practice that was never legally in effect, the comparison tells the reviewer nothing about whether the change retreats from the actual status quo.

Benchmark selection itself generated disputes. The benchmark is ordinarily the last legally enforceable practice, the plan or procedure in effect and properly precleared before the proposed change. Complications arise when the existing practice was itself adopted without preclearance, when a court ordered plan stands in for a legislatively adopted one, or when demographic shifts have rendered the old plan’s numbers stale. In each situation the question is which state of affairs genuinely represents the status quo ante against which backsliding must be measured. The resolution of that question can determine the outcome before any comparison is performed, which is why benchmark disputes occupied a significant share of the substantive litigation.

The second step is to measure minority voting strength under the benchmark and under the proposal. In the redistricting context, where retrogression analysis is most developed, this typically means examining the districts in which minority voters have the opportunity to elect candidates of their choice, and asking whether the proposed plan reduces that opportunity relative to the benchmark. The analysis is concrete rather than abstract: it looks at district by district results, demographic composition, and the history of electoral outcomes to determine whether the proposed plan curtails what minority voters have actually been able to achieve at the ballot box.

The measure of minority voting strength, in the redistricting cases that dominated the doctrine, is the practical opportunity to elect candidates of choice. The inquiry looks past raw population figures to the reality of electoral behavior: whether minority voters, voting cohesively, have been able to elect their preferred candidates under the benchmark, and whether the proposed plan preserves that capacity. A plan that maintains the number of districts in which minority voters can elect candidates of choice is not retrogressive even if its demographics shift at the margins; a plan that eliminates or dilutes such districts is. The focus on practical electoral opportunity, rather than on abstract demographic ratios, gives the retrogression standard its distinctive character: it protects results that have actually been achieved, not theoretical shares that might have been.

The third step is the comparison itself. If the proposed change leaves minority voters in the same or a better position, it is not retrogressive, and the effect prong is satisfied. If it leaves them worse off, the change is retrogressive and preclearance must be denied on effect grounds, regardless of the jurisdiction’s stated intentions. The standard is thus protective of the status quo without being aspirational: it locks in existing minority voting strength against backsliding while leaving jurisdictions free to design new arrangements that maintain or improve that strength.

Congress addressed the purpose prong directly in the 2006 reauthorization. The Supreme Court had held in Reno v. Bossier Parish School Board, 528 U.S. 320 (2000), that the purpose inquiry under Section 5 asked only whether the jurisdiction acted with a retrogressive purpose, a purpose to make minority voters worse off, rather than with any discriminatory purpose in a broader sense. The 2006 amendments rejected that narrowing. A voting change adopted with any discriminatory purpose fails the preclearance standard, even when its effect is not retrogressive. The effect inquiry, and the benchmark comparison at its heart, continues to be governed by the Beer retrogression standard, so the framework for a submitting jurisdiction is twofold: show that the change was not adopted for a discriminatory purpose of any kind, and show through the benchmark comparison that it does not leave minority voters worse off.

Applying retrogression in practice requires judgment, and the doctrine acknowledges that judgment through its treatment of the totality of the circumstances. A plan that reduces minority voting strength in one district while increasing it in another must be assessed as a whole; the question is the net effect on minority voters’ opportunities across the jurisdiction. A change that appears retrogressive on demographic figures alone may be shown, through evidence of actual voting behavior, to preserve minority voters’ practical opportunities; conversely, a change that looks neutral on paper may be shown to diminish those opportunities in fact. The standard thus invites, and in close cases requires, an evidentiary record that goes beyond headcounts to the lived reality of elections.

For the jurisdiction preparing a submission, the retrogression standard supplies the analytical framework the supporting materials must address. The submission should identify the benchmark, present the before and after comparison, and explain why the change does not diminish minority voting strength. For the reviewer, whether in the Department of Justice or in the District of Columbia court, the standard supplies the question to answer: compare the proposal to the status quo ante, and ask whether minority voters come out worse. The elegance of the Beer standard lies in its relative administrability. It does not require the reviewer to decide what a perfectly fair electoral system would look like, a question on which reasonable people and reasonable courts will always disagree. It requires only a judgment about direction: forward, sideways, or back.

Together, the broad definition of a voting change and the retrogression standard define the working life of Section 5 preclearance. Allen tells jurisdictions what must be submitted; Beer tells them, and tells the reviewers, how the submission will be judged. The two doctrines interlock: the broad reach ensures that the comparison cannot be evaded by relabeling, and the comparative standard ensures that the broad reach does not become a demand for perfection. A jurisdiction need not show that its electoral arrangements are ideal. It need only show that they do not move backward.

5. Exemptions, thresholds, and escape hatches

The coverage formula was never meant to operate as a permanent sentence. From the outset, Congress paired the trigger with a set of exits and side entrances: a covered jurisdiction that compiled a genuinely clean record could earn its way out of coverage, and a court that found a proven constitutional violation could pull an uncovered jurisdiction in. Bailout and bail-in were the two great hinges of the system, and alongside them ran the thresholds of time itself, the temporary authorizations and the administrative clock that gave preclearance its rhythm. These mechanisms deserve close attention because they reveal what Section 5 was designed to be. It was not a standing federal receivership over the elections of the covered states. It was a conditional regime, one that endured only so long as the conditions justifying it persisted, and one that contained, from its first day, the procedures for its own partial dismantling: the conditions for entry, the conditions for exit, and the clock by which everything ran.

The exit was Section 4(a), the bailout provision. It allowed a covered state or political subdivision to file a declaratory judgment action in the United States District Court for the District of Columbia, asking the court to terminate the coverage that the formula had imposed. The burden of proof rested entirely on the jurisdiction seeking release, and the showing it had to make was exacting. Over the ten years preceding its filing, the jurisdiction had to demonstrate a record free of the practices the Act was written to reach: no test or device used with a discriminatory purpose or with a discriminatory effect on account of race, color, or membership in a language minority group; no final judgment of any court finding that it had denied or abridged the right to vote on such grounds; no consent decree, settlement, or agreement in which it had abandoned a challenged voting practice; and no assignment of federal examiners to observe or supervise its elections. It had to show complete compliance with Section 5 itself, meaning every covered change submitted for review, no objection left standing unreversed, and no submission left pending without a determination. And the statute demanded more than the mere absence of violations. The jurisdiction had to present affirmative evidence of constructive citizenship in election administration: genuine efforts to eliminate harassment and intimidation of voters, expanded opportunities for minority citizens to register and to vote conveniently, the appointment of minority citizens as election officials, and other concrete steps demonstrating that it had moved beyond the era the coverage formula was written to address. A successful bailout ended the matter entirely, releasing the jurisdiction from the special provisions as a package, preclearance included, rather than merely suspending individual obligations.

The burden was the point. Congress deliberately made the jurisdiction prove a negative across a full decade, and courts construed the criteria strictly, because bailout was meant to reward only the jurisdiction whose record affirmatively demonstrated that federal supervision was no longer necessary. A jurisdiction that had simply avoided detection, or that had complied only sporadically, could not carry the showing. The provision thus functioned as both an incentive and a test: it gave covered jurisdictions a reason to build the clean record the statute described, and it ensured that release went only to those that had.

The ten-year window was measured backward from the date of filing, which gave the provision a rolling and unforgiving character. A disqualifying event within the decade, a discriminatory test, an adverse judgment, an unreversed objection, defeated the suit, and the jurisdiction’s only recourse was to wait until the event aged out of the window and try again. The clock rewarded patience as much as virtue: a jurisdiction that stumbled could rehabilitate its record and return, but only after the statute’s full decade had run clean.

For decades, almost no jurisdiction walked through that door. Bailout suits were vanishingly rare, and the reasons were as practical as they were legal. A bailout action meant litigating in Washington before a specialized court, assembling a decade of election records, and carrying a demanding burden of proof on every element of the statutory test, all for a reward that many local officials regarded as abstract. The units of government that actually administered elections, the counties, cities, school boards, and special districts, had an additional reason to stay away. The statute’s description of the entities eligible to seek bailout was widely read as excluding many of them, so a subdivision nested inside a covered state saw little prospect of winning release on its own record while the state around it remained covered. Coverage was conditional in theory and nearly permanent in practice, and the escape hatch Congress had built gathered dust for a generation.

The doorway widened in 2009. In Northwest Austin Municipal Utility District No. 1 v. Holder, the Supreme Court took up the eligibility question and resolved it in favor of breadth, holding that every political subdivision, including small special-purpose districts like the Texas utility district that brought the case, could seek bailout on the strength of its own record. The Court’s reading meant that such a unit’s own decade of compliance, rather than the history of the state containing it, determined its eligibility. The Court decided the case on statutory grounds, construing the bailout provision to make the remedy available to subunits rather than only to the states and larger units that earlier practice had assumed, while noting the serious constitutional concerns that a narrower reading would have raised about the coverage scheme. The substantive standard did not change; the ten-year clean-record showing remained exactly as demanding as Congress had written it. What changed was access. A city, a county, or a utility district inside a covered state could now pursue exemption independently, measured against its own conduct rather than the conduct of the state around it. The escape hatch became, for the first time, a realistic option for the local governments that did most of the actual work of running elections.

If bailout was the exit, Section 3(c) was the side door. Known as bail-in, it empowered a court adjudicating a voting rights case to impose preclearance on a jurisdiction as part of its judgment, even where the coverage formula had never applied. The provision operated only after the court found a violation of the voting guarantees of the Fourteenth or Fifteenth Amendments. It was not a shortcut around proof of discrimination but a consequence of proven discrimination, and the constitutional predicate mattered: bail-in required a judicial finding of the kind of wrongdoing the Constitution itself prohibits, not merely a statistical trigger. Upon such a finding, the court could order that for a period it deemed appropriate, the jurisdiction submit its voting changes for federal approval before enforcing them, and the court could retain jurisdiction to supervise compliance and to hear disputes about whether particular practices fell within the order. The court’s order typically specified which practices were covered and how long the obligation would run, and retained jurisdiction allowed the court to modify the order as circumstances changed. The orders often emerged from cases in which plaintiffs had also challenged practices under Section 2 of the Act, with the court anchoring the preclearance remedy in its constitutional findings. Bail-in thus extended the preclearance obligation beyond the formula’s geography. A jurisdiction in a region the trigger dates had never captured could still find itself subject to the same duties if a court concluded that its record warranted them, with the duration and scope of the obligation tailored to the violation the court had found. Where the formula sorted jurisdictions by history, bail-in sorted them by adjudicated conduct, and the two mechanisms together meant that preclearance was neither strictly territorial nor strictly historical but remedial in both directions.

The third set of thresholds concerned time itself, and they expressed the most basic premise of the regime: preclearance was born temporary. Congress authorized the special provisions, the coverage formula together with the preclearance requirement and the examiner provisions, for five years, on the understanding that extraordinary federal intervention in the administration of state elections required periodic reexamination supported by fresh legislative findings. As each expiration approached, Congress chose to extend the provisions rather than let them lapse, in 1970, in 1975, in 1982, and again in 2006, updating the formula’s trigger dates and the length of the authorization as the legislative record warranted. The 1975 extension also carried the Act’s protections to language minority citizens, adding a new dimension to the coverage the formula imposed. The 1982 extension ran for 25 years, and the 2006 extension added another 25, yet each renewal preserved the formal character of the regime as a temporary measure, renewed only after Congress had compiled a new record of conditions in the covered jurisdictions. the history of those extensions is traced in the reauthorizations article.

Within that framework of temporary authorizations, the deadline that governed any single voting change was the administrative review clock. Once a covered jurisdiction submitted a change to the Attorney General, a 60-day period began to run, measured from the receipt of a complete submission. If the Attorney General requested additional information, the clock started anew when the response arrived, which gave the reviewing authority a practical means of ensuring that submissions were genuinely complete before the countdown ran. If the 60 days expired without an objection, the change was precleared by the statute’s own operation, and the jurisdiction could enforce it; silence was approval. An objection, when interposed, was not the end of the road. The jurisdiction could modify or abandon the change and resubmit it, or it could seek a declaratory judgment in the District Court for the District of Columbia that the change had neither a discriminatory purpose nor a discriminatory effect, obtaining judicial preclearance after a full de novo proceeding. A jurisdiction could also bypass the Attorney General entirely and file the declaratory judgment action in the first instance, trading the speed of administrative review for the formality of a trial. The two routes ran in parallel throughout the life of the regime, giving covered jurisdictions both a swift executive review and a complete judicial hearing as paths to lawful implementation, and the choice between them was itself one of the regime’s quiet thresholds.

In practice, the review was a paper proceeding with real stakes on both sides. The Attorney General was not required to wait out the full 60 days; a determination of no objection could arrive well before the clock ran, and many routine submissions were cleared quickly. When the Attorney General did object, the objection came in writing with the reasons stated, which gave the jurisdiction something concrete to answer. The jurisdiction could then ask the Attorney General to reconsider the objection in light of new information or a modified proposal, or it could take the judicial route. The written objection thus functioned less as a final verdict than as the opening of a negotiation conducted under the shadow of the declaratory judgment action, with the 60-day clock setting the tempo.

A final threshold worth noting is the effective reach of the obligation. The duty to preclear attached to changes put into effect after coverage began, even if those changes had been adopted earlier; what mattered was administration, not enactment. Coverage thus froze the electoral status quo as it stood when the formula’s trigger caught the jurisdiction, and required federal approval for everything after. Combined with the 60-day clock, this created the distinctive tempo of Section 5 administration. A covered jurisdiction planning an election lived by the submission calendar and built the review period into its preparations, because a change that had not cleared review by election day could not lawfully be used.

6. Enforcement and remedies

A preclearance regime is only as real as its enforcement, and Section 5 was given force at the point where it mattered most: the election itself. A voting change subject to preclearance that had not been approved had no legal effect. The Supreme Court established the principle in Allen v. State Board of Elections (1969), the decision that first gave Section 5 its full scope. Arising from challenges to election-law changes in covered states, Allen held both that the preclearance requirement reached any change affecting voting, however minor its drafters considered it, and that covered changes were unenforceable until the jurisdiction obtained approval through one of the statute’s two routes. The consequence was stark, and deliberately so. A redistricting plan, an annexation, a shift to at-large elections, or a relocated polling place that skipped the required review existed, for legal purposes, as though it had never been adopted. Elections conducted under such a change rested on void authority, and the officials who administered them did so at their legal peril. The rule applied with equal force to a change that had been submitted and met with an objection: an objected-to change was, for enforcement purposes, the same as a change never submitted at all.

What was the legal consequence of skipping preclearance?

A covered change implemented without preclearance had no legal force and was treated as a legal nullity for election purposes. Elections conducted under it could be challenged, and the Attorney General or affected private plaintiffs could sue in federal court to enjoin its enforcement until the jurisdiction obtained approval.

The enforcement suit was the mechanism that gave this principle practical bite. The Attorney General could bring an action to enjoin a covered jurisdiction from enforcing an unprecleared change, and private parties could do the same. Allen recognized an implied private right of action under Section 5, so voters harmed by the change were not required to wait for federal officials to move; the people most directly affected by an unlawful election practice could go to court themselves. This mattered enormously in practice, because it meant enforcement did not depend on the priorities or resources of any particular administration. These suits were brought in the District Court for the District of Columbia, generally before a three-judge court, with direct review available in the Supreme Court, a procedural channel that reflected the national importance Congress attached to the preclearance obligation. The relief was injunctive. A court could bar the jurisdiction from holding an election under the unapproved change, direct officials to revert to the last precleared practice, and hold a noncompliant jurisdiction in contempt if it proceeded in defiance of the order. The incentive structure was unmistakable: implement first and seek approval later, and the implementation would be stopped before it could govern a single ballot.

The voidness principle did not stop at the courthouse door on election eve. Where an election had already been conducted under an unprecleared change, the results rested on the same void authority, and challengers could seek judicial relief reaching those results. Courts handling such claims balanced the violation against the disruption of unwinding a completed election, but the underlying rule never wavered: a change that had not been precleared could not supply lawful authority for an election, whenever the challenge arrived.

The regime policed the boundary between approved and unapproved changes with equal care. A jurisdiction that had obtained preclearance for a redistricting plan could not administer a materially altered version of that plan on the theory that the approval carried over; the change actually put into effect had to be the change that was reviewed, and deviations required fresh review. Where officials doubted whether a modification counted as a change at all, the structure of the statute pushed strongly toward submission, because the cost of guessing wrong was the invalidation of the electoral machinery built on the guess. This is why covered jurisdictions developed the habit of submitting expansively, erring on the side of review rather than risking an enforcement action over a judgment call. The cautious administrator treated the preclearance requirement as covering everything plausibly within its reach, and the 60-day clock, short by design, made such caution affordable.

The calendar gave these abstractions their urgency. A covered jurisdiction planning an election worked backward from election day through the 60-day review period, submitting changes early enough that an objection would still leave time to revise, resubmit, or litigate. An objection that arrived late in the cycle could force the jurisdiction to scrap months of preparation and revert to the last precleared plan, redrawing districts or reprinting ballots on a compressed schedule. The savvy election administrator therefore treated the submission date, not election day, as the real deadline, and built the federal review into the critical path of every electoral change. Preclearance was not a postscript to local decisionmaking; it was a step in the decision itself.

The reversion remedy deserves a final word, because it showed how concretely the voidness principle bit. Directing officials to revert to the last precleared practice was straightforward for an administrative change like a polling place: the old location simply stayed in use. For a redistricting plan struck down at the enforcement stage, it could mean conducting the election under the prior precleared map, even one drawn for the previous decade, or under an interim plan ordered by the court while the jurisdiction went back through review. Either way, the jurisdiction lost control of the electoral ground rules it had tried to change unilaterally, which was precisely the deterrent the remedy was designed to be.

None of this left a jurisdiction that disagreed with an adverse decision without a forum. The District Court for the District of Columbia was the exclusive venue for the declaratory judgment action, which existed precisely for the jurisdiction convinced that an objection was mistaken: a full trial, de novo, on whether the change had a discriminatory purpose or a discriminatory effect, with the burden of proof resting on the jurisdiction and the court deciding the question afresh rather than deferring to the Attorney General’s review. And a jurisdiction chafing under coverage itself could pursue bailout under Section 4(a), asking the same court to end the obligation entirely. Enforcement under Section 5 was therefore severe without being arbitrary. Every coercive edge of the regime, the voiding of unprecleared changes and the injunctions that stopped them, was paired with a judicial forum in which the jurisdiction could contest the determination behind it, and the availability of those forums was part of what made the severity sustainable as law.

For the officials, advocates, and voters who had to navigate the system in practice, the moving parts reduced to a single operational question: for this type of change, what had to happen before it could lawfully take effect? The table below presents the article’s findable artifact, the preclearance decision tree, mapping the five recurring categories of voting changes against the route available for approval, the decisionmaker, the standard applied, and the consequence of proceeding without approval.

The preclearance decision tree

Type of change	Route available	Decisionmaker	Standard applied	Consequence of proceeding without approval
Redistricting plan	Administrative submission or declaratory judgment action	Attorney General or District Court for the District of Columbia	No discriminatory purpose; no retrogressive effect	Unenforceable; subject to injunction until approved
Municipal annexation	Administrative submission or declaratory judgment action	Attorney General or District Court for the District of Columbia	No discriminatory purpose; no retrogressive effect	Unenforceable; subject to injunction until approved
Shift from district to at-large elections	Administrative submission or declaratory judgment action	Attorney General or District Court for the District of Columbia	No discriminatory purpose; no retrogressive effect	Unenforceable; subject to injunction until approved
Polling-place relocation or consolidation	Administrative submission or declaratory judgment action	Attorney General or District Court for the District of Columbia	No discriminatory purpose; no retrogressive effect	Unenforceable; subject to injunction until approved
Change to voter-registration procedures	Administrative submission or declaratory judgment action	Attorney General or District Court for the District of Columbia	No discriminatory purpose; no retrogressive effect	Unenforceable; subject to injunction until approved

Read together, the tree and the enforcement cases tell a single story about how Section 5 worked as law rather than as aspiration. The standard never varied with the type of change: purpose and effect, with effect measured against retrogression, applied to a polling place the same as to a redistricting plan. The routes never varied either, and neither did the consequence. That uniformity was the point. Preclearance did not grade voting changes by importance or presume that small administrative adjustments were harmless; it required that all of them, great and small, pass through the same federal review before they could govern an election. The severity of the remedy, voidness backed by injunction, was what made the review meaningful, and the availability of judicial forums was what made the severity legitimate. Together they formed the enforcement equilibrium on which the entire preclearance regime rested.

7. Provisions most often misread

Preclearance is among the most summarized and least understood mechanisms in American election law. Its reputation has traveled much farther than its text, and in the traveling, three misreadings have hardened into conventional wisdom. Each deserves to be met at full strength, because each carries a grain of truth large enough to explain its survival, and each, left uncorrected, distorts everything built on top of it.

The map was smaller than the legend. At full strength, the first misreading runs like this: the Voting Rights Act was a national statute answering a national disgrace, so its centerpiece must have operated nationally. General explainers describe the Act as placing American elections under federal supervision. Textbooks write that jurisdictions could not change their voting rules without federal approval. From there it is a short step to picturing a single checkpoint through which every ballot measure, district line, and polling place in the country had to pass, a kind of federal permit office for democracy itself. The logic feels airtight, and it flatters both sides of every argument about the law: supporters get a sweeping national safeguard, critics get a sweeping federal intrusion, and neither side has much incentive to check the statute.

The statute never built that office. Section 5 applied only to covered jurisdictions, and coverage was defined by the formula in Section 4(b), which sorted jurisdictions by their past rather than their promises: whether they had used a test or device as a prerequisite to voting, and whether fewer than half of their voting-age residents had registered or turned out in the presidential elections of 1964, 1968, or 1972. The jurisdictions caught by that formula were specific and nameable: nine states in their entirety, Alabama, Alaska, Arizona, Georgia, Louisiana, Mississippi, South Carolina, Texas, and Virginia, along with counties and townships scattered through a handful of other states. Everywhere else the preclearance duty simply did not exist. A county in a noncovered state could redraw its commission districts, relocate every polling place, and change its method of election without sending a single submission to Washington, and nothing in Section 5 gave federal reviewers a basis to ask for one.

The misreading matters because it flattens an extraordinary remedy into an ordinary one. Preclearance was not routine federal regulation of elections. It was a targeted suspension of the usual American presumption that states administer their own elections, imposed only where Congress found the historical record severe enough to reverse the normal order of things. And it came with its own adjustments, which the national-checkpoint image also erases: bailout under Section 4(a), by which a jurisdiction with a sustained clean record can exit coverage, and bail-in under Section 3(c), by which a court can impose the preclearance duty on a proven violator anywhere in the country. Read the map correctly and the mechanism looks like what it was: a cordon drawn around a defined territory, with a gate out and a gate in, not a blanket thrown over the nation. Arguments that treat it as a blanket, whether deployed to praise its reach or to condemn its overreach, are arguing with a law Congress did not write.

The burden pointed the unfamiliar way. The second misreading is the more consequential of the three because it reverses the mechanism’s procedural and moral polarity. Stated at full strength: preclearance worked like any other enforcement proceeding. The government suspected a voting change was discriminatory, the government carried the burden of proving it, and the jurisdiction stood as the defendant, entitled to the benefit of the doubt. This is how enforcement works nearly everywhere else in American law, from criminal prosecution to regulatory action, so the assumption imports itself without anyone noticing. It also tells a comforting story in which federal power is always the constrained party and local governments the rights-bearers, a story that makes preclearance sound far more conventional, and far less intrusive, than it was.

Section 5 ran the other way, and deliberately so. The submitting jurisdiction bore the burden of proving that its proposed change had neither a discriminatory purpose nor a discriminatory effect, with effect measured by the retrogression standard the Supreme Court set in Beer v. United States: the question was whether the change left minority voters worse off than the existing practice did. The Department of Justice did not have to build a case for discrimination. It had sixty days to interpose an objection, and if it did not, the change was precleared by the silence itself. A jurisdiction seeking clearance could bypass the Department entirely and seek a declaratory judgment from the United States District Court for the District of Columbia, but the polarity stayed the same in that forum too: the jurisdiction was the moving party, the one with something to prove, asking a court to declare that its change cleared the statutory bar.

Congress designed the inversion for a reason rooted in the record before 1965. Case-by-case litigation, with the burden on the government or on private plaintiffs to prove discrimination after each new barrier had already taken effect, had been outrun for the better part of a century by jurisdictions that could invent a replacement device faster than courts could strike down the old one. A literacy test fell and a poll tax rose; the tax fell and a white primary took its place; each victory arrived after the election it was meant to protect had already happened. Pausing the change until clearance, and making the changer prove the negative, was the whole innovation, the break with a century of losing races against the calendar. Miss the flip and you miss the mechanism. You will picture a cautious federal agency hunting for proof of discrimination, when the statute actually pictured a jurisdiction obliged to demonstrate its own compliance before acting. Every debate about whether preclearance was fair to covered jurisdictions, or fair to the voters it protected, has to start from the direction the burden actually ran, because the familiar direction produces the familiar arguments, and the familiar arguments answer a different law.

Section 5 was not Section 2 in a hurry. The third misreading collapses the Act’s two great engines into one, and at full strength it sounds almost scholarly: Sections 5 and 2 are two procedural routes to the same substantive destination, the ban on discriminatory voting practices. Section 5 is the administrative shortcut, Section 2 the courtroom long way around, and the distinction is one of forum and timing rather than of law. The phrasing is tidy, which is why it survives in classrooms and newsrooms and in the mouths of commentators who should know better. It is wrong in every dimension that matters, and the tidiness is the tell: real statutes are rarely that symmetrical.

Section 2 is permanent, nationwide, and prohibitory. It forbids any voting practice that results in the denial or abridgment of the right to vote on account of race, it is enforced through litigation brought after a practice takes effect, and since the 1982 amendments its test has been results-based: a court examines the totality of the circumstances to decide whether minority voters have less opportunity to participate in the political process and elect representatives of their choice, without requiring proof of discriminatory intent. Section 5 is temporary, geographically confined, and administrative, reauthorized in 2006 for an additional twenty-five years. It forbids covered jurisdictions from enforcing any change in voting practice until the change has been precleared, and the preclearance inquiry asks two narrower questions: does the change have a discriminatory purpose, and is it retrogressive.

The two sections thus ask different questions, at different moments, in different forums, under different standards, with different lifespans, and the practical upshot is that the standards do not collapse into each other no matter how often they are described as one. A voting change that merely preserves existing minority electoral strength satisfies Section 5’s retrogression test even though the underlying system might still be challenged under Section 2’s broader results inquiry, which can attack the status quo itself rather than just changes to it. Conversely, the Supreme Court has held that a change’s vulnerability under Section 2 does not by itself supply the basis for a Section 5 objection. The two provisions can point in different directions on the same facts, and that is by design: one freezes backsliding within a defined territory, the other polices discriminatory results everywhere, and neither was written to do the other’s job. Treat the two sections as one and you will misjudge both what preclearance could do and what it could not. It was a brake on retrogression where Congress had found the risk highest, not a national guarantee of optimal fairness. The nationwide guarantee, to the extent the Act provides one, lives in Section 2 and always has.

8. The honest complication: preclearance rarely said no

Here is the fact that complicates every account of preclearance, including the admiring ones. The mechanism that supporters describe as the Act’s most powerful provision, and that covered jurisdictions experienced as a standing federal veto over their election laws, almost never issued a veto. Since 1965 the Department of Justice received 120,868 Section 5 submissions and interposed 1,401 objections, a rate just over one percent, according to testimony of a former Acting Assistant Attorney General for Civil Rights cited in the 2006 report of the United States Commission on Civil Rights on the Voting Rights Act. The reauthorization record tells the same story in miniature: in the ten years before the 2006 reauthorization hearings, the Department interposed only 37 objections, an annual rate below two-tenths of one percent, as shown in the hearing record. The National Commission on the Voting Rights Act, organized by the Lawyers’ Committee for Civil Rights Under Law, reported 626 objections between 1982 and 2004, about four per month across a period in which submissions arrived by the thousands. Any honest treatment of preclearance has to begin by sitting with these numbers rather than hurrying past them, because they are the ground on which the sharpest arguments about the mechanism’s legitimacy have been fought, and they will not be waved away.

Why did objections fall below one percent of submissions?

Because most submitted changes were genuinely minor, covered jurisdictions learned to screen proposals before filing, and more-information requests prompted hundreds of quiet withdrawals, so the low objection rate reflects routine approvals and pre-filing discipline as much as it reflects the quiet administrative character of the review.

Part of the answer lies in what the submission stream actually contained. The tens of thousands of filings were not tens of thousands of dramatic showdowns over minority voting strength. The great majority were routine administrative adjustments: annexations, polling place relocations, precinct boundary tweaks, changes in filing deadlines and office hours. A jurisdiction that wanted to move a polling place across the street filed a submission; a city that annexed a subdivision filed a submission; the sixty-day clock ran, no objection issued, and the change took effect. A review process that handles that kind of volume will clear the overwhelming majority of what crosses its desk for the same reason a building inspector approves the overwhelming majority of permits: most of what is submitted was never problematic. The objection rate, taken alone, measures the share of proposals that were both submitted and found retrogressive or purposeful, and it says nothing about the proposals that were never submitted at all.

The withdrawal evidence is where the low objection rate stops being the whole story. Fraga and Ocampo studied the Department’s use of more-information requests, the formal device by which reviewers ask a submitting jurisdiction for additional material before the sixty-day clock runs out, and found that over a six-year period, 262 proposed voting changes were withdrawn or modified after such a request. The study also identified 95 submissions that drew more-information requests and then no response at all from the jurisdiction, cases of silent abandonment, and the authors conservatively excluded those 95 from their deterrence estimate rather than counting them as evidence that scrutiny had worked. The evidentiary limits deserve plain statement, because they cut against the temptation to treat withdrawals as hidden objections. A withdrawn submission is not an adjudicated act of discrimination. Jurisdictions may withdraw changes to avoid the cost and delay of extended review, to rework proposals for entirely nondiscriminatory reasons, or because the political coalition behind a change dissolved on its own. The study measures the shadow of the mechanism rather than its direct hits, and shadows are inherently harder to count than objections. What it establishes, within those limits, is that the objection tally understates the mechanism’s footprint: hundreds of proposed changes never reached the point of a formal yes or no because scrutiny arrived earlier in the process, at the information-request stage, and the proposal did not survive the encounter.

Now give each side its strongest reading, in its own voice, because both readings take the same numbers seriously and they cannot both be right about what the numbers mean. The critics’ reading: a one-percent objection rate is the regime’s own verdict on itself. If covered jurisdictions were still systematically manipulating election rules against minority voters, the objection rate would not sit below two-tenths of one percent per year in the decade before reauthorization. The coverage formula sorts jurisdictions by the election returns of the 1960s and early 1970s, which means the preclearance duty tracks geography and history rather than present conduct, and a remedy that fires one time in a hundred has become a ritual of submission rather than a shield against discrimination. On this view the Fraga and Ocampo withdrawals prove little, because a jurisdiction that abandons a proposal after a federal information request has not been found to have done anything wrong, and counting non-objections as evidence of deterrence turns the absence of findings into a finding. If the patient has been healthy for decades, the critics ask, why is the quarantine still in force.

The defenders’ reading: the low rate is exactly what a working preclearance regime looks like, and the critics have the causal arrow backward. A checkpoint that deters does not need to interdict often; the jurisdictions that knew their proposals would not survive review stopped submitting them, or reshaped them before filing, which is why the withdrawal figures matter and why the critics’ arithmetic is upside down. On this view the 262 withdrawn or modified changes are the mechanism’s real output, the formal objections are the residue of cases where deterrence failed, and judging preclearance by its objection rate alone is like judging a dam by the number of floods it failed to stop. The formula’s backward look is not a bug but the point: the jurisdictions with the worst histories are precisely the ones whose proposals warrant screening, and the decades of low objection rates are the payoff of the screening, not proof of its irrelevance.

The evidence, taken together, supports two modest conclusions and resists a third. It supports that formal objections were rare, across every period and every source that has counted them, and it supports that pre-objection withdrawals were numerous, in the hundreds over a six-year window by one careful count. It does not by itself establish whether those withdrawals represent deterred discrimination or merely deterred paperwork, and that is the honest limit of what the record, as it stands, can tell us. Readers who want a different measure of the mechanism’s footprint, one taken in voters rather than in filings, should know that what the mechanism produced in registration and turnout is measured in the impact article.

9. Closing: the three-question preclearance test

Strip away the commentary and every preclearance question reduces to three smaller questions asked in order: (1) was the jurisdiction covered, (2) was the action a change in voting practice, (3) was the change retrogressive. The first question is answered by the Section 4(b) formula and the bailout and bail-in provisions that adjust its edges. The third is answered by the retrogression standard. And the disputes that look the most complicated, the fights over annexations and redistricting plans and polling place consolidations and shifts from district to at-large elections, almost always turn on the second question: whether what the jurisdiction did counts as a change in voting practice at all, and therefore whether the preclearance duty attached in the first place. An annexation that adds new voters to a city is a change; a redistricting that redraws the lines is a change; a polling place moved across town can be a change; and each had to be filtered through that threshold before anyone reached the question of purpose or retrogression. Learn to ask the second question first and the mechanism stops looking mysterious. It becomes a gate, and like any gate it only matters if you can tell what is trying to pass through it.

The statute’s architecture can then be held in a single view. Section 4(b) draws the map, defining which jurisdictions carry the preclearance duty by reference to their history. Section 5 imposes the duty itself, barring covered jurisdictions from enforcing voting changes until those changes have been precleared. Section 4(a) provides the exit, allowing jurisdictions with sustained clean records to bail out of coverage and return to the ordinary presumption. Section 3(c) provides the court-ordered entrance, allowing judges to bail in violators anywhere in the country, covered or not, when litigation has proven the need. Map, duty, exit, entrance: four provisions, one mechanism, and the whole of it designed around a single insight, that some barriers to the ballot are erected faster than lawsuits can dismantle them, so the law must sometimes arrive before the barrier does. Keep the three questions and the four provisions straight, and the rest of the cluster will read as elaboration rather than as surprise.

10. Study: keeping the mechanism straight

A reader working through this cluster will accumulate a tangle of distinctions that resist casual memory: which section draws the map and which imposes the duty, which standard asks about backsliding and which about the totality of circumstances, which provisions open the coverage and which close it, which questions belong to the courtroom and which to the sixty-day administrative clock. The way to keep them straight is to keep a running section map beside the reading, one page per provision, with the citations that anchor each claim and the chronology of reauthorizations that keeps the temporary provisions in their proper tense. That is the kind of document that repays the effort of building it once and consulting it often, because every later article in the cluster will assume these distinctions are already in hand, and the cost of re-deriving them each time is the slow erosion of the reading itself.

keep your statute notes, citations, and case chronologies together free on VaultBook, where a dedicated legislation notebook gives each section its own page and each citation its own line, so the map, the duty, the exit, and the entrance stay sorted as the cluster grows. The habit is simple: read with the notebook open, file each provision under its number, note the standard each one applies and the forum each one uses, and let the chronology accumulate entry by entry. By the time the cluster is finished, the notebook will be the more valuable of the two documents, because it will be the one written in the reader’s own hand, and the distinctions that now require effort will have become the furniture of the mind.

Frequently Asked Questions

Q: What is preclearance under the Voting Rights Act?

Preclearance was the Voting Rights Act’s requirement that certain state and local governments obtain federal approval before putting any new voting rule into effect. Instead of waiting for a discriminatory law to harm voters and then suing after the fact, Congress reversed the usual burden: a covered jurisdiction had to prove to the federal government that a proposed change had neither a discriminatory purpose nor a discriminatory effect before it could be enforced. Approval came through one of two routes, administrative review by the Attorney General or a declaratory judgment action in federal court in Washington, D.C. The idea was to freeze the status quo in places with the worst histories of voting discrimination so that new barriers could be stopped before Election Day rather than litigated for years afterward.

Q: What did Section 5 of the Voting Rights Act require?

Section 5 required every jurisdiction covered by the Act’s coverage formula to submit any new or changed voting law, practice, or procedure to the federal government and receive approval, called preclearance, before enforcing it. A covered state, county, or city could not implement the change, even for a single election, until the Attorney General declined to object or a federal court declared it nondiscriminatory. The provision applied to the full range of voting changes, from redistricting plans and annexations to polling place moves and new registration rules. The burden of proof rested on the jurisdiction, which had to show the change had neither a discriminatory purpose nor a discriminatory effect. Enforcement was suspended while review was pending, so the federal government held a genuine gate over election administration in covered places.

Q: Which states were covered by the Voting Rights Act coverage formula?

At enactment in 1965 the formula covered Alabama, Alaska, Georgia, Louisiana, Mississippi, South Carolina, and Virginia in their entirety, along with counties in Arizona, Hawaii, Idaho, and North Carolina. Congress later widened the net in two steps: the 1970 amendments added jurisdictions meeting the formula measured against the 1968 election, and the 1975 amendments added those measured against the 1972 election plus a language minority expansion that brought in places using English only election materials where more than five percent of voting age citizens belonged to a single language minority group. Coverage therefore grew in stages rather than all at once, and each expansion reflected the same two part test, a test or device in use on the trigger date plus registration or turnout below half of the voting age population.

Q: What was Section 4(b) of the Voting Rights Act?

Section 4(b) was the coverage formula, the mechanical trigger that decided which jurisdictions had to obey the Act’s special provisions, including Section 5 preclearance. A state or political subdivision was covered if two conditions were both true: it maintained a test or device, such as a literacy test, as a prerequisite to voting on a specified trigger date, and fewer than half of its voting age residents were registered to vote or fewer than half voted in the presidential election of that year. Congress designed the formula to capture, without naming names, the jurisdictions with the deepest records of racial discrimination in voting. Section 4(b) did not itself ban anything; it switched on the Act’s strongest remedies, preclearance, the suspension of tests and devices, and the authorization of federal examiners, wherever its two prongs were satisfied.

Q: How did a jurisdiction get Voting Rights Act preclearance approved?

A jurisdiction seeking preclearance chose between two routes. It could submit the change to the Attorney General for administrative review; if the Attorney General did not object within sixty days of receiving a complete submission, the change was precleared. Alternatively, it could file a declaratory judgment action before a three judge panel of the federal district court in Washington, D.C., asking the court to declare the change nondiscriminatory. Under either route the burden rested on the jurisdiction to prove the change had neither a discriminatory purpose nor a discriminatory effect, measured against the retrogression standard. If the Attorney General objected, the jurisdiction could still seek a court judgment, since the court proceeding was fresh and independent. The system processed an enormous volume, more than 120,000 submissions since 1965, with objections in roughly one percent of cases.

Q: What is bailout under the Voting Rights Act?

Bailout, provided by Section 4(a), was the exit ramp from coverage. A covered jurisdiction could file suit in the federal district court in Washington, D.C., and win release from the Act’s special provisions, including preclearance, by proving a clean record over the preceding ten years. It had to show, among other things, that it had not used a test or device for a discriminatory purpose, that no court had found voting discrimination against it, that no federal examiners had been assigned, that it had complied fully with preclearance, and that it had taken constructive steps such as expanding minority registration and appointing minority election officials. Congress liberalized the provision in 1982, and after Northwest Austin Municipal Utility District No. 1 v. Holder in 2009 the courts confirmed that even small political subdivisions could bail out.

Q: What is the retrogression standard in Voting Rights Act cases?

The retrogression standard, announced by the Supreme Court in Beer v. United States in 1976, was the test for whether a voting change had a forbidden discriminatory effect under Section 5. The question was comparative: would the change leave minority voters worse off than they were under the existing practice, the benchmark? If the new plan diminished minority voting strength relative to the status quo, it was retrogressive and could be blocked; if it preserved or improved the minority community’s position, it passed, even if it did not maximize minority representation. Preclearance therefore policed backsliding rather than demanding ideal fairness. A redistricting plan that reduced the number of districts where minority voters could elect their preferred candidates, for example, would fail the test, while a plan that kept effective districts intact would satisfy it.

Q: What is Section 3 bail-in under the Voting Rights Act?

Section 3(c) allowed federal courts to impose preclearance on jurisdictions that were not covered by the Section 4(b) formula, a remedy nicknamed bail-in. When a court found that a state or political subdivision had violated the voting guarantees of the Fourteenth or Fifteenth Amendment, it could order the jurisdiction to preclear future voting changes for a period the court specified, essentially subjecting it to Section 5 style review as punishment and prevention. Unlike the formula’s automatic coverage, bail-in was discretionary and case specific, tailored to jurisdictions with proven records of intentional discrimination that the formula had missed. It was the mirror image of bailout: bailout let covered jurisdictions earn their way out of preclearance, while bail-in pulled uncovered jurisdictions in through a court order.

Q: What counts as a voting change under Section 5?

The Supreme Court read the phrase broadly. In Allen v. State Board of Elections in 1969 the Court held that Section 5 reached any enactment or practice that affected voting, not just laws that directly regulated the ballot. Covered changes included redistricting plans, annexations that altered a city’s electorate, switches from district based to at large elections, new candidate qualifications, changes in voter registration procedures, polling place relocations, and shifts in the form of local government. The touchstone was effect on the voting process: if a measure had the potential to affect who could vote, how votes were cast and counted, or the weight of minority votes, it required preclearance before enforcement. Even administrative decisions that looked minor on paper fell within the provision when they touched the mechanics of voting.

Q: How long did the Attorney General have to review a preclearance submission?

The Attorney General had sixty days from receipt of a complete submission to interpose an objection. The clock ran on the administrative review route only, and it started when the submission contained enough information for review. If the sixty days expired without an objection, the change was precleared by operation of law and the jurisdiction could enforce it. If the Attorney General objected within the period, the jurisdiction could not implement the change unless it later won a declaratory judgment from the federal court in Washington, D.C. A request for more information stopped the process and restarted the sixty day period once the jurisdiction responded. The deadline gave the Justice Department a fixed window to evaluate thousands of submissions while giving jurisdictions a predictable answer date.

Q: What is a test or device under the Voting Rights Act?

Section 4(c) defined a test or device as any requirement that a person demonstrate literacy, educational achievement, or knowledge of a subject, pass a test of good moral character, or produce vouchers from registered voters, as a condition of registering or voting. Literacy tests were the notorious example: registrars in the Jim Crow South used deliberately tricky or subjectively graded tests to reject Black applicants while waving white applicants through. Because these devices were the principal machinery of disenfranchisement, Congress made their presence the first prong of the coverage formula and separately suspended their use in covered jurisdictions. Congress later banned tests and devices nationwide. For coverage purposes, what mattered was whether the jurisdiction maintained such a requirement on the statute’s trigger date.

Q: Could a covered jurisdiction move a polling place without preclearance?

No. Moving a polling place was a classic voting change under Section 5 and required preclearance before the new location could be used. Where voters cast ballots directly affects access: a move can lengthen travel, confuse voters who arrive at the old site, or shift a polling place out of a minority neighborhood. The jurisdiction therefore had to submit the relocation to the Attorney General or the federal court in Washington and wait for approval, even when the reason seemed purely practical, such as a building being sold or damaged. Implementing the move first and seeking approval later violated the Act. Voters or the Justice Department could sue in the three judge district court in Washington to block an election conducted at an unprecleared site.

Q: Did Section 5 preclearance apply to redistricting plans?

Yes. Redistricting plans were among the most closely scrutinized voting changes under Section 5. Every new congressional and state legislative map drawn by a covered jurisdiction had to be precleared before it could be used in an election, which made the post census redistricting cycles major preclearance events. The Justice Department and the courts measured each plan against the retrogression standard, comparing minority voters’ ability to elect their preferred candidates under the proposed map with their position under the existing map. Plans that reduced that ability drew objections. Because line drawing so directly determines the weight of minority votes, redistricting generated some of the fiercest preclearance fights, and jurisdictions often submitted alternative plans after an objection rather than litigate.

Q: What is the difference between Section 5 preclearance and Section 2 litigation?

Section 5 and Section 2 were complementary enforcement tools with opposite designs. Section 5 applied only to jurisdictions covered by the formula, operated before a change took effect, and placed the burden on the government proposing the change to prove it was nondiscriminatory. Section 2 applied nationwide, operated after the fact through lawsuits, and placed the burden on challengers, usually private plaintiffs or the Justice Department, to prove that a voting practice resulted in discrimination under a totality of the circumstances test. Section 5 asked only whether a change made minority voters worse off than the status quo; Section 2 could strike down longstanding practices, not just new changes. In short, preclearance was a federal gate that stopped suspect changes at the door, while Section 2 was a courtroom remedy for discrimination already in place.

Q: Which elections set the Section 4(b) trigger dates?

The coverage formula measured jurisdictions against presidential elections. As enacted in 1965, Section 4(b) looked to the presidential election of 1964, with November 1, 1964 as the date for determining whether a test or device was maintained. The 1970 amendments added a second trigger tied to the presidential election of 1968, with November 1, 1968 as the measuring date. The 1975 amendments added a third trigger tied to the presidential election of 1972, measured as of November 1, 1972. In each round, a jurisdiction was covered if it maintained a test or device on that November date and fewer than half of its voting age residents were registered or voted in that year’s presidential election. Congress deliberately chose presidential elections because their high turnout made low participation a telling signal of suppressed registration.

Q: What is a declaratory judgment action under Section 5?

A declaratory judgment action was the judicial route to preclearance. Instead of submitting a voting change to the Attorney General, a covered jurisdiction could sue the United States in the federal district court for the District of Columbia, where a specially convened three judge panel would decide, in a fresh proceeding, whether the change had neither a discriminatory purpose nor a discriminatory effect. The burden of proof rested on the jurisdiction, and the court applied the same retrogression standard the Attorney General used. The action was also the fallback after an administrative objection: a jurisdiction that drew an objection could still obtain preclearance by winning in court. Appeals went directly to the Supreme Court. Although most jurisdictions chose the faster administrative route, the declaratory judgment action gave them a judicial forum with a full evidentiary record.

Q: What information did a preclearance submission have to include?

A preclearance submission had to give the Justice Department enough material to judge the change’s purpose and effect. Justice Department regulations required the text of the new law or a clear description of the practice, an explanation of the prior procedure it replaced, the reasons for the change, and its anticipated impact. For redistricting, that meant maps, population and voting age population broken down by race, and election data showing how minority voters fared under the old and proposed plans. Submissions routinely included registration and turnout figures by race, descriptions of minority participation in the political process, and any public comments. If the package was too thin to evaluate, the Attorney General could issue a more information request rather than guess, which restarted the sixty day review clock.

Q: Could voters challenge a precleared change in court?

Yes. Preclearance was not a seal of approval against every legal claim. A precleared change could still be challenged under Section 2, which let voters sue over any voting practice that produced a discriminatory result, even one the Attorney General had declined to block. Voters could also enforce Section 5 itself: if a covered jurisdiction implemented a change without obtaining preclearance at all, private plaintiffs or the Justice Department could sue before a three judge court in Washington to stop it. What voters could not do, under Morris v. Gressette in 1977, was ask a court to second guess the Attorney General’s decision not to object; that choice was committed to executive discretion and was not judicially reviewable. In practice, then, preclearance cleared only the Section 5 retrogression question, leaving Section 2 claims fully alive.

Q: What is a more information request in the preclearance process?

A more information request was the Attorney General’s tool for dealing with an incomplete submission. If the materials a jurisdiction sent in did not contain enough data to assess whether the change had a discriminatory purpose or effect, the Justice Department could write back asking for specific additional items, such as missing demographic tables, maps, or explanations, instead of issuing an objection on a thin record. The request was not a ruling against the change, but it had a procedural consequence: the sixty day review clock restarted when the jurisdiction supplied the requested information. Jurisdictions sometimes chose to withdraw a submission and refile a fuller one rather than answer piecemeal. The device kept the review honest by preventing approval, or objection, based on guesswork.

Q: What did it mean for a test or device to be maintained on the trigger date?

For coverage purposes, a test or device was maintained on the trigger date if the jurisdiction had it in force and in use as of that November 1 date, 1964, 1968, or 1972 depending on the round. The statute asked about the jurisdiction’s actual practice on the measuring date, not its later promises: a state that operated a literacy test for registrars on November 1, 1964 satisfied the first prong of the formula even if it repealed the test the following year. Paper repeals after the fact could not erase coverage, because Congress fixed the snapshot date to capture the machinery of disenfranchisement as it existed when the formula was drawn. Combined with the second prong, registration or turnout below half of the voting age population, maintenance of a test or device on the trigger date is what pulled a jurisdiction into preclearance.

Appendix B

ARTICLE 13 (voting-rights-act-section-5-preclearance)

vra|change|counts|voting # What counts as a voting change under Section 5?
vra|ag|long|preclearance|review|submission # How long did the Attorney General have to review a preclearance submission?
vra|device|test # What is a test or device under the Voting Rights Act?
vra|covered|jurisdiction|move|place|polling|preclearance|without # Could a covered jurisdiction move a polling place without preclearance?
vra|apply|plans|preclearance|redistricting # Did Section 5 preclearance apply to redistricting plans?
vra|difference|litigation|preclearance # What is the difference between Section 5 preclearance and Section 2 litigation?
vra|dates|elections|set|trigger # Which elections set the Section 4(b) trigger dates?
vra|action|declaratory|judgment # What is a declaratory judgment action under Section 5?
vra|include|information|preclearance|submission # What information did a preclearance submission have to include?
vra|challenge|change|court|precleared|voters # Could voters challenge a precleared change in court?
vra|information|preclearance|process|request # What is a more information request in the preclearance process?
vra|date|device|maintained|mean|test|trigger # What did it mean for a test or device to be maintained on the trigger date?