Concept
The Katzenbach Framework
Fifteenth Amendment enforcement power
The Katzenbach Framework
South Carolina v. Katzenbach (1966) was the Supreme Court’s first measurement of the Voting Rights Act of 1965 against Section 2 of the Fifteenth Amendment, and its reasoning remains the baseline for every later enforcement-power question. South Carolina challenged the Act in an original action, a suit filed directly in the Supreme Court rather than appealed from a lower court. The Act’s new remedies were three: a five-year suspension of literacy tests and similar voting qualifications, running from the last occurrence of substantial voting discrimination; a suspension of new voting regulations pending federal review; and the assignment of federal examiners to list qualified applicants.1 The Court sustained the new remedies.2
The opinion’s reasoning falls into three parts, and the subsections below take them in turn. The first sets the standard of review: Congress may use any rational means to enforce the Amendment. The second concerns subject matter: Congress may forbid practices that Section 1 would not itself condemn. The third concerns geography: Congress may confine its strongest remedies to the places where discrimination is concentrated. Together they mark the baseline of deference against which later decisions are measured. Shelby County v. Holder is the clearest example, because it held that Congress may not subject some States but not others to preclearance on the strength of decades-old coverage data.3 Seeing what Katzenbach allowed, and why, shows what Shelby County took away.
Rational Means Standard
Section 2 authorizes “appropriate legislation,” and a court cannot measure a statute against that word until it knows what the word requires.4 Katzenbach answered in one sentence: as against the reserved powers of the States, Congress may use any rational means to effectuate the constitutional prohibition of racial discrimination in voting.5 The Court drew that rule from the language and purpose of the Fifteenth Amendment, the prior decisions construing its provisions, and the general doctrines of constitutional interpretation, all of which pointed to the same principle.6
The sentence rests on an older test. Katzenbach held that the basic test in a Section 2 case is the same as in all cases concerning the express powers of Congress with relation to the reserved powers of the States, and it identified that test as the one formulated in McCulloch v. Maryland; Katzenbach v. Morgan later repeated the point.7 The McCulloch formulation reads: “Let the end be legitimate, let it be within the scope of the constitution, and all means which are appropriate, which are plainly adapted to that end, which are not prohibited, but consist with the letter and spirit of the constitution, are constitutional.”8
The formulation has two halves, and each does different work. The end must be legitimate and within the scope of the Constitution; under Section 2, the end is the prohibition of racial discrimination in voting. The means must be appropriate, plainly adapted to that end, not prohibited, and consistent with the letter and spirit of the Constitution. The Court had long described the Civil War Amendments in the same terms, treating as within congressional power whatever legislation is adapted to carry out their objects and tends to enforce submission to their prohibitions, if the Constitution does not prohibit it.9
The means half is generous because the Court refused to confine Congress to general commands. South Carolina argued that Congress may do no more than forbid violations of the Amendment in general terms and leave specific remedies, and their application to particular localities, to the courts. The Court called that an artificial rule. Congress’s power under Section 2, like its other powers, “is complete in itself, may be exercised to its utmost extent, and acknowledges no limitations, other than are prescribed in the constitution.”10 The structure of Section 2 explains the generosity. By adding the enforcement clause, the Framers indicated that Congress was to be chiefly responsible for implementing the rights that Section 1 creates.11 A court that picked the best remedy would be doing work the Amendment assigned to Congress.
Katzenbach shows the generous reading at work. Congress had found that case-by-case litigation was inadequate against widespread and persistent discrimination, because of the time and energy needed to overcome the obstruction these lawsuits invariably met. After nearly a century of resistance to the Amendment, the Court said, Congress might decide to shift the advantage of time and inertia from the perpetrators of the evil to its victims.12 A State that answered that the courts remain open could not defeat that choice, because the premise of the choice was that the courts had proved too slow. The Court then asked whether the specific remedies were appropriate means. It called the coverage formula “rational in both practice and theory,” treated the suspension of literacy tests as an uncommon but justified use of congressional power, and found exceptional conditions behind the freeze on new voting rules.13
Because the rule asks only whether the means are rational, a challenger who shows that a gentler remedy would have worked has not yet shown a defect. The court asks whether Congress could sensibly have chosen the means; it does not choose among remedies for Congress.14 Deference on means does not widen the ends Congress may pursue, however. On the rare occasions when the Court has held an enforcement statute unconstitutional, it reasoned that Congress had attacked evils the Fifteenth Amendment does not comprehend, as in United States v. Reese and James v. Bowman.15 In those cases the failure lay in the end Congress pursued, so the quality of the means never mattered.
The first step is where a challenger can win, since a statute aimed at something other than racial discrimination in voting attacks the end. The third is where challengers usually lose, since an argument that a gentler remedy would have sufficed attacks only the wisdom of the means. That deference reflected the record Congress built in 1965, and the Court later asked more of a reauthorization that rested on stale data, as the discussion of targeted remedies below shows.
Reaching Beyond Section 1
Congress may prohibit a practice that a court would not itself hold unconstitutional under Section 1, so long as the prohibition enforces Section 1. The rule depends on separating two propositions that sound alike: a court would hold the practice violates Section 1, and Congress may prohibit the practice as a means of enforcing Section 1. The first is a judicial conclusion about what the guarantee forbids. The second is a legislative judgment about what the guarantee needs to be effective. Because the enforcement power is remedial and prophylactic, the second can be true when the first is false: Congress may prohibit actions that do not directly violate the amendment if it reasonably believes they would or might lead to violations.17 Three decisions show how far the principle runs: the literacy-test suspension in Katzenbach, the effects-based bar upheld in City of Rome, and the nationwide ban upheld in Oregon v. Mitchell.
The 1965 suspension of literacy tests shows the principle most clearly. Lassiter v. Northampton County Board of Elections had held that a literacy test fair on its face and not employed in a discriminatory fashion did not violate Section 1.18 Yet Katzenbach upheld the Act’s per se ban on such tests. The Court recognized the prohibition as an appropriate method of enforcing the Amendment because, for many years, most of the covered jurisdictions had imposed literacy tests to effect voting discrimination, and continued use of even fairly administered tests would “freeze the effect” of past discrimination.19 The Act suspended these tests and similar devices for five years from the last occurrence of substantial voting discrimination, a response the Court called legitimate and well supported by precedent in Fifteenth Amendment cases.20
The freeze worked in three steps. Jurisdictions had allowed white illiterates to vote for years, so their rolls already held unqualified white registrants. A test applied evenhandedly from then on would screen only new applicants: an illiterate Black applicant would be turned away, while the white illiterates already registered would stay on the rolls. Congress knew that continuing the tests, however fairly administered in the future, would freeze the effect of past discrimination in favor of unqualified white registrants.21 Suspension was the measured remedy. The Court added that States that had let white illiterates vote could not sincerely complain of “dilution” when Black illiterates registered, and that Congress permissibly rejected a complete re-registration of all voters as too harsh on whites who had voted throughout their adult lives.22
City of Rome v. United States later drew the general rule from that holding: under Section 2, Congress may prohibit state action that is not itself a violation of Section 1 but perpetuates the effects of past discrimination.23 Lassiter therefore survives. A court applying Section 1 alone would still find no violation in a fair, evenhandedly administered test, while Congress, acting under Section 2, may ban the same test where it would carry the past forward.
City of Rome also carried the reasoning from lingering effects to present effects. It read Katzenbach’s treatment of the literacy-test ban to show that Congress may prohibit voting practices that have only a discriminatory effect.24 It upheld the 1975 seven-year reauthorization on the ground that Congress could rationally have concluded that electoral changes by jurisdictions with a demonstrable history of intentional racial discrimination in voting create the risk of purposeful discrimination, so that it was proper to prohibit changes with a discriminatory impact.25 Two links carry that reasoning. If Congress reasonably fears that a jurisdiction will commit intentional discrimination, it may remedy that discrimination in advance by banning a practice that would hurt minority voters, whether or not there is proof of discriminatory purpose. And the demonstrable history ties that preventive ban to Section 1, because a record of intentional exclusion makes a harmful change a likely vehicle for more of it.
The same logic reached beyond the covered jurisdictions. In Oregon v. Mitchell, the Court unanimously upheld a five-year nationwide ban on literacy tests and similar registration requirements, concluding that Congress could rationally have determined these provisions to be appropriate methods of attacking the perpetuation of earlier, purposeful racial discrimination, whether or not the banned practices were discriminatory only in effect.26 Even Justice Harlan, who took the most restrictive view of congressional power on the Warren Court, agreed that the expanded reach fell within Congress’s remedial powers. He reasoned that, despite the lack of evidence of specific instances of discriminatory application or effect, Congress could have determined that racial prejudice is prevalent throughout the Nation and that literacy tests unduly lend themselves to discriminatory application.27 A nationwide ban thus rests on a congressional judgment that the danger of discrimination in each place supports the prohibition without proof of it there.
The power is broad but bounded, and Oregon v. Mitchell also marks the boundary. Opinions in that case identified at least three limits on the enforcement clauses of the Civil War Amendments: Congress may not repeal other provisions of the Constitution; the power was not meant to strip the States of their power to govern themselves or to convert the national government of enumerated powers into one of unrestrained authority over the whole Nation; and Congress may only “enforce” the amendments by “appropriate legislation,” so it has no power to undercut their guarantees of personal equality and freedom from discrimination.28 The same discussion explained why the second limit gives way to remedies for racial discrimination. The division of power between state and national governments was expressly qualified by the Civil War Amendments’ ban on racial discrimination, so where Congress attempts to remedy such discrimination, its authority is enhanced by the avowed intention of the framers.29 A measure with no tie to racial discrimination earns none of that enhancement.
The most important limit on an exam is the one built into the word “enforce.” In construing the parallel power under the Fourteenth Amendment, the Court in Katzenbach v. Morgan emphasized that Congress’s power is limited to adopting measures that enforce the guarantees and gives it no power to restrict, abrogate, or dilute them. A statute authorizing States to maintain racially segregated school systems, the Court explained, could not be a measure “to enforce” the Equal Protection Clause, because the clause forbids such laws of its own force.30 The Fifteenth Amendment works the same way: a statute that authorized a State to deny the vote on account of race would weaken the guarantee, while the literacy-test ban left Section 1 untouched and added protection around it. Congress may reach beyond what the Court itself would condemn only in service of the guarantee, never in place of it.
Reaching beyond Section 1 also depends on a record that supports the connection. City of Rome rested on a demonstrable history of intentional discrimination, and Shelby County later held that Congress may not subject some States but not others to preclearance on the strength of decades-old coverage data.31 A statute that outruns Section 1 therefore needs two things: an aim of preventing or undoing racial discrimination in voting, and a showing, current enough to count, that the prohibited practice threatens to carry that discrimination forward.
Targeted Remedies and Equal Sovereignty
A second question is whether Congress may impose its strongest remedies on some States and subdivisions and spare the rest. Katzenbach answered yes. The Act intentionally confined its new remedies to a small number of States and political subdivisions, most of them familiar to Congress by name, and the Court held that this was a permissible method of dealing with the problem.33 Three points support the holding: the reason for targeting at all, the answer to the objection that States must be treated equally, and the rationality of the coverage formula, the bailout provision, and preclearance.
Congress had learned that substantial voting discrimination presently occurred in certain sections of the country, and it knew no way of accurately forecasting whether the evil might spread elsewhere in the future. It chose, in acceptable legislative fashion, to limit its attention to the geographic areas where immediate action seemed necessary.34 The alternatives show why. A nationwide remedy would have burdened jurisdictions against which Congress had no evidence of the evil. A remedy withheld until Congress could predict where discrimination would appear next would have left the documented problem untouched, and Congress was entitled to shift the advantage of time and inertia to the victims of that problem.35 Targeting places the burden where the evidence is.
South Carolina answered with the doctrine of the equality of States. The Court held that the doctrine applies only to the terms on which States are admitted to the Union, and not to remedies for local evils that appeared afterward.36 That answer works because of what the doctrine regulates. A rule about admission governs the conditions Congress may attach when a State enters, so it is measured at the moment of entry. The Voting Rights Act attaches no condition to anyone’s entry, and it responds to conduct that occurred after statehood in particular places. If the doctrine reached every differential remedy, Congress could address a problem concentrated in a few jurisdictions only by burdening all States alike.
The earlier discussion of the coverage formula described how it operates; the constitutional question is why it survived review. The evidence came in two grades. Federal courts had repeatedly found substantial voting discrimination in Alabama, Louisiana, and Mississippi, while Georgia, South Carolina, and part of North Carolina were covered on more fragmentary evidence gathered by the Justice Department and the Civil Rights Commission.37 The areas with evidence of actual discrimination shared two traits that Congress built into the formula: the use of tests and devices for voter registration, and a 1964 presidential voting rate at least 12 points below the national average. Each trait carries its own logic. Tests and devices have a long history as a tool of voting discrimination, and a low voting rate is relevant because widespread disenfranchisement must reduce the number of actual voters. For those reasons the Court called the formula “rational in both practice and theory.”38
The Court then drew two further conclusions from the record. No State or subdivision exempted from coverage had a record of recent racial discrimination involving tests and devices, which confirmed the rationality of the line.39 And it was permissible to impose the new remedies on the few remaining covered areas, at least in the absence of proof that they had been free of substantial voting discrimination in recent years. Because the formula works somewhat like a presumption, the Court noted that Congress is not bound by the rules governing statutory presumptions in criminal cases when it prescribes civil remedies against other organs of government under Section 2.40
A formula built from proxies will sweep in some jurisdictions where the danger never materialized, and the Act addressed that risk directly. Acknowledging the possibility of overbreadth, it allows a State or subdivision to terminate coverage if the danger of substantial voting discrimination has not materialized during the preceding five years, and the Court approved limiting that litigation to a single court in the District of Columbia.41 Bailout converts a general worry about overbreadth into a question each covered jurisdiction can settle with its own record. The burden of proof it places on the jurisdiction is bearable, because the facts about its voting practices are peculiarly within its own knowledge.42
Preclearance was the most intrusive remedy and needed the strongest justification, which the Court found in the history of evasion. The Act suspends new voting regulations pending federal scrutiny, and the Court conceded that this may have been an uncommon exercise of congressional power. Its answer was that exceptional conditions can justify legislative measures not otherwise appropriate.43 The exceptional condition was specific. Congress knew that some covered States had resorted to the extraordinary stratagem of contriving new rules of various kinds for the sole purpose of perpetuating voting discrimination in the face of adverse federal court decrees, and it had reason to suppose they might try similar maneuvers to evade the Act’s own remedies.44 Suspending only the rules in force in 1965 would have left the substitution tactic available, and freezing and reviewing new rules closes it. The justification is tied to the record about these jurisdictions, so it reaches only as far as that record does.
That dependence on the record is what Shelby County later enforced. When Congress subjected some States but not others to preclearance on a coverage formula built on decades-old data, the Court held that this violated the Fifteenth Amendment’s enforcement power, because the Tenth Amendment and principles of equal sovereignty require extraordinary justification for such disparate treatment.45 Katzenbach’s answer on equal States concerned the terms of admission. The later Court held that equal-sovereignty principles demand extraordinary justification for disparate coverage and that decades-old data cannot supply it. The sections ahead take up how the Court framed that inquiry.