Holder v. Hall, 512 U.S. 874 (1994)

Argued: October 4, 1993
Decided: June 30, 1994
Syllabus

OCTOBER TERM, 1993

Syllabus

HOLDER, INDIVIDUALLY AND IN HIS OFFICIAL CAPACITY AS COUNTY COMMISSIONER FOR BLECKLEY COUNTY, GEORGIA, ET AL. v. HALL ET AL.

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE ELEVENTH CIRCUIT

No. 91-2012. Argued October 4, 1993-Decided June 30, 1994

Bleckley County, Georgia, has always had a form of government whereby a single commissioner holds all legislative and executive authority. In 1985, the state legislature authorized the county to adopt by referendum a multimember commission consisting of five members elected from single-member districts and a chair elected at large, but voters defeated the proposal, although they had previously approved a five-member district plan for the county school board. Respondents, black voters and the local chapter of the National Association for the Advancement of Colored People, filed this action. The District Court rejected their constitutional claim that the single-member commission was enacted or maintained with an intent to exclude or limit the political influence of the county's black community in violation of the Fourteenth and Fifteenth Amendments. The court also ruled against their claim that the commission's size violated § 2 of the Voting Rights Act of 1965, finding that respondents satisfied only one of the three preconditions established in Thornburg v. Gingles, 478 U. S. 30. The Court of Appeals reversed on the statutory claim, holding that the totality of the circumstances supported § 2 liability and remanding for a formulation of a remedy, which it suggested could be modeled after the county's school board election system.

Held: The judgment is reversed, and the case is remanded. 955 F.2d 1563, reversed and remanded.

JUSTICE KENNEDY, joined by THE CHIEF JUSTICE and JUSTICE O'CONNOR, concluded in Parts I, II-A, and III:

1. The size of a governing authority is not subject to a vote dilution challenge under § 2. Along with determining whether the Gingles preconditions are met and whether the totality of the circumstances support a liability finding, a court in a § 2 suit must find a reasonable alternative practice as a benchmark against which to measure the existing voting practice. However, there is no objective and workable standard for choosing a reasonable benchmark where, as here, the challenge is brought to the government body's size. There is no reason why one


875

size should be picked over another. Respondents have offered no convincing reasons why the benchmark should be a hypothetical fivemember commission. That such a commission is the most common form of governing authority in the State does not bear on dilution, since a sole commissioner system has the same impact on voting strength whether it is shared by none, or by all, of Georgia's counties. That the county was authorized to expand its commission, and that it adopted a five-member school board, are likewise irrelevant considerations. At most, they indicate that the county could change the size of its governing body with minimal disruption, but the failure to do so says nothing about the effects the current system has on the county citizens' voting power. Pp.880-882.

2. The case is remanded for consideration of respondents' constitutional claim. P. 885.

JUSTICE KENNEDY, joined by THE CHIEF JUSTICE, concluded in Part II-B that a voting practice subject to the preclearance requirement of § 5 of the Act is not necessarily subject to a dilution challenge under § 2. The sections differ in structure, purpose, and application; and in contrast to § 2 cases, a baseline for comparison under § 5 exists by definition: A proposed voting practice is measured against the existing practice to determine whether retrogression would result from the proposed change. pp. 882-885.

JUSTICE O'CONNOR concluded that precedent compels the conclusion that the size of a governing authority is both a "standard, practice, or procedure" under § 2 and a "standard, practice, or procedure with respect to voting" under § 5, but agreed that a § 2 dilution challenge to a governing authority's size cannot be maintained because there can never be an objective alternative benchmark for comparison. Pp. 885-888.

JUSTICE THOMAS, joined by JUSTICE SCALIA, concluded that the size of a governing body cannot be attacked under § 2 because it is not a "standard, practice, or procedure" within the terms of § 2. An examination of § 2's text makes it clear that those terms refer only to practices that affect minority citizens' access to the ballot. Districting systems and electoral mechanisms that may affect the "weight" given to a ballot duly cast and counted are simply beyond the purview of the Act. The decision in Thornburg v. Gingles, 478 U. S. 30, which interprets § 2 to reach claims of vote "dilution," should be overruled. Gingles was based upon a flawed method of statutory construction and has produced an interpretation of § 2 that is at odds with the text of the Act and that has proved unworkable in practice. Pp. 891-946.

KENNEDY, J., announced the judgment of the Court and delivered an opinion, in which REHNQUIST, C. J., joined, and in all but Part II-B of


876
Full Text of Opinion


Opinions
OCTOBER TERM, 1993

Syllabus

HOLDER, INDIVIDUALLY AND IN HIS OFFICIAL CAPACITY AS COUNTY COMMISSIONER FOR BLECKLEY COUNTY, GEORGIA, ET AL. v. HALL ET AL.

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE ELEVENTH CIRCUIT

No. 91-2012. Argued October 4, 1993-Decided June 30, 1994

Bleckley County, Georgia, has always had a form of government whereby a single commissioner holds all legislative and executive authority. In 1985, the state legislature authorized the county to adopt by referendum a multimember commission consisting of five members elected from single-member districts and a chair elected at large, but voters defeated the proposal, although they had previously approved a five-member district plan for the county school board. Respondents, black voters and the local chapter of the National Association for the Advancement of Colored People, filed this action. The District Court rejected their constitutional claim that the single-member commission was enacted or maintained with an intent to exclude or limit the political influence of the county's black community in violation of the Fourteenth and Fifteenth Amendments. The court also ruled against their claim that the commission's size violated § 2 of the Voting Rights Act of 1965, finding that respondents satisfied only one of the three preconditions established in Thornburg v. Gingles, 478 U. S. 30. The Court of Appeals reversed on the statutory claim, holding that the totality of the circumstances supported § 2 liability and remanding for a formulation of a remedy, which it suggested could be modeled after the county's school board election system.

Held: The judgment is reversed, and the case is remanded. 955 F.2d 1563, reversed and remanded.

JUSTICE KENNEDY, joined by THE CHIEF JUSTICE and JUSTICE O'CONNOR, concluded in Parts I, II-A, and III:

1. The size of a governing authority is not subject to a vote dilution challenge under § 2. Along with determining whether the Gingles preconditions are met and whether the totality of the circumstances support a liability finding, a court in a § 2 suit must find a reasonable alternative practice as a benchmark against which to measure the existing voting practice. However, there is no objective and workable standard for choosing a reasonable benchmark where, as here, the challenge is brought to the government body's size. There is no reason why one


875

size should be picked over another. Respondents have offered no convincing reasons why the benchmark should be a hypothetical fivemember commission. That such a commission is the most common form of governing authority in the State does not bear on dilution, since a sole commissioner system has the same impact on voting strength whether it is shared by none, or by all, of Georgia's counties. That the county was authorized to expand its commission, and that it adopted a five-member school board, are likewise irrelevant considerations. At most, they indicate that the county could change the size of its governing body with minimal disruption, but the failure to do so says nothing about the effects the current system has on the county citizens' voting power. Pp.880-882.

2. The case is remanded for consideration of respondents' constitutional claim. P. 885.

JUSTICE KENNEDY, joined by THE CHIEF JUSTICE, concluded in Part II-B that a voting practice subject to the preclearance requirement of § 5 of the Act is not necessarily subject to a dilution challenge under § 2. The sections differ in structure, purpose, and application; and in contrast to § 2 cases, a baseline for comparison under § 5 exists by definition: A proposed voting practice is measured against the existing practice to determine whether retrogression would result from the proposed change. pp. 882-885.

JUSTICE O'CONNOR concluded that precedent compels the conclusion that the size of a governing authority is both a "standard, practice, or procedure" under § 2 and a "standard, practice, or procedure with respect to voting" under § 5, but agreed that a § 2 dilution challenge to a governing authority's size cannot be maintained because there can never be an objective alternative benchmark for comparison. Pp. 885-888.

JUSTICE THOMAS, joined by JUSTICE SCALIA, concluded that the size of a governing body cannot be attacked under § 2 because it is not a "standard, practice, or procedure" within the terms of § 2. An examination of § 2's text makes it clear that those terms refer only to practices that affect minority citizens' access to the ballot. Districting systems and electoral mechanisms that may affect the "weight" given to a ballot duly cast and counted are simply beyond the purview of the Act. The decision in Thornburg v. Gingles, 478 U. S. 30, which interprets § 2 to reach claims of vote "dilution," should be overruled. Gingles was based upon a flawed method of statutory construction and has produced an interpretation of § 2 that is at odds with the text of the Act and that has proved unworkable in practice. Pp. 891-946.

KENNEDY, J., announced the judgment of the Court and delivered an opinion, in which REHNQUIST, C. J., joined, and in all but Part II-B of


876

Opinion of KENNEDY, J.

which O'CONNOR, J., joined. O'CONNOR, J., filed an opinion concurring in part and concurring in the judgment, post, p. 885. THOMAS, J., filed an opinion concurring in the judgment, in which SCALIA, J., joined, post, p. 891. BLACKMUN, J., filed a dissenting opinion, in which STEVENS, SouTER, and GINSBURG, JJ., joined, post, p. 946. GINSBURG, J., filed a dissenting opinion, post, p. 956. STEVENS, J., filed a separate opinion, in which BLACKMUN, SOUTER, and GINSBURG, JJ., joined, post, p. 957.

R. Napier Murphy argued the cause for petitioners.

With him on the briefs was W Lonnie Barlow.

Christopher Coates argued the cause for respondents.

With him on the brief were Laughlin McDonald, Kathleen Wilde, Neil Bradley, Mary Wyckoff, John A. Powell, and Steven R. Shapiro. *

JUSTICE KENNEDY announced the judgment of the Court and delivered an opinion, in which THE CHIEF JUSTICE joined, and in all but Part II-B of which JUSTICE O'CONNOR joined.

This case presents the question whether the size of a governing authority is subject to a vote dilution challenge under § 2 of the Voting Rights Act of 1965, 42 U. S. C. § 1973.

I

The State of Georgia has 159 counties, one of which is Bleckley County, a rural county in central Georgia. Black persons make up nearly 20% of the eligible voting population in Bleckley County. Since its creation in 1912, the county has had a single-commissioner form of government for the exercise of "county governing authority." See Ga. Code Ann. § 1-3-3(7) (Supp. 1993). Under this system, the

*Briefs of amici curiae urging affirmance were filed for the United States by Acting Solicitor General Bryson, Acting Assistant Attorney General Turner, Acting Deputy Solicitor General Kneedler, Michael R. Dreeben, and Dennis J. Dimsey; and for the Lawyers' Committee for Civil Rights Under Law by Antonia B. Ianniello, Herbert M. Wachtell, William H. Brown III, Norman Redlich, Thomas J. Henderson, Frank R. Parker, and Brenda Wright.


877

Bleckley County Commissioner performs all of the executive and legislative functions of the county government, including the levying of general and special taxes, the directing and controlling of all county property, and the settling of all claims. Ga. Code Ann. § 36-5-22.1 (1993). In addition to Bleckley County, about 10 other Georgia counties use the single-commissioner system; the rest have multimember commissions.

In 1985, the Georgia Legislature authorized Bleckley County to adopt a multimember commission consisting of five commissioners elected from single-member districts and a single chairman elected at large. 1985 Ga. Laws, p. 4406. In a referendum held in 1986, however, the electorate did not adopt the change to a multimember commission. (In a similar referendum four years earlier, county voters had approved a five-member district plan for the election of the county school board.)

In 1985, respondents (six black registered voters from Bleckley County and the CochranlBleckley County Chapter of the National Association for the Advancement of Colored People) challenged the single-commissioner system in a suit filed against petitioners (Jackie Holder, the incumbent county commissioner, and Probate Judge Robert Johnson, the superintendent of elections). The complaint raised both a constitutional and a statutory claim.

In their constitutional claim, respondents alleged that the county's single-member commission was enacted or maintained with an intent to exclude or to limit the political influence of the county's black community in violation of the Fourteenth and Fifteenth Amendments. At the outset, the District Court made extensive findings of fact about the political history and dynamics of Bleckley County. The court found, for example, that when the county was formed in 1912, few, if any, black citizens could vote. Indeed, until passage of federal civil rights laws, Bleckley County "enforced racial segregation in all aspects of local government-courthouse,


878

Opinion of KENNEDY, J.

jails, public housing, governmental services-and deprived its black citizens of the opportunity to participate in local government." 757 F. Supp. 1560, 1562 (MD Ga. 1991). And even today, though legal segregation no longer exists, "more black than white residents of Bleckley County continue to endure a depressed socio-economic status." Ibid. No black person has run for or been elected to the office of Bleckley County Commissioner, and the District Judge stated that, having run for public office himself, he "wouldn't run if [he] were black in Bleckley [C]ounty." See 955 F.2d 1563, 1571 (CAll 1992).

The court rejected respondents' constitutional contention, however, concluding that respondents "ha[d] failed to provide any evidence that Bleckley County's single member county commission [wa]s the product of original or continued racial animus or discriminatory intent." 757 F. Supp., at 1571. Nor was there evidence that the system was maintained "for tenuous reasons" or that the commissioner himself was unresponsive to the "particularized needs" of the black community. Id., at 1564. There was no "slating process" to stand as a barrier to black candidates, and there was testimony from respondents that they were unaware of any racial appeals in recent elections. Id., at 1562, n. 2, 1583.

In their statutory claim, respondents asserted that the county's single-member commission violated § 2 of the Voting Rights Act of 1965, 79 Stat. 437, as amended, 42 U. S. C. § 1973. Under the statute, the suit contended, Bleckley County must have a county commission of sufficient size that, with single-member election districts, the county's black citizens would constitute a majority in one of the single-member districts. Applying the § 2 framework established in Thornburg v. Gingles, 478 U. S. 30 (1986), the District Court found that respondents satisfied the first of the three Gingles preconditions because black voters were sufficiently numerous and compact that they could have constituted a majority in one district of a multimember commission. In particular,


879

the District Court found that "[i]f the county commission were increased in number to six commissioners to be elected from five single member districts and if the districts were the same as the present school board election districts, a black majority 'safe' district ... would result." 757 F. Supp., at 1565. The court found, however, that respondents failed to satisfy the second and third Gingles preconditions-that whites vote as a bloc in a manner sufficient to defeat the black-preferred candidate and that blacks were politically cohesive.

The Court of Appeals for the Eleventh Circuit reversed on the statutory claim. Relying on its decision in Carrollton Branch of NAACP v. Stallings, 829 F.2d 1547 (1987), the court first held that a challenge to the single-commissioner system was subject to the same analysis as that used in Gingles. Applying that analysis, the Court of Appeals agreed with the District Court that respondents had satisfied the first Gingles precondition by showing that blacks could constitute a majority of the electorate in one of five singlemember districts. The court explained that it was "appropriate to consider the size and geographical compactness of the minority group within a restructured form of the challenged system when the existing structure is being challenged as dilutive." 955 F. 2d, at 1569. The Court of Appeals further found that the District Court had erred in concluding that the second and third Gingles preconditions were not met. Turning to the totality of the circumstances, the court found that those circumstances supported a finding of liability under § 2. The court therefore concluded that respondents had proved a violation of § 2, and it remanded for formulation of a remedy, which, it suggested, "could well be modeled" after the system used to elect the Bleckley County school board. 955 F. 2d, at 1573-1574, and n. 20. Because of its statutory ruling, the Court of Appeals did not consider the District Court's ruling on respondents' constitutional claim.


880

Opinion of KENNEDY, J.

We granted certiorari to review the statutory holding of the Court of Appeals. 507 U. S. 959 (1993).

II A

Section 2 of the Voting Rights Act of 1965 provides that "[n]o voting qualification or prerequisite to voting, or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner which results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color." 42 U. S. C. § 1973(a). In a § 2 vote dilution suit, along with determining whether the Gingles preconditions are met 1 and whether the totality of the circumstances supports a finding of liability, a court must find a reasonable alternative practice as a benchmark against which to measure the existing voting practice. See post, at 887 (O'CONNOR, J., concurring in part and concurring in judgment). As JUSTICE O'CONNOR explained in Gingles: "The phrase vote dilution itself suggests a norm with respect to which the fact of dilution may be ascertained .... [I]n order to decide whether an electoral system has made it harder for minority voters to elect the candidates they prefer, a court must have an idea in mind of how hard it should be for minority voters to elect their preferred candidates under an acceptable system." 478 U. S., at 88 (opinion concurring in judgment) (internal quotation marks omitted).

In certain cases, the benchmark for comparison in a § 2 dilution suit is obvious. The effect of an anti-single-shot voting rule, for instance, can be evaluated by comparing the

1 Gingles requires a showing that "the minority group ... is sufficiently large and geographically compact to constitute a majority in a singlemember district," 478 U. S., at 50, that the minority group is politically cohesive, and that the majority group "votes sufficiently as a bloc to enable it-in the absence of special circumstances ... usually to defeat the minority's preferred candidate," id., at 51.


881

system with that rule to the system without that rule. But where there is no objective and workable standard for choosing a reasonable benchmark by which to evaluate a challenged voting practice, it follows that the voting practice cannot be challenged as dilutive under § 2. See post, at 887-891 (O'CONNOR, J., concurring in part and concurring in judgment).

As the facts of this case well illustrate, the search for a benchmark is quite problematic when a § 2 dilution challenge is brought to the size of a government body. There is no principled reason why one size should be picked over another as the benchmark for comparison. Respondents here argue that we should compare Bleckley County's sole commissioner system to a hypothetical five-member commission in order to determine whether the current system is dilutive. Respondents and the United States as amicus curiae give three reasons why the single-commissioner structure should be compared to a five-member commission (instead of, say, a 3-, 10-, or 15-member body): (1) because the five-member commission is a common form of governing authority in the State; (2) because the state legislature had authorized Bleckley County to adopt a five-member commission if it so chose (it did not); and (3) because the county had moved from a single superintendent of education to a school board with five members elected from single-member districts. See Brief for United States as Amicus Curiae 17-18.

These referents do not bear upon dilution. It does not matter, for instance, how popular the single-member commission system is in Georgia in determining whether it dilutes the vote of a minority racial group in Bleckley County. That the single-member commission is uncommon in the State of Georgia, or that a five-member commission is quite common, tells us nothing about its effects on a minority group's voting strength. The sole commissioner system has the same impact regardless of whether it is shared by none, or by all, of the other counties in Georgia. It makes little


882

Opinion of KENNEDY, J.

sense to say (as do respondents and the United States) that the sole commissioner system should be subject to a dilution challenge if it is rare-but immune if it is common.

That Bleckley County was authorized by the State to expand its commission, and that it adopted a five-member school board, are likewise irrelevant considerations in the dilution inquiry. At most, those facts indicate that Bleckley County could change the size of its commission with minimal disruption. But the county's failure to do so says nothing about the effects the sole commissioner system has on the voting power of Bleckley County's citizens. Surely a minority group's voting strength would be no more or less diluted had the State not authorized the county to alter the size of its commission, or had the county not enlarged its school board. One gets the sense that respondents and the United States have chosen a benchmark for the sake of having a benchmark. But it is one thing to say that a benchmark can be found, quite another to give a convincing reason for finding it in the first place.

B

To bolster their argument, respondents point out that our § 5 cases may be interpreted to indicate that covered jurisdictions may not change the size of their government bodies without obtaining preclearance from the Attorney General or the federal courts. Brief for Respondents 29; see Presley v. Etowah County Comm'n, 502 U. S. 491, 501-503 (1992); City of Lockhart v. United States, 460 U. S. 125, 131-132 (1983); City of Rome v. United States, 446 U. S. 156, 161 (1980). Respondents contend that these § 5 cases, together with the similarity in language between §§ 2 and 5 of the Act, compel the conclusion that the size of a government body must be subject to a dilution challenge under § 2. It is true that in Chisom v. Roemer, 501 U. S. 380, 401-402 (1991), we said that the coverage of §§ 2 and 5 is presumed to be the same (at least if differential coverage would be anomalous). We did not adopt a conclusive rule to that effect, however,


883

and we do not think that the fact that a change in a voting practice must be precleared under § 5 necessarily means that the voting practice is subject to challenge in a dilution suit under §2.

To be sure, if the structure and purpose of § 2 mirrored that of § 5, then the case for interpreting §§ 2 and 5 to have the same application in all cases would be convincing. But the two sections differ in structure, purpose, and application.2 Section 5 applies only in certain jurisdictions specified by Congress and "only to proposed changes in voting procedures." Beer v. United States, 425 U. S. 130, 138 (1976); see 42 U. S. C. § 1973b(b) (specifying jurisdictions where § 5 applies). In those covered jurisdictions, a proposed change in a voting practice must be approved in advance by the Attorney General or the federal courts. § 1973c. The purpose of this requirement "has always been to insure that no voting-procedure changes would be made that would lead to a retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise." 425 U. S., at 141. Under § 5, then, the proposed voting practice is measured against the existing voting practice to determine whether retrogression would result from the proposed change. See ibid. The baseline for comparison is present by definition; it is the existing status. While there may be difficulty in determining whether a pro-

2 Section 2 provides that "[n]o voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner which results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color." 42 U. S. C. § 1973(a).

Section 5 requires preclearance approval by a court or by the Attorney General "[w]henever a [covered] State or political subdivision ... shall enact or seek to administer any voting qualification or prerequisite to voting, or standard, practice, or procedure with respect to voting ... different from that [previously] in force or effect" so as to ensure that it "does not have the purpose and will not have the effect of denying or abridging the right to vote on account of race or color .... " 42 U. S. C. § 1973c.


884

Opinion of KENNEDY, J.

posed change would cause retrogression, there is little difficulty in discerning the two voting practices to compare to determine whether retrogression would occur. See 28 CFR § 51.54(b) (1993).

Retrogression is not the inquiry in § 2 dilution cases. 42 U. S. C. § 1973(a) (whether voting practice "results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color"); S. Rep. No. 97417, p. 68, n. 224 (1982) ("Plaintiffs could not establish a Section 2 violation merely by showing that a challenged reapportionment or annexation, for example, involved a retrogressive effect on the political strength of a minority group"). Unlike in § 5 cases, therefore, a benchmark does not exist by definition in § 2 dilution cases. And as explained above, with some voting practices, there in fact may be no appropriate benchmark to determine if an existing voting practice is dilutive under § 2. For that reason, a voting practice that is subject to the preclearance requirements of § 5 is not necessarily subject to a dilution challenge under § 2.

This conclusion is quite unremarkable. For example, in Perkins v. Matthews, 400 U. S. 379, 388 (1971), we held that a town's annexation ofland was covered under § 5. Notwithstanding that holding, we think it quite improbable to suggest that a § 2 dilution challenge could be brought to a town's existing political boundaries (in an attempt to force it to annex surrounding land) by arguing that the current boundaries dilute a racial group's voting strength in comparison to the proposed new boundaries. Likewise, in McCain v. Lybrand, 465 U. S. 236 (1984), we indicated that a change from an appointive to an elected office was covered under § 5. Here, again, we doubt Congress contemplated that a racial group could bring a § 2 dilution challenge to an appointive office (in an attempt to force a change to an elective office) by arguing that the appointive office diluted its voting strength in comparison to the proposed elective office. We think these examples serve to show that a voting practice is


885

not necessarily subject to a dilution challenge under § 2 even when a change in that voting practice would be subject to the preclearance requirements of § 5.

III

With respect to challenges to the size of a governing authority, respondents fail to explain where the search for reasonable alternative benchmarks should begin and end, and they provide no acceptable principles for deciding future cases. The wide range of possibilities makes the choice "inherently standardless," post, at 889 (O'CONNOR, J., concurring in part and concurring in judgment), and we therefore conclude that a plaintiff cannot maintain a § 2 challenge to the size of a government body, such as the Bleckley County Commission. The judgment of the Court of Appeals is reversed, and the case is remanded for consideration of respondents' constitutional claim.

It is so ordered.

JUSTICE O'CONNOR, concurring in part and concurring in the judgment.

I agree with JUSTICES KENNEDY and THOMAS that a plaintiff cannot maintain a § 2 vote dilution challenge to the size of a governing authority, though I reach that conclusion by a somewhat different rationale. JUSTICE THOMAS rejects the notion that § 2 covers any dilution challenges, and would hold that § 2 is limited to "state enactments that regulate citizens' access to the ballot or the processes for counting a ballot." Post, at 945. As JUSTICE STEVENS points out, however, stare decisis concerns weigh heavily here. Post, at 963-966 (opinion of STEVENS, J.); see also Thornburg v. Gingles, 478 U. S. 30, 84 (1986) (O'CONNOR, J., concurring in judgment) ("We know that Congress intended to allow vote dilution claims to be brought under § 2"); id., at 87 ("I agree with the Court that proof of vote dilution can establish a violation of § 2"). These concerns require me to reject JUSTICE


886

Opinion of O'CONNOR, J.

THOMAS' suggestion that we overhaul our established reading of §2.

I also agree with JUSTICE BLACKMUN, see post, at 946950, that our precedents compel the conclusion that the size of the Bleckley County Commission is both a "standard, practice, or procedure" under § 2 and a "standard, practice, or procedure with respect to voting" under § 5. See, e. g., Presley v. Etowah County Comm'n, 502 U. S. 491, 503 (1992) (change in size is a change in a "standard, practice, or procedure" because the change "increase[s] or diminish[es] the number of officials for whom the electorate may vote"); City of Lockhart v. United States, 460 U. S. 125, 131-132 (1983) (change from three-member commission to five-member commission is subject to § 5 preclearance); City of Rome v. United States, 446 U. S. 156, 160-161 (1980) (it "is not disputed" that an expansion in the size of a board of education is subject to § 5 preclearance); Bunton v. Patterson, decided with Allen v. State Bd. of Elections, 393 U. S. 544, 569-571 (1969) (change from elected to appointed office is subject to § 5 preclearance); id., at 566-567 (§ 2 should be given "the broadest possible scope").

As JUSTICES KENNEDY and BLACKMUN both recognize, in these cases we have consistently said that a change in size is a "standard, practice, or procedure with respect to voting" that is subject to § 5 preclearance. See ante, at 882 (opinion of KENNEDY, J.); post, at 946-948 (BLACKMUN, J., dissenting). And though our cases involving size have concerned § 5, I do not think it possible to read the terms of § 2 more narrowly than the terms of § 5. Section 2 covers any "standard, practice, or procedure," while § 5 covers any "standard, practice, or procedure with respect to voting." As a textual matter, I cannot see how a practice can be a "standard, practice, or procedure with respect to voting," yet not be a "standard, practice, or procedure." Indeed, the similarity in language led to our conclusion in Chisom v. Roemer, 501 U. S. 380,


887

401-402 (1991), that, at least for determining threshold coverage, §§ 2 and 5 have parallel scope.

But determining the threshold scope of coverage does not end the inquiry, at least so far as § 2 dilution challenges are concerned. As JUSTICES KENNEDY and BLACKMUN agree, the fact that the size of a governing authority is a "standard, practice, or procedure" does not answer the question whether respondents may maintain a § 2 vote dilution challenge. See ante, at 880 (opinion of KENNEDY, J.); post, at 951 (BLACKMUN, J., dissenting). Section 2 vote dilution plaintiffs must establish that the challenged practice is dilutive. In order for an electoral system to dilute a minority group's voting power, there must be an alternative system that would provide greater electoral opportunity to minority voters. "Put simply, in order to decide whether an electoral system has made it harder for minority voters to elect the candidates they prefer, a court must have an idea in mind of how hard it 'should' be for minority voters to elect their preferred candidates under an acceptable system." Gingles, 478 U. S., at 88 (O'CONNOR, J., concurring in judgment). As we have said, "[u]nless minority voters possess the potential to elect representatives in the absence of the challenged structure or practice, they cannot claim to have been injured by that structure or practice." Id., at 50, n. 17 (emphasis in original); see also id., at 99 (O'CONNOR, J., concurring in judgment) ("[T]he relative lack of minority electoral success under a challenged plan, when compared with the success that would be predicted under the measure of undiluted minority voting strength the court is employing, can constitute powerful evidence of vote dilution") (emphasis added).

Accordingly, to determine whether voters possess the potential to elect representatives of choice in the absence of the challenged structure, courts must choose an objectively reasonable alternative practice as a benchmark for the dilution comparison. On this, there is general agreement. See ante, at 880 (opinion of KENNEDY, J.) ("[A] court must find a


888

Opinion of O'CONNOR, J.

reasonable alternative practice as a benchmark against which to measure the existing voting practice"); post, at 951 (BLACKMUN, J., dissenting) ("[T]he allegedly dilutive mechanism must be measured against the benchmark of an alternative structure or practice that is reasonable and workable under the facts of the specific case"). We require preclearance of changes in size under § 5, because in a § 5 case the question of an alternative benchmark never arises-the benchmark is simply the former practice employed by the jurisdiction seeking approval of a change. See ante, at 883 (opinion of KENNEDY, J.).

But § 2 dilution challenges raise more difficult questions.

This case presents the question whether, in a § 2 dilution challenge to size, there can ever be an objective alternative benchmark for comparison. And I agree with JUSTICE KENNEDY that there cannot be. As JUSTICE KENNEDY points out, ante, at 880, the alternative benchmark is often self-evident. In a challenge to a multimember at-large system, for example, a court may compare it to a system of multiple single-member districts. See Gingles, supra, at 38, 50; Davidson, Minority Vote Dilution: An Overview, in Minority Vote Dilution 5 (C. Davidson ed. 1984). Similarly, a court may assess the dilutive effect of majority vote requirements, numbered posts, staggered terms, residency requirements, or anti-single-shot rules by comparing the election results under a system with the challenged practice to the results under a system without the challenged practice. Cf. City of Rome, supra, at 183-185; U. S. Comm'n on Civil Rights, The Voting Rights Act: Ten Years After, pp. 206-208 (1975); Note, Application of Section 2 of the Voting Rights Act to Runoff Primary Election Laws, 91 Colum. L. Rev. 1127, 1148 (1991). Though there may be disagreements about the precise appropriate alternative practice in these cases, see Gingles, supra, at 88-89 (O'CONNOR, J., concurring in judgment), there are at least some objectively determinable constraints on the dilution inquiry.


889

This is not so with § 2 dilution challenges to size, however.

In a dilution challenge to the size of a governing authority, choosing the alternative for comparison-a hypothetical larger (or smaller) governing authority-is extremely problematic. See ante, at 881-882 (opinion of KENNEDY, J.). The wide range of possibilities makes the choice inherently standardless. Here, for example, respondents argued that the single-member commission structure was dilutive in comparison to a five-member structure, in which AfricanAmericans would probably have been able to elect one representative of their choice. Some groups, however, will not be able to constitute a majority in one of five districts. Once a court accepts respondents' reasoning, it will have to allow a plaintiff group insufficiently large or geographically compact to form a majority in one of five districts to argue that the jurisdiction's failure to establish a 10-, 15-, or 25commissioner structure is dilutive. See, e. g., Romero v. Pomona, 883 F.2d 1418, 1425, n. 10 (CA9 1989); Heath, Managing the Political Thicket: Developing Objective Standards in Voting Rights Litigation, 21 Stetson L. Rev. 819, 827 (1992) ("[O]nce one departs from the current number of districts or other objective standard, the test loses its validity as a threshold standard").

Respondents argue that this concern with arbitrary and standardless intrusions into the size of local governing authority is overstated. Respondents' principal support for this conclusion is that a five-member commission is the most common size for Georgia. But a five-member commission is not the only common size in Georgia: 22 Georgia counties have three-member commissions (and one county has an 11member commission). Moreover, there is no good reason why the search for benchmarks should be limited to Georgia. Expanding the search nationwide produces many 20-person county commissions in Tennessee, and 40-member commissions in Wisconsin. DeSantis, County Government: A Century of Change, in The Municipal Yearbook 1989, pp. 80, 83.


890

Opinion of O'CONNOR, J.

In sum, respondents do not explain how common an alternative practice must be before it can be a reliable alternative benchmark for the dilution comparison, nor do they explain where the search for alternative benchmarks should begin and end.

Respondents' failure to provide any meaningful principles for deciding future cases demonstrates the difficulty with allowing dilution challenges to the size of a governing authority. Under respondents' open-ended test, a wide range of state governmental bodies may be subject to a dilution challenge. Within each State there are many forms of government, including county commissions that range dramatically in size. For example, the majority of county commissions in New Jersey have seven members, but three counties have smaller commissions and one has a larger commission. Id., at 76. Similarly, in South Carolina the norm is a sevenmember commission, but a number of counties deviate. Id., at 79. In Tennessee, the average size for a county commission is 19 members, but one county has as few as 9 and another has as many as 40. Id., at 80. And in Wisconsin the average size is 27 members, but the commission sizes range from 7 to 46. Id., at 83.

Nor are deviations from the norm limited to counties.

Statewide governing authorities also range dramatically in size, and often do not correlate to the size of the State. For example, Texas has only 31 members in its State Senate, while tiny Rhode Island has 50. Council of State Governments, State Elective Officials and the Legislatures 1993-94, p. vi. The Texas Senate is smaller than the national average and the Rhode Island Senate is larger. Similarly, California has an unusually small 80-person Assembly, while New Hampshire has a 400-person House. Ibid.

The discrepancies in size among state and local governing authorities reinforce my concern that the limiting principle offered by respondents will in practice limit very little. Though respondents purport to present Bleckley County as


891

unique, it is not. County commissions throughout New Jersey, South Carolina, Tennessee, and Wisconsin, and the state legislatures of Texas, Rhode Island, California, and New Hampshire are ripe for a dilution challenge under respondents' theory, since they do not fit the norm for their State. Moreover, though my examples are some of the more extreme ones, they are not alone. In these cases, and perhaps in many more, the potential reach of allowing dilution challenges to size will not be meaningfully circumscribed by the open-ended requirement that the alternative benchmark be "reasonable and workable." Post, at 951 (BLACKMUN, J., dissenting).

For these reasons, I concur in the conclusion that respondents' dilution challenge to the size of the Bleckley County Commission cannot be maintained under § 2 of the Voting Rights Act, and I join Parts I, II-A, and III of JUSTICE KENNEDY'S opinion. Because the Court appropriately reverses the judgment below and remands for consideration of respondents' constitutional claim of intentional discrimination, I also concur in the judgment.

JUSTICE THOMAS, with whom JUSTICE SCALIA joins, concurring in the judgment.

We are asked in this case to determine whether the size of a local governing body is subject to challenge under § 2 of the Voting Rights Act of 1965 as a "dilutive" practice. While I agree with JUSTICES KENNEDY and O'CONNOR that the size of a governing body cannot be attacked under § 2, I do not share their reasons for reaching that conclusion. JUSTICE KENNEDY persuasively demonstrates that there is no principled method for determining a benchmark against which the size of a governing body might be compared to determine whether it dilutes a group's voting power. Both he and JUSTICE O'CONNOR rely on that consideration to conclude that size cannot be challenged under § 2 of the Act. See ante, at 880-882, 885 (opinion of KENNEDY, J.);


892

THOMAS, J., concurring in judgment

ante, at 888-891 (O'CONNOR, J., concurring in part and concurring in judgment).

While the practical concerns JUSTICES KENNEDY and O'CONNOR point out can inform a proper construction of the Act, I would explicitly anchor analysis in this case in the statutory text. Only a "voting qualification or prerequisite to voting, or standard, practice, or procedure" can be challenged under § 2. I would hold that the size of a governing body is not a "standard, practice, or procedure" within the terms of the Act. In my view, however, the only principle limiting the scope of the terms "standard, practice, or procedure" that can be derived from the text of the Act would exclude, not only the challenge to size advanced today, but also challenges to allegedly dilutive election methods that we have considered within the scope of the Act in the past.

I believe that a systematic reassessment of our interpretation of § 2 is required in this case. The broad reach we have given the section might suggest that the size of a governing body, like an election method that has the potential for diluting the vote of a minority group, should come within the terms of the Act. But the gloss we have placed on the words "standard, practice, or procedure" in cases alleging dilution is at odds with the terms of the statute and has proved utterly unworkable in practice. A review of the current state of our cases shows that by construing the Act to cover potentially dilutive electoral mechanisms, we have immersed the federal courts in a hopeless project of weighing questions of political theory-questions judges must confront to establish a benchmark concept of an "undiluted" vote. Worse, in pursuing the ideal measure of voting strength, we have devised a remedial mechanism that encourages federal courts to segregate voters into racially designated districts to ensure minority electoral success. In doing so, we have collaborated in what may aptly be termed the racial "balkaniz[ation]" of the Nation. Shaw v. Reno, 509 U. S. 630, 658 (1993).


893

I can no longer adhere to a reading of the Act that does not comport with the terms of the statute and that has produced such a disastrous misadventure in judicial policymaking. I would hold that the size of a government body is not a "standard, practice, or procedure" because, properly understood, those terms reach only state enactments that limit citizens' access to the ballot.

I

If one surveys the history of the Voting Rights Act, 42 U. S. C. § 1973 et seq., one can only be struck by the sea change that has occurred in the application and enforcement of the Act since it was passed in 1965. The statute was originally perceived as a remedial provision directed specifically at eradicating discriminatory practices that restricted blacks' ability to register and vote in the segregated South. Now, the Act has grown into something entirely different. In construing the Act to cover claims of vote dilution, we have converted the Act into a device for regulating, rationing, and apportioning political power among racial and ethnic groups. In the process, we have read the Act essentially as a grant of authority to the federal judiciary to develop theories on basic principles of representative government, for it is only a resort to political theory that can enable a court to determine which electoral systems provide the "fairest" levels of representation or the most "effective" or "undiluted" votes to minorities.

Before I turn to an analysis of the text of § 2 to explain why, in my view, the terms of the statute do not authorize the project that we have undertaken in the name of the Act, I intend first simply to describe the development of the basic contours of vote dilution actions under the Voting Rights Act.1 An examination of the current state of our decisions

1 Of course, many of the basic principles I will discuss are equally applicable to constitutional vote dilution cases. Indeed, prior to the amendment of the Voting Rights Act in 1982, dilution claims typically were


894

THOMAS, J., concurring in judgment

should make obvious a simple fact that for far too long has gone unmentioned: Vote dilution cases have required the federal courts to make decisions based on highly political judgments-judgments that courts are inherently ill-equipped to make. A clear understanding of the destructive assumptions that have developed to guide vote dilution decisions and the role we have given the federal courts in redrawing the political landscape of the Nation should make clear the pressing need for us to reassess our interpretation of the Act.

A

As it was enforced in the years immediately following its enactment, the Voting Rights Act of 1965, Pub. L. 89-110,79 Stat. 437, was perceived primarily as legislation directed at eliminating literacy tests and similar devices that had been used to prevent black voter registration in the segregated South. See A. Thernstrom, Whose Votes Count? Affirmative Action and Minority Voting Rights 17-27 (1987) (hereinafter Thernstrom). See also Guinier, The Representation of Minority Interests: The Question of Single-Member Districts, 14 Cardozo L. Rev. 1135, 1151 (1993) (referring to actions securing access to the ballot as the "first generation" of Voting Rights Act claims).2 This focus in enforcement flowed, no doubt, from the emphasis on access to the ballot apparent in the central provision of the Act, § 4, which used a mathematical formula based on voter registration and

brought under the Equal Protection Clause. See, e. g., White v. Regester, 412 U. S. 755 (1973); Whitcomb v. Chavis, 403 U. S. 124 (1971); Burns v. Richardson, 384 U. S. 73 (1966). The early development of our voting rights jurisprudence in those cases provided the basis for our analysis of vote dilution under the amended §2 in Thornburg v. Gingles, 478 U. S. 30 (1986).

2 Cf. L. Guinier, The Tyranny of the Majority 49, n. 58 (1994) (hereinafter Guinier) ("The first generation of voting litigation, and the 1965 statute which represented the congressional response, were concerned with the complete and total exclusion of blacks from the electoral process").


895

turnout in 1964 to define certain "covered" jurisdictions in which the use of literacy tests was immediately suspended. Pub. L. 89-110, §4, 79 Stat. 438. Section 6 of the Act reflected the same concern for registration as it provided that federal examiners could be dispatched to covered jurisdictions whenever the Attorney General deemed it necessary to supervise the registration of black voters. 42 U. S. C. § 1973d. And to prevent evasion of the requirements of § 4, § 5 required that covered jurisdictions obtain "preclearance" from the Department of Justice before altering any "voting qualification or prerequisite to voting, or standard, practice, or procedure with respect to voting." § 1973c.

The Act was immediately and notably successful in removing barriers to registration and ensuring access to the ballot. For example, in Mississippi, black registration levels skyrocketed from 6.7% to 59.8% in a mere two years; in Alabama the increase was from 19.3% to 51.6% in the same time period. See Thernstrom 18. By the end of 1967, black voter registration had reached at least 50% in every covered State. See B. Grofman, L. Handley, & R. Niemi, Minority Representation and the Quest for Voting Equality 22 (1992).

The Court's decision in Allen v. State Bd. of Elections, 393 U. S. 544 (1969), however, marked a fundamental shift in the focal point of the Act. In an opinion dealing with four companion cases, the Allen Court determined that the Act should be given "the broadest possible scope." Id., at 567. Thus, in Fairley v. Patterson, the Court decided that a covered jurisdiction's switch from a districting system to an at-large system for election of county supervisors was a "standard, practice, or procedure with respect to voting," subject to preclearance under § 5. Id., at 569. Stating that the Act "was aimed at the subtle, as well as the obvious, state regulations which have the effect of denying citizens their right to vote because of their race," id., at 565, the Court reasoned that § 5's preclearance provisions should apply, not only to changes in electoral laws that pertain to


896

THOMAS, J., concurring in judgment

registration and access to the ballot, but to provisions that might "dilute" the force of minority votes that were duly cast and counted. See id., at 569. The decision in Allen thus ensured that the terms "standard, practice, or procedure" would extend to encompass a wide array of electoral practices or voting systems that might be challenged for reducing the potential impact of minority votes.

As a consequence, Allen also ensured that courts would be required to confront a number of complex and essentially political questions in assessing claims of vote dilution under the Voting Rights Act. The central difficulty in any vote dilution case, of course, is determining a point of comparison against which dilution can be measured. As Justice Frankfurter observed several years before Allen, "[t]alk of 'debasement' or 'dilution' is circular talk. One cannot speak of 'debasement' or 'dilution' of the value of a vote until there is first defined a standard of reference as to what a vote should be worth." Baker v. Carr, 369 U. S. 186, 300 (1962) (dissenting opinion). See also Thornburg v. Gingles, 478 U. S. 30, 88 (1986) (O'CONNOR, J., concurring in judgment) ("[I]n order to decide whether an electoral system has made it harder for minority voters to elect the candidates they prefer, a court must have an idea in mind of how hard it 'should' be for minority voters to elect their preferred candidates under an acceptable system"). But in setting the benchmark of what "undiluted" or fully "effective" voting strength should be, a court must necessarily make some judgments based purely on an assessment of principles of political theory. As Justice Harlan pointed out in his dissent in Allen, the Voting Rights Act supplies no rule for a court to rely upon in deciding, for example, whether a multimember at-large system of election is to be preferred to a single-member district system; that is, whether one provides a more "effective" vote than another. "Under one system, Negroes have some influence in the election of all officers; under the other, minority groups have more influence in the selection of fewer officers." Allen,


897

supra, at 586 (opinion concurring in part and dissenting in part). The choice is inherently a political one, and depends upon the selection of a theory for defining the fully "effective" vote-at bottom, a theory for defining effective participation in representative government. In short, what a court is actually asked to do in a vote dilution case is "to choose among competing bases of representation-ultimately, really, among competing theories of political philosophy." Baker, supra, at 300 (Frankfurter, J., dissenting).

Perhaps the most prominent feature of the philosophy that has emerged in vote dilution decisions since Allen has been the Court's preference for single-member districting schemes, both as a benchmark for measuring undiluted minority voting strength and as a remedial mechanism for guaranteeing minorities undiluted voting power. See, e. g., Growe v. Emison, 507 U. S. 25, 40 (1993); Gingles, supra, at 50, n. 17 (declaring that the "single-member district is generally the appropriate standard against which to measure minority group potential to elect"); Mobile v. Bolden, 446 U. S. 55, 66, n. 12 (1980) (plurality opinion) (noting that singlemember districts should be preferred in court-ordered remedial schemes); Connor v. Finch, 431 U. S. 407, 415 (1977) (same). Indeed, commentators surveying the history of voting rights litigation have concluded that it has been the objective of voting rights plaintiffs to use the Act to attack multimember districting schemes and to replace them with single-member districting systems drawn with majorityminority districts to ensure minority control of seats. See Guinier, 14 Cardozo L. Rev., at 1151; Guinier 49-54; Thernstrom 193.

It should be apparent, however, that there is no principle inherent in our constitutional system, or even in the history of the Nation's electoral practices, that makes single-member districts the "proper" mechanism for electing representatives to governmental bodies or for giving "undiluted" effect to the votes of a numerical minority. On the contrary, from


898

THOMAS, J., concurring in judgment

the earliest days of the Republic, multimember districts were a common feature of our political systems. The Framers left unanswered in the Constitution the question whether congressional delegations from the several States should be elected on a general ticket from each State as a whole or under a districting scheme and left that matter to be resolved by the States or by Congress. See U. S. Const., Art. I, § 4, cl. 1. It was not until 1842 that Congress determined that Representatives should be elected from single-member districts in the States. See Act of June 25, 1842, ch. 47, 5 Stat. 491.3 Single-member districting was no more the rule in the States themselves, for the Constitutions of most of the 13 original States provided that representatives in the state legislatures were to be elected from multimember districts.4 Today, although they have come under increasing attack under the Voting Rights Act, multimember district systems continue to be a feature on the American political landscape, especially in municipal governments. See The Municipal Yearbook 14 (table) (1988) (over 60% of American cities use at-large election systems for their governing bodies).

The obvious advantage the Court has perceived in singlemember districts, of course, is their tendency to enhance the ability of any numerical minority in the electorate to gain control of seats in a representative body. See Gingles, supra, at 50-51. But in choosing single-member districting as a benchmark electoral plan on that basis the Court has made a political decision and, indeed, a decision that itself depends on a prior political choice made in answer to Justice Harlan's question in Allen. Justice Harlan asked whether a

3 At that time, seven States elected their congressional delegations on a statewide ticket. See Wesberry v. Sanders, 376 U. S. 1, 8, n. 11 (1964).

4 See, e. g., Ga. Const., Art. IV (1777); Mass. Const., Part II, ch. I, § II, Arts. I, II (1780); N. H. Const., Part II (1784); N. J. Const., Art. III (1776); N. Y. Const., Art. IV (1777); S. C. Const., Art. XIII (1778). See also Klain, A New Look at the Constituencies: The Need for a Recount and a Reappraisal, 49 Am. Pol. Sci. Rev. 1105, 1112-1113 (1955).


899

group's votes should be considered to be more "effective" when they provide influence over a greater number of seats, or control over a lesser number of seats. See 393 U. S., at 586. In answering that query, the Court has determined that the purpose of the vote-or of the fully "effective" vote-is controlling seats. In other words, in an effort to develop standards for assessing claims of dilution, the Court has adopted the view that members of any numerically significant minority are denied a fully effective use of the franchise unless they are able to control seats in an elected body.5 Under this theory, votes that do not control a representative are essentially wasted; those who cast them go unrepresented and are just as surely disenfranchised as if they had been barred from registering. Cf. id., at 569 (equating denial of the ability to elect candidates with denial of the vote). Such conclusions, of course, depend upon a certain theory of the "effective" vote, a theory that is not inherent in the concept of representative democracy itself.6

5 See, e. g., Gingles, 478 U. S., at 88 (O'CONNOR, J., concurring in judgment) (noting that the Court has determined that "minority voting strength is to be assessed solely in terms of the minority group's ability to elect candidates it prefers") (emphasis deleted). See also Abrams, "Raising Politics Up": Minority Political Participation and Section 2 of the Voting Rights Act, 63 N. Y. U. L. Rev. 449, 456, n. 43, 468-471 (1988) (criticizing the Court's "electoral focus" as a narrow conception of "political opportunity"); Guinier 49 (arguing that since Gingles, courts "have measured black political representation and participation solely by reference to the number and consistent election of black candidates").

6 Undoubtedly, one factor that has prompted our focus on control of seats has been a desire, when confronted with an abstract question of political theory concerning the measure of effective participation in government, to seize upon an objective standard for deciding cases, however much it may oversimplify the issues before us. If using control of seats as our standard does not reflect a very nuanced theory of political participation, it at least has the superficial advantage of appealing to the "most easily measured indicia of political power." Davis v. Bandemer, 478 U. S. 109, 157 (1986) (O'CONNOR, J., concurring in judgment).


900

THOMAS, J., concurring in judgment

In fact, it should be clear that the assumptions that have guided the Court reflect only one possible understanding of effective exercise of the franchise, an understanding based on the view that voters are "represented" only when they choose a delegate who will mirror their views in the legislative halls. See generally H. Pitkin, The Concept of Representation 60-91 (1967).7 But it is certainly possible to construct a theory of effective political participation that would accord greater importance to voters' ability to influence, rather than control, elections. And especially in a two-party system such as ours, the influence of a potential "swing" group of voters composing 10% to 20% of the electorate in a given district can be considerable.8 Even such a focus on practical influence, however, is not a necessary component of the definition of the "effective" vote. Some conceptions of representative government may primarily emphasize the formal value of the vote as a mechanism for participation in

7 Indeed, the assumptions underpinning the Court's conclusions largely parallel principles that John Stuart Mill advanced in proposing a system of proportional representation as an electoral reform in Great Britain. See J. S. Mill, Considerations on Representative Government (1861). In Mill's view, a just system of representative government required an electoral system that ensured "a minority of the electors would always have a minority of the representatives." Id., at 133. To Mill, a system that allowed a portion of the population that constituted a majority in each district to control the election of all representatives and to defeat the minority's choice of candidates was unjust because it operated to produce a "complete disfranchisement of minorities." Id., at 132.

8 We ourselves have tacitly acknowledged that our current view of what constitutes an effective vote may be subject to reevaluation, or at least that it may not provide an exclusive definition of effective voting power, as we repeatedly have reserved the question whether a vote dilution claim may be brought for failure to create minority "influence" districts. See, e. g., Voinovich v. Quilter, 507 U. S. 146, 154 (1993) (citing cases). Cf. Bandemer, supra, at 132 (noting that "the power to influence the political process is not limited to winning elections"); Gingles, supra, at 99 (O'CONNOR, J., concurring in judgment) (suggesting that the Court should not focus solely on a minority group's ability to elect representatives in assessing the effectiveness of the group's votes).


901

the electoral process, whether it results in control of a seat or not. Cf. id., at 14-59.9 Under such a theory, minorities unable to control elected posts would not be considered essentially without a vote; rather, a vote duly cast and counted would be deemed just as "effective" as any other. If a minority group is unable to control seats, that result may plausibly be attributed to the inescapable fact that, in a majoritarian system, numerical minorities lose elections.10

In short, there are undoubtedly an infinite number of theories of effective suffrage, representation, and the proper apportionment of political power in a representative democracy that could be drawn upon to answer the questions posed in Allen. See generally Pitkin, supra. I do not pretend to have provided the most sophisticated account of the various possibilities; but such matters of political theory are beyond the ordinary sphere of federal judges. And that is precisely the point. The matters the Court has set out to resolve in vote dilution cases are questions of political philosophy, not questions of law.ll As such, they are not readily subjected

9 Cf. also Levinson, Gerrymandering and the Brooding Omnipresence of Proportional Representation, 33 UCLA L. Rev. 257, 260-261 (1985).

10 There are traces of this view in our cases as well. See Whitcomb, 403 U. S., at 153, 155; id., at 160 ("The short of it is that we are unprepared to hold that district-based elections decided by plurality vote are unconstitutional in either single- or multi-member districts simply because the supporters of losing candidates have no legislative seats assigned to them"). See also League of United Latin American Citizens v. Midland Independent School Dist., 812 F.2d 1494, 1507 (CA5) (Higginbotham, J., dissenting) ("I had supposed that the essence of our republican arrangement is that voting minorities lose"), vacated on rehearing, 829 F.2d 546 (1987) (en bane) (per curiam).

11 The point is perhaps so widely accepted at this date that it needs little further demonstration. See, e. g., L. Tribe, American Constitutional Law § 13-7, p. 1076, n. 7 (2d ed. 1988) (stating that "no strategy [in vote dilution cases] can avoid the necessity for at least some hard substantive decisions of political theory by the federal judiciary"); Howard & Howard, The Dilemma of the Voting Rights Act-Recognizing the Emerging Political Equality Norm, 83 Colum. L. Rev. 1615, 1633, 1635 (1983) (hereinafter


902

THOMAS, J., concurring in judgment

to any judicially manageable standards that can guide courts in attempting to select between competing theories.

But the political choices the Court has had to make do not end with the determination that the primary purpose of the "effective" vote is controlling seats or with the selection of single-member districting as the mechanism for providing that control. In one sense, these were not even the most critical decisions to be made in devising standards for assessing claims of dilution, for, in itself, the selection of singlemember districting as a benchmark election plan will tell a judge little about the number of minority districts to create. Single-member districting tells a court "how" members of a minority are to control seats, but not "how many" seats they should be allowed to control.

But "how many" is the critical issue. Once one accepts the proposition that the effectiveness of votes is measured in terms of the control of seats, the core of any vote dilution claim is an assertion that the group in question is unable to control the "proper" number of seats-that is, the number of seats that the minority's percentage of the population would enable it to control in the benchmark "fair" system. The claim is inherently based on ratios between the numbers of the minority in the population and the numbers of seats controlled. As JUSTICE O'CONNOR has noted, "any theory of vote dilution must necessarily rely to some extent on a measure of minority voting strength that makes some reference to the proportion between the minority group and the electorate at large." Gingles, 478 U. S., at 84 (opinion concurring in judgment). As a result, only a mathematical calculation can answer the fundamental question posed by a claim of vote dilution. And once again, in selecting the proportion that will be used to define the undiluted strength of a minor-

Howard & Howard) (arguing that the Court has developed a "substantive theory of representative government" and a theory of "allocating political power" in vote dilution cases).


903

ity-the ratio that will provide the principle for decision in a vote dilution case-a court must make a political choice.

The ratio for which this Court has opted, and thus the mathematical principle driving the results in our cases, is undoubtedly direct proportionality. Indeed, four Members of the Court candidly recognized in Gingles that the Court had adopted a rule of roughly proportional representation, at least to the extent proportionality was possible given the geographic dispersion of minority populations. See id., at 85, 91, 98-99 (O'CONNOR, J., concurring in judgment). While in itself that choice may strike us intuitively as the fairest or most just rule to apply, opting for proportionality is still a political choice, not a result required by any principle of law.

B

The dabbling in political theory that dilution cases have prompted, however, is hardly the worst aspect of our vote dilution jurisprudence. Far more pernicious has been the Court's willingness to accept the one underlying premise that must inform every minority vote dilution claim: the assumption that the group asserting dilution is not merely a racial or ethnic group, but a group having distinct political interests as well. Of necessity, in resolving vote dilution actions we have given credence to the view that race defines political interest. We have acted on the implicit assumption that members of racial and ethnic groups must all think alike on important matters of public policy and must have their own "minority preferred" representatives holding seats in elected bodies if they are to be considered represented at all.

It is true that in Gingles we stated that whether a racial group is "politically cohesive" may not be assumed, but rather must be proved in each case. See 478 U. S., at 51, 56. See also Growe, 507 U. S., at 40-41. But the standards we have employed for determining political cohesion have proved so insubstantial that this "precondition" does not present much of a barrier to the assertion of vote dilution


904

THOMAS, J., concurring in judgment

claims on behalf of any racial groUp.12 Moreover, it provides no test-indeed, it is not designed to provide a test-of whether race itself determines a distinctive political community of interest. According to the rule adopted in Gingles, plaintiffs must show simply that members of a racial group tend to prefer the same candidates. See 478 U. S., at 61-67 (opinion of Brennan, J.). There is no set standard defining how strong the correlation must be, and an inquiry into the cause for the correlation (to determine, for example, whether it might be the product of similar socioeconomic interests rather than some other factor related to race) is unnecessary. Ibid. See also id., at 100 (O'CONNOR, J., concurring in judgment).13 Thus, whenever similarities in political preferences along racial lines exist, we proclaim that the cause of the correlation is irrelevant, but we effectively rely on the fact of the correlation to assume that racial groups have unique political interests.

12 Cf. Citizens for a Better Gretna v. Gretna, 834 F.2d 496, 501-502 (CA5 1987) (emphasizing that political cohesion under Gingles can be shown where a "significant number" of minority voters prefer the same candidate, and suggesting that data showing that anywhere from 49% to 67% of the members of a minority group preferred the same candidate established cohesion), cert. denied, 492 U. S. 905 (1989).

13JUSTICE O'CONNOR agreed with Justice Brennan in Gingles that, insofar as determining political cohesion was concerned, the cause for a correlation between race and candidate preference was irrelevant. She maintained, however, that evidence of the cause of the correlation would still be relevant to the overall vote dilution inquiry and particularly to the question whether a white majority will usually vote to defeat the minority's preferred candidate. See 478 U. S., at 100 (opinion concurring in judgment). The splintering of opinions in Gingles on this point has produced, at best, "uncertainty," Overton v. Austin, 871 F.2d 529, 538 (CA5 1989), and has allowed bivariate regression analysis-that is, an analysis that measures merely the correlation between race and candidate preference and that does not directly control for other factors-to become the norm for determining cohesion in vote dilution cases. See id., at 539. But cf. League of United Latin American Citizens v. Clements, 999 F.2d 831, 850-851 (CA5 1993), cert. denied, 510 U. S. 1071 (1994).


905

As a result, Gingles' requirement of proof of political cohesiveness, as practically applied, has proved little different from a working assumption that racial groups can be conceived of largely as political interest groups. And operating under that assumption, we have assigned federal courts the task of ensuring that minorities are assured their "just" share of seats in elected bodies throughout the Nation.

To achieve that result through the currently fashionable mechanism of drawing majority-minority single-member districts, we have embarked upon what has been aptly characterized as a process of "creating racially 'safe boroughs.'" United States v. Dallas County Comm'n, 850 F.2d 1433, 1444 (CAll 1988) (Hill, J., concurring specially), cert. denied, 490 U. S. 1030 (1989). We have involved the federal courts, and indeed the Nation, in the enterprise of systematically dividing the country into electoral districts along racial lines-an enterprise of segregating the races into political homelands that amounts, in truth, to nothing short of a system of "political apartheid." Shaw, 509 U. S., at 647. See also id., at 657 (noting that racial gerrymandering "may balkanize us into competing racial factions"). Blacks are drawn into "black districts" and given "black representatives"; Hispanics are drawn into Hispanic districts and given "Hispanic representatives"; and so on. Worse still, it is not only the courts that have taken up this project. In response to judicial decisions and the promptings of the Justice Department, the States themselves, in an attempt to avoid costly and disruptive Voting Rights Act litigation, have begun to gerrymander electoral districts according to race. That practice now promises to embroil the courts in a lengthy process of attempting to undo, or at least to minimize, the damage wrought by the system we created. See, e. g., Shaw, supra; Hays v. Louisiana, 839 F. Supp. 1188 (WD La. 1993), appeal pending, No. 93-1539.

The assumptions upon which our vote dilution decisions have been based should be repugnant to any nation that


906

THOMAS, J., concurring in judgment

strives for the ideal of a color-blind Constitution. "The principle of equality is at war with the notion that District A must be represented by a Negro, as it is with the notion that District B must be represented by a Caucasian, District C by a Jew, District D by a Catholic, and so on." Wright v. Rockefeller, 376 U. S. 52, 66 (1964) (Douglas, J., dissenting). Despite Justice Douglas' warning sounded 30 years ago, our voting rights decisions are rapidly progressing toward a system that is indistinguishable in principle from a scheme under which members of different racial groups are divided into separate electoral registers and allocated a proportion of political power on the basis of race. Cf. id., at 63-66. Under our jurisprudence, rather than requiring registration on racial rolls and dividing power purely on a population basis, we have simply resorted to the somewhat less precise expedient of drawing geographic district lines to capture minority populations and to ensure the existence of the "appropriate" number of "safe minority seats."

That distinction in the practical implementation of the concept, of course, is immaterial.14 The basic premises underlying our system of safe minority districts and those behind the racial register are the same: that members of the racial group must think alike and that their interests are so distinct that the group must be provided a separate body of representatives in the legislature to voice its unique point of view. Such a "system, by whatever name it is called, is a divisive force in a community, emphasizing differences between candidates and voters that are irrelevant." Id., at 66. Justice Douglas correctly predicted the results of state sponsorship of such a theory of representation: "When racial or religious

14 Cf. Lijphart, Proportionality by Non-PR Methods: Ethnic Representation in Belgium, Cyprus, Lebanon, New Zealand, West Germany, and Zimbabwe, in Electoral Laws and Their Political Consequences 113, 116 (B. Grofman & A. Lijphart eds. 1986) (describing methods other than separate electoral registers to allocate political power on the basis of ethnicity or race).


907

lines are drawn by the State, ... antagonisms that relate to race or to religion rather than to political issues are generated; communities seek not the best representative but the best racial or religious partisan." Id., at 67. In short, few devices could be better designed to exacerbate racial tensions than the consciously segregated districting system currently being constructed in the name of the Voting Rights Act.

As a practical political matter, our drive to segregate political districts by race can only serve to deepen racial divisions by destroying any need for voters or candidates to build bridges between racial groups or to form voting coalitions. "Black-preferred" candidates are assured election in "safe black districts"; white-preferred candidates are assured election in "safe white districts." Neither group needs to draw on support from the other's constituency to win on election day. As one judge described the current trend of voting rights cases: "We are bent upon polarizing political subdivisions by race. The arrangement we construct makes it unnecessary, and probably unwise, for an elected official from a white majority district to be responsive at all to the wishes of black citizens; similarly, it is politically unwise for a black official from a black majority district to be responsive at all to white citizens." Dallas County Comm'n, 850 F. 2d, at 1444 (Hill, J., concurring specially).

As this description suggests, the system we have instituted affirmatively encourages a racially based understanding of the representative function. The clear premise of the system is that geographic districts are merely a device to be manipulated to establish "black representatives" whose real constituencies are defined, not in terms of the voters who populate their districts, but in terms of race. The "black representative's" function, in other words, is to represent the "black interest." Cf. Shaw, 509 U. S., at 650 (recognizing that systems that "classify and separate voters by race" threaten "to undermine our system of representative democ-


908

THOMAS, J., concurring in judgment

racy by signaling to elected officials that they represent a particular racial group rather than their constituency as a whole").

Perhaps not surprisingly, the United States has now adopted precisely this theory of racial group representation, as the arguments advanced in another case decided today, Johnson v. De Grandy, post, p. 997, should show. The case involved a claim that an apportionment plan for the Florida Legislature should have provided another Hispanic district in Dade County. Florida responded to the claim of vote dilution by arguing that the plan already provided Dade County Hispanics with seats in proportion to their numbers. According to the Solicitor General, this claim of proportionality should have been evaluated, not merely on the basis of the population in the Dade County area where the racial gerrymandering was alleged to have occurred, but on a statewide basis. It did not matter, in the Solicitor General's view, that Hispanic populations elsewhere in the State could not meet the Gingles geographic compactness test, see 478 U. S., at 50, and thus could not possibly have controlled districts of their own. After all, the Solicitor General reasoned, the Hispanic legislators elected from Hispanic districts in Dade County would represent, not just the interests of the Dade County Hispanics, but the interests of all the Hispanics in the State. Brief for United States in Johnson v. De Grandy, O. T. 1993, No. 92-519, p. 20. As the argument shows, at least some careful observers have recognized the racial gerrymandering in our vote dilution cases for what it is: a slightly less precise mechanism than the racial register for allocating representation on the basis of race.

C

While the results we have already achieved under the Voting Rights Act might seem bad enough, we should recognize that our approach to splintering the electorate into racially designated single-member districts does not by any means


909

mark a limit on the authority federal judges may wield to rework electoral systems under our Voting Rights Act jurisprudence. On the contrary, in relying on single-member districting schemes as a touchstone, our cases so far have been somewhat arbitrarily limited to addressing the interests of minority voters who are sufficiently geographically compact to form a majority in a single-member district. See Gingles, supra, at 49-50. There is no reason a priori, however, that our focus should be so constrained. The decision to rely on single-member geographic districts as a mechanism for conducting elections is merely a political choice-and one that we might reconsider in the future. Indeed, it is a choice that has undoubtedly been influenced by the adversary process:

In the cases that have come before us, plaintiffs have focused largely upon attacking multimember districts and have offered single-member schemes as the benchmark of an "undiluted" alternative.

But as the destructive effects of our current penchant for majority-minority districts become more apparent, cf. Shaw, supra, courts will undoubtedly be called upon to reconsider adherence to geographic districting as a method for ensuring minority voting power. Already, some advocates have criticized the current strategy of creating majority-minority districts and have urged the adoption of other voting mechanisms-for example, cumulative voting 15 or a system using

15 Under a cumulative voting scheme, a system commonly used in corporations to protect the interests of minority shareholders, see R. Clark, Corporate Law § 9.1.3, pp. 361-366 (1986), each voter has as many votes as there are posts to be filled, and the voter may cast as many of his votes as he wishes for a single candidate. The system thus allows a numerical minority to concentrate its voting power behind a given candidate without requiring that the minority voters themselves be concentrated into a single district. For a complete description of the mechanics of cumulative voting, see Zimmerman, The Federal Voting Rights Act and Alternative Election Systems, 19 Wm. & Mary L. Rev. 621, 654-657 (1978).


910

THOMAS, J., concurring in judgment

transferable votes 16-that can produce proportional results without requiring division of the electorate into racially segregated districts. Cf., e. g., Guinier 14-15, 94-101; Howard & Howard 1660; Karlan, Maps and Misreadings: The Role of Geographic Compactness in Racial Vote Dilution Litigation, 24 Harv. Civ. Rights-Civ. Lib. L. Rev. 173, 174-175,231236 (1989) (hereinafter Karlan); Taebel, Engstrom, & Cole, Alternative Electoral Systems As Remedies for Minority Vote Dilution, 11 Hamline J. of Public Law & Policy 19 (1990); Note, Reconciling the Right to Vote with the Voting Rights Act, 92 Colum. L. Rev. 1810, 1857-1865 (1992).

Such changes may seem radical departures from the electoral systems with which we are most familiar. Indeed, they may be unwanted by the people in the several States who purposely have adopted districting systems in their electoral laws. But nothing in our present understanding of the Voting Rights Act places a principled limit on the authority of federal courts that would prevent them from instituting a system of cumulative voting as a remedy under § 2, or even from establishing a more elaborate mechanism for securing proportional representation based on transferable votes.17 As some Members of the Court have already recog-

16 A system utilizing transferable votes is designed to ensure proportional representation with "mathematical exactness." Id., at 640. Under such a system, each voter rank orders his choices of candidates. To win, a candidate must receive a fixed quota of votes, which may be set by any of several methods. Ballots listing a given candidate as the voter's first choice are counted for that candidate until the candidate has secured the quota of votes necessary for election. Remaining first-choice ballots for that candidate are then transferred to another candidate, usually the one listed as the second choice on the ballot. See id., at 640-642. Like cumulative voting, the system allows a minority group to concentrate its voting power without requiring districting, and it has the additional advantage of ensuring that "surplus" votes are transferred to support the election of the minority voters' next preference.

17 Such methods of voting cannot be rejected out-of-hand as bizarre concoctions of Voting Rights Act plaintiffs. The system of transferable votes was a widely celebrated, although unsuccessful, proposal for English par-


911

nized, geographic districting is not a requirement inherent in our political system. See, e. g., Davis v. Bandemer, 478 U. S. 109, 159 (1986) (O'CONNOR, J., concurring in judgment) ("Districting itself represents a middle ground between winner-take-all statewide elections and proportional representation for political parties"); id., at 160 (noting that our current practice of accepting district-based elections as a given is simply a "political judgment"). Rather, districting is merely another political choice made by the citizenry in the drafting of their state constitutions. Like other political choices concerning electoral systems and models of representation, it too is presumably subject to a judicial override if it comes into conflict with the theories of representation and effective voting that we may develop under the Voting Rights Act.

Indeed, the unvarnished truth is that all that is required for districting to fall out of favor is for Members of this Court to further develop their political thinking. We should not be surprised if voting rights advocates encourage us to "revive our political imagination," Guinier, 14 Cardozo L. Rev., at 1137, and to consider "innovative and nontraditional remedies" for vote dilution, Karlan 221, for under our Voting Rights Act jurisprudence, it is only the limits on our "political imagination" that place restraints on the standards we may select for defining undiluted voting systems. Once we candidly recognize that geographic districting and other aspects of electoral systems that we have so far placed beyond question are merely political choices, those practices, too,

liamentary reform in the last century. See generally T. Hare, Election of Representatives (4th ed. 1873); J. S. Mill, Considerations on Representative Government (1861). And while it is an oddity in American political history, cumulative voting in an at-large system has been employed in some American jurisdictions. See Weaver, Semi-Proportional and Proportional Representation Systems in the United States, in Choosing an Electoral System 191, 198 (A. Lijphart & B. Grofman eds. 1984); Hyneman & Morgan, Cumulative Voting in Illinois, 32 Ill. L. Rev. 12 (1937). See also Ill. Const., Art. IV, §§ 7, 8 (1870).


912

THOMAS, J., concurring in judgment

may fall under suspicion of having a dilutive effect on minority voting strength. And when the time comes to put the question to the test, it may be difficult indeed for a Court that, under Gingles, has been bent on creating roughly proportional representation for geographically compact minorities to find a principled reason for holding that a geographically dispersed minority cannot challenge districting itself as a dilutive electoral practice. In principle, cumulative voting and other non-district-based methods of effecting proportional representation are simply more efficient and straightforward mechanisms for achieving what has already become our tacit objective: roughly proportional allocation of political power according to race.

At least one court, in fact, has already abandoned districting and has opted instead for cumulative voting on a countywide basis as a remedy for a Voting Rights Act violation. The District Court for the District of Maryland recently reasoned that, compared to a system that divides voters into districts according to race, "[c]umulative voting is less likely to increase polarization between different interests," and that it "will allow the voters, by the way they exercise their votes, to 'district' themselves," thereby avoiding government involvement in a process of segregating the electorate. Cane v. Worcester County, 847 F. Supp. 369, 373 (1994). Cf. Guinier, 14 Cardozo L. Rev., at 1135-1136 (proposing a similar analysis of the benefits of cumulative voting); Karlan 236 (same). If such a system can be ordered on a countywide basis, we should recognize that there is no limiting principle under the Act that would prevent federal courts from requiring it for elections to state legislatures as well.

D

Such is the current state of our understanding of the Voting Rights Act. That our reading of the Act has assigned the federal judiciary the task of making the decisions I have described above should suggest to the Members of this Court


913

that something in our jurisprudence has gone awry.18 We would be mighty Platonic guardians indeed if Congress had granted us the authority to determine the best form of local government for every county, city, village, and town in America. But under our constitutional system, this Court is not a centralized politburo appointed for life to dictate to the provinces the "correct" theories of democratic representation, the "best" electoral systems for securing truly "representative" government, the "fairest" proportions of minority political influence, or, as respondents would have us hold today, the "proper" sizes for local governing bodies. We should be cautious in interpreting any Act of Congress to grant us power to make such determinations.

JUSTICE BLACKMUN suggests that, if we were to interpret the Act to allow challenges to the size of governmental bodies under § 2, the Court's power to determine the structure that local governing bodies must take would be bounded by the constraints that local customs provide in the form of benchmarks. Post, at 952-953. But as JUSTICE O'CONNOR rightly points out, such benchmarks are themselves arbitrarily selected and would provide no assured limits on judicial power. Ante, at 888-891. In my view, the local standards to which JUSTICE BLACKMUN points today are little different from the various standards to which the Court has resorted in the past as touchstones of undiluted voting systems. The appeal to such standards, which are necessarily arbitrarily chosen, should not serve to obscure the assumption in the Court's vote dilution jurisprudence of a sweeping

18JUSTICE STEVENS suggests that the discussion above outlines policy arguments best addressed to Congress. See post, at 957. In one sense, that is precisely my point. The issues I have discussed above involve policy decisions that are matters best left to Congress. Our interpretation of the Voting Rights Act, however, has required federal courts to take over the policymaking role in the area of voting rights and has forced judges to make decisions on matters beyond the normal sphere of judicial competence.


914

THOMAS, J., concurring in judgment

authority to select the electoral systems to be used by every governing body in each of the 50 States, and to do so based upon little more than the passing preference of five Members of this Court for one political theory over another.

A full understanding of the authority that our current interpretation of the Voting Rights Act assigns to the federal courts, and of the destructive effects that our exercise of that authority is presently having upon our body politic, compels a single conclusion: A systematic reexamination of our interpretation of the Act is required.

II

Section 2(a) of the Voting Rights Act provides that "[n]o voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner which results in a denial or abridgement of the right of any citizen of the United States to vote" on account of race, color, or membership in one of the language minority groups defined in the Act. 42 U. S. C. § 1973(a). Respondents contend that the terms "standard, practice, or procedure" should extend to cover the size of a governmental body. An examination of the text of § 2 makes it clear, however, that the terms of the Act do not reach that far; indeed, the terms of the Act do not allow many of the challenges to electoral mechanisms that we have permitted in the past. Properly understood, the terms "standard, practice, or procedure" in § 2(a) refer only to practices that affect minority citizens' access to the ballot. Districting systems and electoral mechanisms that may affect the "weight" given to a ballot duly cast and counted are simply beyond the purview of the Act.

A

In determining the scope of § 2(a), as when interpreting any statute, we should begin with the statutory language. See Connecticut Nat. Bank v. Germain, 503 U. S. 249, 253-


915

254 (1992). Under the plain terms of the Act, § 2(a) covers only a defined category of state actions. Only "voting qualification[s]," "prerequisite[s] to voting," or "standard[s], practice[s], or procedure[s]" are subject to challenge under the Act. The first two items in this list clearly refer to conditions or tests applied to regulate citizens' access to the ballot. They would cover, for example, any form of test or requirement imposed as a condition on registration or on the process of voting on election day.

Taken in isolation, the last grouping of terms-"standard, practice, or procedure"-may seem somewhat less precise. If we give the words their ordinary meanings, however-for they have no technical significance and are not defined in the Act-they would not normally be understood to include the size of a local governing body. Common sense indicates that the size of a governing body and other aspects of government structure do not comfortably fit within the terms "standard, practice, or procedure." Moreover, we need not simply treat the terms in isolation; indeed, it would be a mistake to do so. Cf. United Sav. Assn. of Tex. v. Timbers of Inwood Forest Associates, Ltd., 484 U. S. 365, 371 (1988). Reading the words in context strongly suggests that § 2(a) must be understood as referring to any standard, practice, or procedure with respect to voting. And thus understood, the terms of the section would not extend to the size of a governmental body; we would not usually describe the size or form of a governing authority as a "practice" or "procedure" concerning voting.

But under our precedents, we have already stretched the terms "standard, practice, or procedure" beyond the limits of ordinary meaning. We have concluded, for example, that the choice of a certain set of district lines is a "procedure," or perhaps a "practice," concerning voting subject to challenge under the Act, see Growe, 507 U. S., at 40-41, even though the drawing of a given set of district lines has nothing to do with the basic process of allowing a citizen to vote-that is,


916

THOMAS, J., concurring in judgment

the process of registering, casting a ballot, and having it counted. Similarly, we have determined that the use of multimember districts, rather than single-member districts, can be challenged under the Act. See Gingles, 478 U. S., at 46-51. Undoubtedly, one of the critical reasons we have read § 2 to reach such districting decisions is that the choice of one districting system over another can affect a minority group's power to control seats in the elected body. See ibid. In that respect, however, the districting practices we have treated as subject to challenge under the Act are essentially similar to choices concerning the size of a governing authority. Just as drawing district lines one way rather than another, or using one type of districting system rather than another, can affect the ability of a minority group to control seats, so can restricting the number of seats that are available. And if how districts are drawn is a "practice" concerning voting, why not conclude that how many districts are drawn is a "practice" as well?

To be sure, a distinction can be made between the size of a local governing body and a districting mechanism. After all, we would ordinarily think that the size of a government has greater independent significance for the functioning of the governmental body than the choice of districting systems apportioning representation. Interfering with the form of government, therefore, might appear to involve a greater intrusion on state sovereignty. But such distinctions between the size of a governing body and other potential "voting practices" do not, at bottom, depend upon how closely each is related to "voting," and thus they are not rooted in any way in the text of § 2(a). On the contrary, while it may seem obvious that the size of a government is not within the reach of the Act, if we look to the text of the statute for the limiting principle that confines the terms "standard, practice, or procedure" and excludes government size from their reach, we must conclude that the only line drawn in § 2 excludes many


917

"practices" that we have already decided are subject to challenge under the Act.

If we return to the Act to reexamine the terms setting out the actions regulated by § 2, a careful reading of the statutory text will reveal a good deal more about the limitations on the scope of the section than suggested above. The terms "standard, practice, or procedure" appear to have been included in § 2 as a sort of catchall provision. They seem phrased with an eye to eliminating the possibility of evasion.19 Nevertheless, they are catchall terms that round out a list, and a sensible and long-established maxim of construction limits the way we should understand such general words appended to an enumeration of more specific items. The principle of ejusdem generis suggests that such general terms should be understood to refer to items belonging to the same class that is defined by the more specific terms in the list. See, e. g., Cleveland v. United States, 329 U. S. 14, 18 (1946).

Here, the specific items described in § 2(a) ("voting qualification[sJ" and "prerequisite[sJ to voting") indicate that Congress was concerned in this section with any procedure, however it might be denominated, that regulates citizens' access to the ballot-that is, any procedure that might erect a barrier to prevent the potential voter from casting his vote. In describing the laws that would be subject to § 2, Congress focused attention upon provisions regulating the interaction between the individual voter and the voting process-on hurdles the citizen might have to cross in the form of "prerequisites" or "qualifications." The general terms in the section are most naturally understood, therefore, to refer to

19 Cf. South Carolina v. Katzenbach, 383 U. S. 301, 335 (1966) (noting that "Congress knew that some of the 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 that "Congress had reason to suppose that these States might try similar maneuvers in the future").


918

THOMAS, J., concurring in judgment

any methods for conducting a part of the voting process that might similarly be used to interfere with a citizen's ability to cast his vote, and they are undoubtedly intended to ensure that the entire voting process-a process that begins with registration and includes the casting of a ballot and having the ballot counted-is covered by the Act. Cf. infra, at 919920. Simply by including general terms in § 2(a) to ensure the efficacy of the restriction imposed, Congress should not be understood to have expanded the scope of the restriction beyond the logical limits implied in the specific terms of the statute. Cf. Cleveland, supra, at 18 ("Under the ejusdem generis rule of construction the general words are confined to the class and may not be used to enlarge it").

Moreover, it is not only in the terms describing the practices regulated under the Act that § 2(a) focuses on the individual voter. The section also speaks only in the singular of the right of "any citizen" to vote. Giving the terms "standard, practice, or procedure" an expansive interpretation to reach potentially dilutive practices, however, would distort that focus on the individual, for a vote dilution claim necessarily depends on the assertion of a group right. Cf. Bandemer, 478 U. S., at 150-151 (O'CONNOR, J., concurring in judgment). At the heart of the claim is the contention that the members of a group collectively have been unable to exert the influence that their numbers suggest they might under an alternative system. Such a group right, however, finds no grounding in the terms of § 2(a).

Of course, the scope of the right that is protected under the Act can provide further guidance concerning the meaning of the terms "standard, practice, or procedure." Under the terms of the Act, only a "standard, practice, or procedure" that may result in the "denial or abridgement of the right ... to vote" is within the reach of § 2(a). But nothing in the language used in § 2(a) to describe the protection provided by the Act suggests that in protecting the "right to vote," the section was meant to incorporate a concept of vot-


919

ing that encompasses a concern for the "weight" or "influence" of votes. On the contrary, the definition of the terms "vote" and "voting" in § 14(c)(1) of the Act focuses precisely on access to the ballot. Thus, § 14(c)(1) provides that the terms "vote" and "voting" shall encompass any measures necessary to ensure "registration" and any "other action required by law prerequisite to voting, casting a ballot, and having such ballot counted properly and included in the appropriate totals of votes cast." 42 U. S. C. § 1973l(c)(1).

It is true that § 14(c)(1) also states that the term "voting" "include[s] all action necessary to make a vote effective," ibid. (emphasis added), and the Court has seized on this language as an indication that Congress intended the Act to reach claims of vote dilution. See Allen, 393 U. S., at 566. But if the word "effective" is not plucked out of context, the rest of § 14(c)(1) makes clear that the actions Congress deemed necessary to make a vote "effective" were precisely the actions listed above: registering, satisfying other voting prerequisites, casting a ballot, and having it included in the final tally of votes cast. These actions are described in the section only as examples of the steps necessary to make a vote effective. See 42 U. S. C. § 1973l(c)(1). And while the list of such actions is not exclusive, the nature of all the examples that are provided demonstrates that as far as the Act is concerned, an "effective" vote is merely one that has been cast and fairly counted. See 393 U. S., at 590, n. 7 (Harlan, J., concurring in part and dissenting in part).

Reading the Act's prohibition of practices that may result in a "denial or abridgement of the right ... to vote" as protecting only access to the ballot also yields an interpretation that is consistent with the Court's construction of virtually identical language in the Fifteenth Amendment. The use of language taken from the Amendment suggests that the section was intended to protect a "right to vote" with the same scope as the right secured by the Amendment itself; certainly, no reason appears from the text of the Act for giving


920

THOMAS, J., concurring in judgment

the language a broader construction in the statute than we have given it in the Constitution. The Court has never decided, however, whether the Fifteenth Amendment should be understood to protect against vote "dilution." See Voinovich v. Quilter, 507 U. S. 146, 159 (1993). See also Beer v. United States, 425 U. S. 130, 142, n. 14 (1976) (noting that there is no decision of this Court holding a legislative apportionment plan violative of the Fifteenth Amendment).2o

While the terms of § 2(a) thus indicate that the section focuses only on securing access to the ballot, it might be argued that reenactment of § 2 in 1982 should be understood as an endorsement of the interpretation contained in cases such as Allen that the terms "standard, practice, or procedure" were meant to reach potentially dilutive practices. See Lorillard v. Pons, 434 U. S. 575, 580-581 (1978). It is true that we generally will assume that reenactment of specific statutory language is intended to include a "settled judicial interpretation" of that language. Pierce v. Underwood, 487 U. S. 552, 567 (1988). And while § 2 was amended in

20 Indeed, in Mobile v. Bolden, 446 U. S. 55 (1980), a plurality of the Court concluded that the Fifteenth Amendment did not address concerns of dilution at all. See id., at 65. Cf. id., at 84, n. 3 (STEVENS, J., concurring in judgment) (noting that the plurality had concluded that the Fifteenth Amendment "applies only to practices that directly affect access to the ballot and hence is totally inapplicable to the case at bar").

Contrary to JUSTICE STEVENS' suggestions, post, at 958, 962, Gomillion v. Lightfoot, 364 U. S. 339 (1960), does not indicate that the Fifteenth Amendment, in protecting the right to vote, incorporates a concern for anything beyond securing access to the ballot. The Gomillion plaintiffs' claims centered precisely on access: Their complaint was not that the weight of their votes had been diminished in some way, but that the boundaries of a city had been drawn to prevent blacks from voting in municipal elections altogether. Id., at 341. Gomillion thus "maintains the distinction between an attempt to exclude Negroes totally from the relevant constituency, and a statute that permits Negroes to vote but which uses the gerrymander to contain the impact of Negro suffrage." Allen v. State Bd. of Elections, 393 U. S. 544, 589 (1969) (Harlan, J., concurring in part and dissenting in part).


921

1982, the amended section did retain the same language that had appeared in the original Act regulating "standard[s], practice [s], or procedure[s]." 21 But it was hardly well settled in 1982 that Allen's broad reading of the terms "standard, practice, or procedure" in § 5 would set the scope of § 2 as a provision reaching claims of vote dilution.

On the contrary, in 1980 in Mobile v. Bolden, 446 U. S. 55, a plurality of the Court construed § 2 in a manner flatly inconsistent with the understanding that those terms were meant to reach dilutive practices. Emphasizing that the section tracked the language of the Fifteenth Amendment by prohibiting the use of practices that might "deny or abridge the right ... to vote," the Bolden plurality determined that § 2 was "intended to have an effect no different from that of the Fifteenth Amendment itself." Id., at 61. In the plurality's view, however, the Fifteenth Amendment did not extend to reach dilution claims; its protections were satisfied as long as members of racial minorities could "'register and vote without hindrance.'" Id., at 65. Bolden remained the last word from this Court interpreting § 2 at the time the section was amended in 1982. Cf. Rogers v. Lodge, 458 U. S. 613, 619, n. 6 (1982). Thus, the reenactment in the amended section of the same language covering any "standard, practice, or procedure" and the retention of virtually identical language protecting against the "denial or abridgement of the right ... to vote" can hardly be understood as an endorsement of a broad reading of the section as a provision reaching claims of vote dilution.22

21 The original § 2 provided that no "standard, practice, or procedure" should be imposed or applied "to deny or abridge the right ... to vote." Pub. L. 89-110, §2, 79 Stat. 437.

22 If anything, applying the Lorillard v. Pons, 434 U. S. 575 (1978), principle of construction might suggest that, by reenacting virtually the same language derived from the Fifteenth Amendment to define the basic interest protected by the Act, Congress intended to preserve the limitation that the Bolden plurality found implicit in that language. It is clear from the terms of the amendments passed in 1982 that where Congress sought


922

THOMAS, J., concurring in judgment

Finally, as our cases have shown, reading § 2(a) to reach beyond laws that regulate in some way citizens' access to the ballot turns the section into a command for courts to evaluate abstract principles of political theory in order to develop rules for deciding which votes are "diluted" and which are not. See generally supra, at 894-903. Common sense would suggest that we should not lightly interpret the Act to require courts to address such matters so far outside the normal bounds of judicial competence, and the mere use of three more general terms at the end of the list of regulated practices in § 2(a) cannot properly be understood to incorporate such an expansive command into the Act.

Properly understood, therefore, § 2(a) is a provision designed to protect access to the ballot, and in regulating "standard[s], practice [s], and procedure[s]," it reaches only "those state laws that [relate to] either voter qualifications or the manner in which elections are conducted." Allen, 393 U. S., at 591 (Harlan, J., concurring in part and dissenting in part). The section thus covers all manner of registration requirements, the practices surrounding registration (including the selection of times and places where registration takes place and the selection of registrars), the locations of polling places, the times polls are open, the use of paper ballots as opposed to voting machines, and other similar aspects of the voting process that might be manipulated to deny any citizen the right to cast a ballot and have it properly counted. The

to alter the understanding of the Act announced in Bolden, it did so explicitly in the text of the statute. As I explain more fully, infra, at 923-925, the 1982 amendments modified § 2 to eliminate the requirement under Bolden that § 2 plaintiffs, like plaintiffs under the Fifteenth Amendment, show that a challenged practice was adopted with a discriminatory intent, see 446 U. S., at 62-63, and replaced that test with specific language in § 2(b) setting a standard based simply on discriminatory results. See Pub. L. 97-205, § 3, 96 Stat. 134. Had Congress intended to alter the understanding that § 2 protects a concept of the "right to vote" that does not extend to prohibit vote dilution, it likely would have addressed that aspect of Bolden explicitly as well.


923

section does not cover, however, the choice of a multimember over a single-member districting system or the selection of one set of districting lines over another, or any other such electoral mechanism or method of election that might reduce the weight or influence a ballot may have in controlling the outcome of an election.

Of course, this interpretation of the terms "standard, practice, or procedure" effectively means that § 2(a) does not provide for any claims of what we have called vote "dilution." But that is precisely the result suggested by the text of the statute. Section 2(a) nowhere uses the term "vote dilution" or suggests that its goal is to ensure that votes are given their proper "weight." And an examination of § 2(b) does not suggest any different result. It is true that in construing § 2 to reach vote dilution claims in Thornburg v. Gingles, 478 U. S. 30 (1986), the Court relied largely on the gloss on § 2(b) supplied in the legislative history of the 1982 amendments to the Act. See id., at 43-46. But the text of § 2(b) supplies a weak foundation indeed for reading the Act to reach such claims.

As the Court concluded in Gingles, the 1982 amendments incorporated into the Act, and specifically into § 2(b), a "results" test for measuring violations of § 2(a). That test was intended to replace, for § 2 purposes, the "intent" test the Court had announced in Bolden for voting rights claims under § 2 of the Voting Rights Act and under the Fourteenth and Fifteenth Amendments. Section 2(a) thus prohibits certain state actions that may "resul[t] in a denial or abridgement" of the right to vote, and § 2(b) incorporates virtually the exact language of the "results test" employed by the Court in White v. Regester, 412 U. S. 755 (1973), and applied in constitutional voting rights cases before our decision in Bolden. The section directs courts to consider whether "based on the totality of circumstances," a state practice results in members of a minority group "hav[ing] less opportunity than other members of the electorate to participate in


924

THOMAS, J., concurring in judgment

the political process and to elect representatives of their choice." 42 U. S. C. § 1973(b). Cf. White, supra, at 766; Whitcomb v. Chavis, 403 U. S. 124, 149 (1971).

But the mere adoption of a "results" test, rather than an "intent" test, says nothing about the type of state laws that may be challenged using that test. On the contrary, the type of state law that may be challenged under § 2 is addressed explicitly in § 2(a). As we noted in Chisom v. Roemer, 501 U. S. 380 (1991), §§ 2(a) and (b) address distinct issues. While § 2(a) defines and explicitly limits the type of voting practice that may be challenged under the Act, § 2(b) provides only "the test for determining the legality of such a practice." Id., at 391. Thus, as an initial matter, there is no reason to think that § 2(b) could serve to expand the scope of the prohibition in § 2(a), which, as I described above, does not extend by its terms to electoral mechanisms that might have a dilutive effect on group voting power.

Even putting that concern aside for the moment, it should be apparent that the incorporation of a results test into the amended section does not necessarily suggest that Congress intended to allow claims of vote dilution under § 2. A results test is useful to plaintiffs whether they are challenging laws that restrict access to the ballot or laws that accomplish some diminution in the "proper weight" of a group's vote. Nothing about the test itself suggests that it is inherently tied to vote dilution claims. A law, for example, limiting the times and places at which registration can occur might be adopted with the purpose of limiting black voter registration, but it could be extremely difficult to prove the discriminatory intent behind such a facially neutral law. The results test would allow plaintiffs to mount a successful challenge to the law under § 2 without such proof.

Moreover, nothing in the language § 2(b) uses to describe the results test particularly indicates that the test was intended to be used under the Act for assessing claims of dilution. Section 2(b) directs courts to consider whether, under


925

the "totality of circumstances," members of a minority group "have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice." 42 U. S. C. § 1973(b). The most natural reading of that language would suggest that citizens have an equal "opportunity" to participate in the electoral process and an equal "opportunity" to elect representatives when they have been given the same free and open access to the ballot as other citizens and their votes have been properly counted. The section speaks in terms of an opportunity-a chance-to participate and to elect, not an assured ability to attain any particular result. And since the ballot provides the formal mechanism for obtaining access to the political process and for electing representatives, it would seem that one who has had the same chance as others to register and to cast his ballot has had an equal opportunity to participate and to elect, whether or not any of the candidates he chooses is ultimately successful.

To be sure, the test in § 2(b) could be read to apply to claims of vote dilution as well. But to conclude, for example, that a multimember districting system had denied a group of voters an equal opportunity to participate in the political process and to elect representatives, a court would have to embark on the extended project in political theory that I described above in Part I of this opinion. In other words, a court would have to develop some theory of the benchmark undiluted voting system that provides minorities with the "fairest" or most "equitable" share of political influence. Undoubtedly, a dizzying array of concepts of political equality might be described to aid in that task, and each could be used to attribute different values to different systems of election. See, e. g., Still, Political Equality and Election Systems, 91 Ethics 375 (1981).23 But the statutory command to deter-

23 See also Banzhaf, Multi-Member Electoral Districts-Do They Violate the "One Man, One Vote" Principle, 75 Yale L. J. 1309 (1966) (suggesting that how close different districting systems come to providing persons


926

THOMAS, J., concurring in judgment

mine whether members of a minority have had an equal "opportunity ... to participate in the political process and to elect representatives" provides no guidance concerning which one of the possible standards setting undiluted voting strength should be chosen over the others. And it would be contrary to common sense to read § 2(b)'s reference to equal opportunity as a charter for federal courts to embark on the ambitious project of developing a theory of political equality to be imposed on the Nation.24

It is true that one factor courts may consider under the results test might fit more comfortably with an interpretation of the Act that reaches vote dilution claims. Section 2(b) provides that "one circumstance" that may be considered in assessing the results test is the "extent to which members of a protected class have been elected to office." 42 U. S. c. § 1973(b). Obviously, electoral outcomes would be relevant to claims of vote dilution (assuming, of course, that control

equal political "power" can be measured by comparing the statistical probability under each system that a person's vote will determine the election result). Cf. Whitcomb, 403 U. S., at 145, n. 23.

24 In addition, in one respect there is a significant tension between the terms of the results test and an interpretation of the Act that reaches vote dilution claims. Section 2(b) provides that a violation may be established where it is shown that members of a minority have less opportunity than other members of the electorate "to participate in the political process and to elect representatives of their choice." 42 U. S. C. § 1973(b) (emphasis added). We have held that any challenged "standard, practice, or procedure" must have both of these effects to violate the test outlined in § 2(b). See Chisom v. Roemer, 501 U. S. 380, 397 (1991). It is not clear, however, that a potentially dilutive districting method can satisfy both prongs of the test. The primary effect of the choice of one districting system over another will be the direct and mathematically quantifiable impact that the system will have on a minority group's ability to control a given number of seats. But even if one assumes that a districting system may therefore be said to impair a group's "opportunity" to "elect representatives of its choice," it is difficult to see how a districting system could be said to impair a group's opportunity to "participate in the political process," at least if participation is understood to have any meaning distinct from controlling seats.


927

of seats has been selected as the measure of effective voting). But in some circumstances, results in recent elections might also be relevant for demonstrating that a particular practice concerning registration or polling has served to suppress minority voting. Better factors to consider would be figures for voter registration or turnout at the last election, broken down according to race. But where such data are not readily available, election results may certainly be "one circumstance" to consider in determining whether a challenged practice has resulted in denying a minority group access to the political process. The Act merely directs courts not to ignore such evidence of electoral outcomes altogether.

Moreover, the language providing that electoral outcomes may be considered as "one circumstance" in the results test is explicitly qualified by the provision in § 2(b) that most directly speaks to the question whether § 2 was meant to reach claims of vote dilution-and which suggests that dilution claims are not covered by the section. The last clause in the subsection states in unmistakable terms that "nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population." 42 U. s. C. § 1973(b). As four Members of the Court observed in Gingles, there is "an inherent tension" between this disclaimer of proportional representation and an interpretation of § 2 that encompasses vote dilution claims. 478 U. S., at 84 (O'CONNOR, J., concurring in judgment). As I explained above, dilution claims, by their very nature, depend upon a mathematical principle. The heart of the claim is an assertion that the plaintiff group does not hold the "proper" number of seats. As a result, the principle for deciding the case must be supplied by an arithmetic ratio. Either the group has attained the "proper" number of seats under the current election system, or it has not.

By declaring that the section provides no right to proportional representation, § 2(b) necessarily commands that the existence or absence of proportional electoral results should


928

THOMAS, J., concurring in judgment

not become the deciding factor in assessing § 2 claims. But in doing so, § 2(b) removes from consideration the most logical ratio for assessing a claim of vote dilution. To resolve a dilution claim under § 2, therefore, a court either must arbitrarily select a different ratio to represent the "undiluted" norm, a ratio that would have less intuitive appeal than direct proportionality, or it must effectively apply a proportionality test in direct contravention of the text of the Act 25-hence the "inherent tension" between the text of the Act and vote dilution claims. Given that § 2 nowhere speaks in terms of "dilution," an explicit disclaimer removing from the field of play the most natural deciding principle in dilution cases is surely a strong signal that such claims do not fall within the ambit of the Act.26

25 As I discuss more fully below, our cases have pursued the latter option. See infra, at 936-944.

26 In Johnson v. De Grandy, post, p. 997, the Court suggests that §2(b) disclaims only a guarantee of success for minority candidates and thus that it has nothing to say concerning remedial schemes designed to provide a minority group proportional control over seats. See post, at 1014, n. 11. See also post, at 1026-1027 (KENNEDY, J., concurring in part and concurring in judgment). Minority control, of course, mayor may not result in the election of minority candidates. The Court's reading of the disclaimer, in my view, distorts the obvious import of the provision. The clause rejecting a group's right to elect its own members in proportion to their numbers must be understood as a disclaimer of a minority group's right to proportional political power. Otherwise, in practical terms, the clause would be reduced to a nullity.

It should be clear that a system that gives a minority group proportional control effectively provides the "right" to elect a proportionate number of minority candidates that the Act disclaims. Whether that right is utilized by minority voters to elect minority candidates is a matter of the voters' choice. The De Grandy Court's position seems to be that the proviso is directed, not at a system intended to guarantee the ability to elect minority candidates in proportion to the minority's numbers, but only at a system that will invariably guarantee the election of a proportionate number of minority candidates. Only one system would fit that description: a system based on a racial register in which a quota of seats are set aside for members of a minority group. I think it would be preposterous


929

It is true that the terms "standard, practice, or procedure" in § 5 of the Act have been construed to reach districting systems and other potentially dilutive electoral mechanisms, see, e. g., Allen, 393 U. S., at 569, and Congress has reenacted § 5 subsequent to our decisions adopting that expansive interpretation. See, e. g., United States v. Sheffield Bd. of Comm'rs, 435 U. S. 110, 134-135 (1978); Georgia v. United States, 411 U. S. 526, 533 (1973). Nevertheless, the text of the section suggests precisely the same focus on measures that relate to access to the ballot that appears in § 2. Section 5 requires covered jurisdictions to obtain preclearance for a change in "any voting qualification or prerequisite to voting, or standard, practice, or procedure with respect to voting." 42 U. S. C. § 1973c. As in § 2, the specific terms in the list of regulated state actions describe only laws that would limit access to the ballot. Moreover, § 5 makes the focus on the individual voter and access to the voting booth even more apparent as the section goes on to state that "no person shall be denied the right to vote for failure to comply with such qualification, prerequisite, standard, practice, or procedure." 42 U. S. C. § 1973c (emphasis added). This command makes it explicit that in regulating standards, practices, or procedures with respect to voting, "Congress was clearly concerned with changes in procedure with which

to suggest that the disclaimer in § 2(b) was intended solely to prohibit the use of such a system. Such a device has never, to my knowledge, been proposed in any voting rights case. Moreover, to the extent that the decisions in White and Whitcomb can inform our understanding of § 2(b), they suggest that in expressing a concern that "proportionality" not be used as the measure of a voting rights violation, Congress was concerned with proportional electoral power, not merely proportional election of minority candidates. See, e. g., Whitcomb, 403 U. S., at 153 (rejecting the "failure of [the minority group] to have legislative seats in proportion to its population" as a sufficient basis for a claim) (emphasis added). The proviso has been understood in the past simply as a disclaimer of a right to proportional representation, see, e. g., Gingles, 478 U. S., at 84-86,94 (O'CONNOR, J., concurring in judgment), and I think that understanding is correct.


930

THOMAS, J., concurring in judgment

voters could comply." Allen, supra, at 587 (Harlan, J., concurring in part and dissenting in part). But it should be obvious that a districting system, or any other potentially dilutive mechanism for that matter, is not something with which a voter can comply. As is the case with § 2, § 5's description of the terms "standard, practice, or procedure" thus suggests a focus on rules that regulate the individual voter's ability to register and cast a ballot, not a more abstract concern with the effect that various electoral systems might have on the "weight" of the votes cast by a group that constitutes a numerical minority in the electorate.

In my view, the tension between the terms of the Act and the construction we have placed on § 5 at the very least suggests that our interpretation of § 5 should not be adopted wholesale to supply the meaning of the terms "standard, practice, or procedure" under § 2. An expansive construction of § 5 was well established in 1980, yet a plurality of the Court in Bolden, after focusing on the terms of the Act, did not adopt a similarly expansive construction of § 2. Rather, the Bolden plurality concluded that § 2 should be strictly limited to have the same reach as the Fifteenth Amendment, which the plurality interpreted as addressing only matters relating to access to the ballot. See Bolden, 446 U. S., at 61, 65. I would reach a similar result here. Where a careful reading of the language of § 2 dictates a narrow interpretation of the section, there is no reason for transplanting our interpretation of the terms of § 5-an interpretation that I believe is in tension with the text of § 5 itself-to another section of the Act.27

27 I need not decide in this case whether I would overrule our decisions construing the terms "standard, practice, or procedure" in § 5; the challenge here involves only § 2. Although in my view our construction of § 5 may well be incorrect as a matter of first impression, stare decisis would suggest that such an error in prior decisions may not in itself justify overruling settled precedent. Determining whether to abandon prior decisions requires weighing a multitude of factors, one of the most important of which is the extent to which the decisions in question have proved unworkable. Cf. infra, at 936-937. In that regard, while the practical


931

B

From the foregoing, it should be clear that, as far as the text of the Voting Rights Act is concerned, "§ 2 does not speak in terms of 'vote dilution.'" Gingles, 478 U. S., at 87 (O'CONNOR, J., concurring in judgment). One might wonder, then, why we have consistently concluded that "[w]e know that Congress intended to allow vote dilution claims to be brought under § 2." Id., at 84. The juxtaposition of the two statements surely makes the result in our cases appear extraordinary, since it suggests a sort of statutory construction through divination that has allowed us to determine that Congress "really meant" to enact a statute about vote dilution even though Congress did not do so explicitly. In truth, our method of construing § 2 has been only little better than that, for the only source we have relied upon for the expansive meaning we have given § 2 has been the legislative history of the Act.

We first considered the amended § 2 in Thornburg v. Gingles. Although the precise scope of the terms "standard, practice, or procedure" was not specifically addressed in that case, Gingles nevertheless established our current interpretation of the amended section as a provision that addresses vote dilution, and in particular it fixed our understanding that the results test in § 2(b) is intended to measure vote dilution in terms of electoral outcomes. See id., at 93 (O'CONNOR, J., concurring in judgment) (stating that Gingles made electoral results the "linchpin" of vote dilution claims). In reaching its interpretation of § 2, the Gingles Court re-

differences in the application of §§ 2 and 5 that JUSTICE KENNEDY points out, see ante, at 883-884, would not, in my view, suggest as an original matter that the same terms in the two sections should be read to have different meanings, JUSTICE KENNEDY'S observations might suggest that different considerations would have a bearing on the question whether our past interpretations should be abandoned in the § 5 and § 2 contexts. Indeed, in the § 5 context they might suggest a contrary conclusion to the result I reach under § 2. See infra, at 936-945. That, however, is a question for another day.


932

THOMAS, J., concurring in judgment

jected the argument advanced by the United States as amicus curiae that § 2(b)'s test based on an equal "opportunity ... to participate in the political process and to elect representatives" suggested a focus on nothing more than securing equal access to the political process, not a focus on measuring the influence of a minority group's votes in terms of electoral outcomes. See Brief for United States as Amicus Curiae in Thornburg v. Gingles, O. T. 1985, No. 83-1968, pp. 7-19. That understanding of § 2 is, of course, compatible with the interpretation I have set out above.

In approaching § 2, the Gingles Court, based on little more than a bald assertion that "the authoritative source for legislative intent lies in the Committee Reports on the bill," 478 U. S., at 43, n. 7, bypassed a consideration of the text of the Act and proceeded to interpret the section based almost exclusively on its legislative history.28 It was from the legislative history that the Court culled its understanding that § 2

28 In offering two citations to support the sweeping proposition that committee reports provide the authoritative source for legislative intent, Gingles plainly misread the import of our prior decisions. Far from giving an unqualified endorsement of committee reports as a guide to congressional intent, the Court in Garcia v. United States, 469 U. S. 70 (1984), merely indicated that, when resort to legislative history is necessary, it is only committee reports, not the various other sources of legislative history, that should be considered. See id., at 76. The Court, however, carefully repeated Justice Jackson's admonition that "[r]esort to legislative history is only justified where the face of the [statute] is inescapably ambiguous." Id., at 76, n. 3 (quoting Schwegmann Brothers v. Calvert Distillers Corp., 341 U. S. 384, 395 (1951) (concurring opinion)). Similarly, in Zuber v. Allen, 396 U. S. 168 (1969), we considered the reliability of committee reports only as a relative matter in comparing them to statements made by individual Congressmen during floor debates. See id., at 186.

Even if I agreed with Justice Jackson that resort to legislative history is permissible when the text of a statute is "inescapably ambiguous," I could not agree with the use the Court has made of legislative history in interpreting § 2. I think it is clear, first, that in interpreting § 2 the Court has never undertaken any inquiry into the meaning of the plain language of the statute to determine whether it is ambiguous, and second, that the text of § 2 is not riddled with such hopeless ambiguity.


933

is a provision encompassing claims that an electoral system has diluted a minority group's vote and its understanding that claims of dilution are to be evaluated based upon how closely electoral outcomes under a given system approximate the outcomes that would obtain under an alternative, undiluted norm. See, e. g., id., at 43-51.

Contrary to the remarkable "legislative history first" method of statutory construction pursued in Gingles, however, I had thought it firmly established that the "authoritative source" for legislative intent was the text of the statute passed by both Houses of Congress and presented to the President, not a series of partisan statements about purposes and objectives collected by congressional staffers and packaged into a committee report. "We have stated time and again that courts must presume that a legislature says in a statute what it means and means in a statute what it says there." Germain, 503 U. S., at 253-254. See also United States v. Ron Pair Enterprises, Inc., 489 U. S. 235, 241-242 (1989); Oneale v. Thornton, 6 Cranch 53, 68 (1810). Nevertheless, our analysis in Gingles was marked conspicuously by the absence of any attempt to pursue a close reading of the text of the Act. As outlined above, had the Court addressed the text, it would have concluded that the terms of the Act do not address matters of vote "dilution."

Moreover, the legislative history of § 2 itself, and the Court's use of it in Gingles, aptly illustrate that legislative history is often used by this Court as "a forensic rather than an interpretive device," Wisconsin Public Intervenor v. Mortier, 501 U. S. 597, 621 (1991) (SCALIA, J., concurring in judgment), and is read selectively to support the result the Court intends to achieve. It is well documented in the history of the 1982 amendments to the Act that § 2 was passed only after a compromise was reached through the addition of the provision in § 2(b) disclaiming any right to proportional representation. See S. Rep. No. 97-417, pp. 2-4 (1982); id., at 94-97 (additional views of Sen. Hatch). But the views of


934

THOMAS, J., concurring in judgment

the author of that compromise, Senator Dole, hardly coincide with the gloss the Court has placed on § 2.

According to Senator Dole, amended § 2 would "[a]bsolutely not" provide any redress to a group of voters challenging electoral mechanisms in a jurisdiction "if the process is open, if there is equal access, if there are no barriers, direct or indirect, thrown up to keep someone from voting or having their vote counted, or registering, whatever the process may include." 128 Congo Rec. 14133 (1982). Contrary to the Court's interpretation of the section in Gingles, Senator Dole viewed § 2 as a provision more narrowly focused on access to the processes surrounding the casting of a ballot, not a provision concerned with ensuring electoral outcomes in accordance with some "undiluted" norm. See S. Rep. No. 97-417, supra, at 193-194 (additional views of Sen. Dole). The legislative history thus hardly provided unambiguous support for the Court's interpretation; indeed, it seems that the Court used what was helpful to its interpretation in the legislative history and ignored what was not. Cf. Mortier, supra, at 617 (SCALIA, J., concurring in judgment).

Of course, as mentioned above, Gingles did not directly address the meaning of the terms "standard, practice, or procedure" in § 2(a). The understanding that those terms extend to a State's laws establishing various electoral mechanisms dates to our decision in Allen, in which we construed the identical terms in § 5 of the Act. But the Court's method of statutory construction in Allen was little different from that pursued in Gingles, and as the analysis of the text of § 5 above demonstrates, it similarly yielded an interpretation in tension with the terms of the Act.

In Allen, after noting that § 14(c)(1) defined "voting" to include "all action necessary to make a vote effective," 42 U. S. C. § 1973l(c)(1), the Court abandoned any further attempt to construe the text of the Act and went on, instead, to conclude that the "legislative history on the whole supports the view that Congress intended to reach any state


935

enactment which altered the election law of a covered State in even a minor way." Allen, 393 U. S., at 566. Not surprisingly, the legislative history relied upon in Allen also displayed the typical flaws that one might expect-it was hardly unequivocal. See id., at 590-591, and n. 9 (Harlan, J., concurring in part and dissenting in part) (noting inconsistencies in the legislative history). Thus, to the extent that Allen implicitly has served as the basis for our subsequent interpretation of the terms of § 2, it hardly can be thought to provide any surer rooting in the language of the Act than the method of statutory construction pursued in Gingles.

Remarkably, thanks to our reliance on legislative history, we have interpreted § 2 in such a way that four Members of this Court at one time candidly admitted that "[t]here is an inherent tension [in § 2] between what Congress wished to do and what it wished to avoid." Gingles, 478 U. S., at 84 (O'CONNOR, J., concurring in judgment). But our understanding of what Congress purportedly "wished to do" -that is, to allow claims of vote "dilution"-depends solely on a selective reading of legislative history, whereas Congress' statement of what it "wished to avoid" appears explicitly in § 2(b)'s disclaimer of a right to proportional representation. I can see no logical reason to import the "inherent tension" between these two imperatives into the Act, when on its face the statute incorporates only one of two potentially contradictory commands. I would have thought the key to resolving any such conflict between the text and the legislative history obvious: The text of the statute must control, and the text of § 2 does not extend the Act to claims of dilution.

Were it our function to interpret and apply committee reports or other pieces of legislative history, rather than Acts of Congress, I might conclude that we had made the best of a bad situation in interpreting § 2 of the Voting Rights Act, and that the quagmire that is § 2 was Congress' creation, not our own. It is apparent, however, that we have arrived at our current understanding of the Act, with all of its attend-


936

THOMAS, J., concurring in judgment

ant pitfalls, only by abandoning proper methods of statutory construction. Our errors in method in past cases ordinarily might not indicate a need to forsake an established line of precedent. But here they have produced an "inherent tension" between our interpretation of § 2 and the text of the Act and have yielded a construction of the statute that, as I discuss below, is so unworkable in practice and destructive in its effects that it must be repudiated.

C

"Stare decisis is not an inexorable command," Payne v.

Tennessee, 501 U. S. 808, 828 (1991). Indeed, "when governing decisions are unworkable or are badly reasoned, this Court has never felt constrained to follow precedent." Id., at 827 (internal quotation marks omitted). The discussion above should make clear that our decision in Gingles interpreting the scope of § 2 was badly reasoned; it wholly substituted reliance on legislative history for analysis of statutory text. In doing so, it produced a far more expansive interpretation of § 2 than a careful reading of the language of the statute would allow.

Our interpretation of § 2 has also proved unworkable. As I outlined above, it has mired the federal courts in an inherently political task-one that requires answers to questions that are ill-suited to principled judicial resolution. Under § 2, we have assigned the federal judiciary a project that involves, not the application of legal standards to the facts of various cases or even the elaboration of legal principles on a case-by-case basis, but rather the creation of standards from an abstract evaluation of political philosophy.

Worse, our interpretation of § 2 has required us to distort our decisions to obscure the fact that the political choice at the heart of our cases rests on precisely the principle the Act condemns: proportional allocation of political power according to race. Continued adherence to a line of decisions that necessitates such dissembling cannot possibly promote what


937

we have perceived to be one of the central values of the policy of stare decisis: the preservation of "the actual and perceived integrity of the judicial process." Payne, supra, at 827.

I have endeavored to explain above that the core of any vote dilution claim is an assertion that the plaintiff group does not hold seats in the proportion that it should.29 There is no logical way to avoid reliance on a simple ratio in evaluating such a claim. And allocation of seats in direct proportion to the minority group's percentage in the population provides the most logical ratio to apply as an "undiluted" norm. But § 2 makes it clear that the Act does not create a right to proportional representation, and thus dictates that proportionality should not provide the rule of decision for § 2 claims. See supra, at 927-928, and n. 26. Nevertheless, despite the statutory command, in deciding claims of vote dilution we have turned to proportionality as a guide, simply for lack of any better alternative.

No formulation of the test for evaluating vote dilution claims has ever dispensed with the inevitable need to consult a mathematical formula to decide a case. The factors listed in White v. Regester, 412 U. S., at 766-767, resurrected in the Senate Report on the 1982 amendments to § 2, see S. Rep. No. 97-417, at 28-29, and finally reincorporated into our decision in Gingles, see 478 U. S., at 44-45, although praised in our cases as a multifaceted test ensuring that vote dilution is determined based on the "totality of circumstances," in reality provide no rule for deciding a vote dilution claim based on anything other than a numerical principle.

29 I assume for purposes of the analysis here that the measure of effective votes is control of seats. That is precisely the measure the Court has applied, both in the past, see, e. g., Gingles, 478 U. S., at 46-51; id., at 93, 99 (O'CONNOR, J., concurring in judgment) (noting that the Court had made electoral results the "linchpin" of dilution claims), and today, see Johnson v. De Grandy, post, at 1014-1015 (equating "political effectiveness" with control of majority-minority districts).


938

THOMAS, J., concurring in judgment

In Gingles, we condensed the import of these "factors" into a formula stating that the "essence" of a vote dilution claim under § 2 is that "a certain electoral law, practice, or structure interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by black and white voters to elect their preferred representatives." Id., at 47. But it should be apparent that whether an electoral practice does or does not reduce the ability of a numerical minority to control the election of representatives can be determined wholly without reference to "social and historical conditions." Ibid. The dilutive effects of various electoral procedures are matters of mathematics. The "social and historical conditions" "interact" with the election mechanism, and thus are relevant in a vote dilution case, only to the extent that they are important for establishing that the minority group does in fact define a distinct political interest group that might assert that its vote has been diluted by the mechanism at issue. Such social and historical considerations, however, cannot supply the answer to the ultimate question whether the group's vote has been diluted.

In reality, the list of White factors provides nothing more than just that: a list of possible considerations that might be consulted by a court attempting to develop a gestalt view of the political and racial climate in a jurisdiction, but a list that cannot provide a rule for deciding a vote dilution claim. Take, for example, a case in which a district court determines that a minority group constituting 34% of the population in a certain jurisdiction has suffered discrimination in the past, that the group currently bears the effects of that discrimination, and that there has been a history of racial campaigning in the jurisdiction. Cf. White, supra, at 766-767. How can these facts possibly answer the question whether the group's votes have been diluted if the group controls two rather than three seats in a lO-member governing body? Will the answer to the ultimate question change if the first two factors are found, but the third is not? Obviously, the various "fac-


939

tors," singly or in any combination, cannot provide a principle for determining the result. What one must know to decide the case is whether 20% of the seats in the government is sufficient to reflect "undiluted" voting strength, or if 30% should be required.

Of course, as suggested above, the White factors may be relevant to determining as a threshold matter whether the minority group is a distinct political group that should be able to assert a claim of dilution. But after Gingles, the inquiry into whether race defines political interest effectively has been boiled down to the weakened test for minority "political cohesiveness" and majority bloc voting embodied in the second and third Gingles preconditions. See 478 U. S., at 51. Once a plaintiff group establishes that it is mathematically possible for it to control another seat (that is, that it satisfies the first Gingles precondition of size and geographic compactness), see id., at 50, and that it is a distinct political group (that is, that it can show political cohesion and majority bloc voting), the only question remaining in the vote dilution claim is whether the current number of seats is the proper number or not. The other White factors have become essentially superfluous. They may be dutifully intoned by courts performing the empty ritual of applying the "totality of circumstances" test, but they can provide no guidance concerning whether the current allocation of seats constitutes "dilution." Cf. Gingles, supra, at 92-93 (O'CONNOR, J., concurring in judgment) (suggesting that the basic contours of a dilution claim require no reference to most of the White factors).

In short, it should be clear that the factors listed in Gingles-in their various incarnations and by whatever names they are known-are nothing but puffery used to fill out an impressive verbal formulation and to create the impression that the outcome in a vote dilution case rests upon a reasoned evaluation of a variety of relevant circumstances. The "totality of circumstances" test outlined in Gingles thus


940

THOMAS, J., concurring in judgment

serves to obscure the inherent conflict between the text of the Act and an underlying reliance on proportionality.

The resort to proportionality in our cases should hardly come as a surprise. Before § 2 was amended in 1982, and thus before the Act explicitly disavowed a right to proportional representation, some Members of the Court recognized the inevitable drift toward proportional representation that would occur if the test outlined in White were used to evaluate vote dilution claims. As Justice Stewart, writing for four Members of the Court, observed, the factors listed in White amounted to little more than "gauzy sociological considerations," and it did not appear that "they could, in any principled manner, exclude the claims of any discrete political group that happens, for whatever reason, to elect fewer of its candidates than arithmetic indicates it might." Bolden, 446 U. S., at 75, n. 22 (emphasis added). Indeed, Justice Stewart was correct in concluding that "the putative limits [imposed by the White factors] are bound to prove illusory if the express purpose informing their application would be," as our vote dilution cases have assumed, "to redress the inequitable distribution of political influence." Ibid. (internal quotation marks omitted).

In fact, the framework established by this Court for evaluating vote dilution claims in Gingles was at its inception frankly, and in my view correctly, labeled as setting a rule of roughly proportional representation. See Gingles, supra, at 91, 93, 97-99 (O'CONNOR, J., concurring in judgment). Nothing has happened in the intervening years to change the basic import of the Gingles test. Yet we have continued to apply the same Gingles framework, see, e. g., Growe v. Emison, 507 U. S. 25 (1993), all the while suggesting that we are pursuing merely a "totality of the circumstances" test.

In another case decided today, the Court reconfirms the unstated centrality of proportional results in an opinion that demonstrates the obfuscation that must come to characterize our Voting Rights Act rulings if we continue to entertain


941

dilution claims while pretending to renounce reliance on proportionality as a rule of decision. In Johnson v. De Grandy, post, p. 997, the Court assures us that proportionality does not provide the principle for deciding vote dilution claims. Post, at 1000, 1017-1021. Rather, the result in each case must depend on a searching inquiry into the ever-nebulously defined "totality of circumstances." Post, at 1000.

But after the Gingles preconditions have been established, post, at 1008-1009, and after White factors such as a history of discrimination have been found, see post, at 1013, where does the Court turn for a deciding principle to give some meaning to these multifarious facts, which taken individually would each appear to count in favor of a finding of vote dilution? Quite simply, the Court turns to proportionality:

"Treating equal political opportunity as the focus of the enquiry, we do not see how these district lines, apparently providing political effectiveness [that is, majority-minority districts] in proportion to voting-age numbers, deny equal political opportunity." Post, at 1014. See also post, at 1013 (noting that in assessing "dilutive effect," the "pertinent features" of the districting system at issue "were majorityminority districts in substantial proportion to the minority's share of voting-age population"); post, at 1025 (O'CONNOR, J., concurring) (the Court's central teaching in De Grandy "is that proportionality-defined as the relationship between the number of majority-minority voting districts and the minority group's share of the relevant population-is always relevant evidence in determining vote dilution"). JUSTICE O'CONNOR'S comment about the Court's holding in Davis v. Bandemer, 478 U. S. 109 (1986), is equally applicable to the course pursued in De Grandy today: "[The Court's decision] ultimately rests on a political preference for proportionality- ... a conviction that the greater the departure from proportionality, the more suspect an apportionment plan becomes." 478 U. S., at 159 (opinion concurring in judgment).


942

THOMAS, J., concurring in judgment

To be sure, the De Grandy Court repeatedly declares that proportionality is not a defense to a vote dilution claim. See post, at 1017-1021. That, of course, must be the stated rule if we are not to abandon openly the explicit disclaimer enacted by Congress in § 2(b). But given the Court's equivocationproportionality is still always relevant-and the Court's ultimate analysis, such assurances ring hollow. The Court decides the question of dilution based upon proportionality. And it is apparent from the reasons the Court gives for rejecting maximization as a rule for decision that proportionality will drive results in future dilution cases as well.

Consider, for example, the hypothetical rehearsed by the Court concerning a jurisdiction with a 10-member elected body and a 40% minority voting population. See post, at 1016-1017. Assume that as currently constituted the districting scheme creates four majority-minority districts. Even if it is established in this hypothetical jurisdiction that all of the Gingles factors have been proved (as was found in De Grandy), and that there are both a history of discrimination and continuing discrimination (as was found in De Grandy), can it be seriously contended that the minority group can succeed, under any combination of facts, in bringing a § 2 challenge to require the creation of the mathematically possible seven majority-minority districts? The Court recognizes that it would be "absurd" to think that § 2 would allow such a result. That, after all, would give the group "effective political power 75 percent above its numerical strength"-that is, above its proportion in the population. Post, at 1017 (emphasis added). But if it is absurd to give the members of the group seven seats, why is it not equally ridiculous to give them six, or five? Or, indeed, anything beyond the four that would secure them seats in proportion to their numbers in the population?

If it is absurd to give members of the group seven seats, that is because, as the Court tacitly acknowledges, we assume that seats in accord with "numerical strength" will


943

ensure the group "equal" "political effectiveness." Thus, deliberately drawing districts so as to give, under the assumptions of the hypothetical, 40% of the population control over 50% of the seats, while leaving 60% of the population with control of a similar 50% of the seats, would seem to us unfair. Greater deviations from proportionality may appear more patently "absurd" than lesser, but the dividing line between what seems fair and what does not remains the same. The driving principle is proportionality.30

Few words would be too strong to describe the dissembling that pervades the application of the "totality of circumstances" test under our interpretation of § 2. It is an empty incantation-a mere conjurer's trick that serves to hide the

30 Of course, throughout this discussion concerning the Court's inevitable resort to proportionality, I have assumed that effective votes will be measured in terms of control of seats. See n. 29, supra. As JUSTICE O'CONNOR suggests in her opinion in De Grandy, if we were to measure the effectiveness of votes not simply in terms of numbers of seats, but in terms of some more amorphous concept of "access to the political process," there would be no need to make proportionality "dispositive." See De Grandy, post, at 1026 (O'CONNOR, J., concurring). Cf. White, 412 U. S., at 765-766. But Gingles made control of seats the determining factor in dilution claims; that is the measure that has been applied in cases under Gingles, and it remains the measure applied in practice in the cases handed down today. In my view, it is unrealistic to think that the Court will now reverse course and establish some broader understanding of "political effectiveness" under the "totality of circumstances" test, after it consistently has pursued a measure of effective voting that makes electoral results the "linchpin" of dilution claims. See 478 U. S., at 93 (O'CONNOR, J., concurring in judgment).

Indeed, any change in course is made more unlikely by one very practical consideration. As the Court's decision in De Grandy perhaps suggests, measuring political effectiveness by any method other than counting numbers of seats can rapidly become a wholly unmanageable task. As I suggested above, see n. 6, supra, one of the reasons the Court seized upon control of seats as a measure of effective political participation is simply that control of seats provides the "most easily measured indicia of political power." Bandemer, 478 U. S., at 157 (O'CONNOR, J., concurring in judgment).


944

THOMAS, J., concurring in judgment

drive for proportionality that animates our decisions. As actions such as that brought in Shaw v. Reno, 509 U. S. 630 (1993), have already started to show, what might euphemistically be termed the benign "creation of majority-minority single-member districts to enhance the opportunity of minority groups to elect representatives of their choice" might also more simply and more truthfully be termed "racial gerrymandering." Similarly, what we might call a "totality of circumstances" test to determine whether an electoral practice "interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by black and white voters to elect their preferred representatives," Gingles, 478 U. S., at 47, might more accurately be called a test for ensuring proportional electoral results according to race. Cf. id., at 97 (O'CONNOR, J., concurring in judgment).

In my view, our current practice should not continue. Not for another Term, not until the next case, not for another day. The disastrous implications of the policies we have adopted under the Act are too grave; the dissembling in our approach to the Act too damaging to the credibility of the Federal Judiciary. The "inherent tension"-indeed, I would call it an irreconcilable conflict-between the standards we have adopted for evaluating vote dilution claims and the text of the Voting Rights Act would itself be sufficient in my view to warrant overruling the interpretation of § 2 set out in Gingles. When that obvious conflict is combined with the destructive effects our expansive reading of the Act has had in involving the Federal Judiciary in the project of dividing the Nation into racially segregated electoral districts, I can see no reasonable alternative to abandoning our current unfortunate understanding of the Act.

Stare decisis is a powerful concern, especially in the field of statutory construction. See Patterson v. McLean Credit Union, 491 U. S. 164, 172 (1989). See also Fogerty v. Fantasy, Inc., 510 U. S. 517, 538-539 (1994) (THOMAS, J., concurring in judgment). But "we have never applied stare decisis


945

mechanically to prohibit overruling our earlier decisions determining the meaning of statutes." Monell v. New York City Dept. of Social Servs., 436 U. S. 658, 695 (1978). Stare decisis should not bind the Court to an interpretation of the Voting Rights Act that was based on a flawed method of statutory construction from its inception and that in every day of its continued existence involves the Federal Judiciary in attempts to obscure the conflict between our cases and the explicit commands of the Act. The Court has noted in the past that stare decisis" 'is a principle of policy,'" Payne, 501 U. S., at 828 (quoting Helvering v. Hallock, 309 U. S. 106, 119 (1940)), and it "'is usually the wise policy, because in most matters it is more important that the applicable rule of law be settled than it be settled right.''' 501 U. S., at 827 (quoting Burnet v. Coronado Oil & Gas Co., 285 U. S. 393, 406 (1932) (Brandeis, J., dissenting)). I cannot subscribe to the view that in our decisions under the Voting Rights Act it is more important that we have a settled rule than that we have the right rule. When, under our direction, federal courts are engaged in methodically carving the country into racially designated electoral districts, it is imperative that we stop to reconsider whether the course we have charted for the Nation is the one set by the people through their representatives in Congress. I believe it is not.

I cannot adhere to the construction of § 2 embodied in our decision in Thornburg v. Gingles. I reject the assumption implicit in that case that the terms "standard, practice, or procedure" in § 2(a) of the Voting Rights Act can be construed to cover potentially dilutive electoral mechanisms. Understood in context, those terms extend the Act's prohibitions only to state enactments that regulate citizens' access to the ballot or the processes for counting a ballot. The terms do not include a State's or political subdivision's choice of one districting scheme over another. The terms certainly do not include, as respondents would argue, the size of a local governing authority.


946

III

For the foregoing reasons, I agree with the Court's conclusion that the size of a governing body is not subject to challenge under § 2 of the Voting Rights Act. I therefore concur in the Court's judgment reversing the judgment below and remanding for consideration of respondents' constitutional claim of intentional discrimination.

JUSTICE BLACKMUN, with whom JUSTICE STEVENS, JusTICE SOUTER, and JUSTICE GINSBURG join, dissenting.

Five Justices today agree that the size of a governing body is a "standard, practice, or procedure" under § 2 of the Voting Rights Act of 1965 (Act), as amended, 42 U. S. C. § 1973. A different five Justices decide, under three separate theories, that voting rights plaintiffs cannot bring § 2 dilution challenges based on size. I, however, believe that the Act, its history, and our own precedent require us to conclude not only that the size of a governing body is a "standard, practice, or procedure" under § 2, but also that minority voters may challenge the dilutive effects of this practice by demonstrating their potential to elect representatives under an objectively reasonable alternative practice. Accordingly, I dissent from the Court's decision that minority voters cannot bring § 2 vote dilution challenges based on the size of an existing government body.

I

Section 2(a) of the Act prohibits the imposition or application of any "voting qualification or prerequisite to voting, or standard, practice, or procedure" that "results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color." 42 U. S. C. § 1973(a) (emphasis added). Section 5 parallels § 2 by requiring certain jurisdictions to preclear with the Attorney General a change in "any voting qualification or prerequisite to voting, or standard, practice, or procedure with respect to voting." 42 U. S. C. § 1973c (emphasis added). Under the broad inter-


947

pretation that this Court, Congress, and the Attorney General consistently have given the Act in general and § 5 in particular, the practice of electing a single commissioner, as opposed to a multimember commission, constitutes a "standard, practice, or procedure" under § 2.

Nearly 30 years of precedent admonish us that the Act, which was adopted "for the broad remedial purpose of 'rid[ding] the country of racial discrimination in voting,'" Chisom v. Roemer, 501 U. S. 380, 403 (1991), quoting South Carolina v. Katzenbach, 383 U. S. 301, 315 (1966), should be given "the broadest possible scope," Allen v. State Bd. of Elections, 393 U. S. 544, 567 (1969). Because "the Act itself nowhere amplifies the meaning of the phrase 'standard, practice, or procedure with respect to voting,'" the Court "ha[s] sought guidance from the history and purpose of the Act." Dougherty County Bd. of Ed. v. White, 439 U. S. 32, 37 (1978); see also McCain v. Lybrand, 465 U. S. 236, 246 (1984) (the Act must "be interpreted in light of its prophylactic purpose and the historical experience which it reflects").

Consistent with the Act's remedial purposes, this Court has held that a wide variety of election- and voting-related practices fit within the term "standard, practice, or procedure." Among the covered practices are the annexation of land to enlarge city boundaries, see Perkins v. Matthews, 400 U. S. 379, 388 (1971), and Pleasant Grove v. United States, 479 U. S. 462, 467 (1987); a rule requiring employees to take leaves of absence while they campaign for elective office, see Dougherty County Bd. of Ed., 439 U. S., at 34; candidate filing dates and other procedural requirements, see Whitley v. Williams, decided with Allen v. State Bd. of Elections, supra; Hadnott v. Amos, 394 U. S. 358, 365 (1969); NAACP v. Hampton County Election Comm'n, 470 U. S. 166, 176-177 (1985); and candidate residency requirements, see City of Rome v. United States, 446 U. S. 156, 160 (1980).

Specifically, this Court long has treated a change in the size of a governing authority as a change in a "standard,


948

practice, or procedure" with respect to voting. In City of Rome, 446 U. S., at 161, it noted that it "is not disputed" that an expansion in the size of a board of education was "within the purview of the Act" and subject to preclearance under § 5. In City of Lockhart v. United States, 460 U. S. 125, 131 (1983), it stated that a change from a three-member commission to a five-member commission was subject to § 5 preclearance. And, most recently, it said that the term "standard, practice, or procedure with respect to voting" included a change in the size of a governing authority or an increase or decrease in the number of elected offices. Presley v. Etowah County Comm'n, 502 U. S. 491, 500 (1992).

This conclusion flowed naturally from the holding in Bunton v. Patterson, 393 U. S. 544 (1969), that a change from an elected to an appointed office was a "standard, practice, or procedure with respect to voting." In Bunton, the Court reasoned that the power of a citizen's vote is affected by the change because the citizen has been "prohibited from electing an officer formerly subject to the approval of the voters." Id., at 570. The reverse is also true: A change from an appointed to an elected office affects a citizen's voting power by increasing the number of officials for whom he may vote. See McCain v. Lybrand, 465 U. S. 236 (1984). And, as the Court recognized in Presley, a change in the size of a governing authority is a "standard, practice, or procedure with respect to voting" because the change "increase[s] or diminish[es] the number of officials for whom the electorate may vote," 502 U. S., at 503; this change bears "on the substance of voting power" and has "a direct relation to voting and the election process." Ibid.

To date, our precedent has dealt with § 5 challenges to a change in the size of a governing authority, rather than § 2 challenges to the existing size of a governing body. I agree with JUSTICE O'CONNOR, ante, at 886-887, that, as a textual matter, "standard, practice, or procedure" under § 2 is at


949

least as broad as "standard, practice, or procedure with respect to voting" under § 5. In fact, because of the "close connection" between §§ 2 and 5, we interpret them similarly. See Chisom v. Roemer, 501 U. S., at 402 (concluding that it would be "anomalous" to do otherwise). And in the context of § 2, the Court stated: "Section 2 protected the right to vote, and it did so without making any distinctions or imposing any limitations as to which elections would fall within its purview." Id., at 392. See also Houston Lawyers' Assn. v. Attorney General of Tex., 501 U. S. 419 (1991) (rejecting a "single-member-office" exception to § 2).

Congress repeatedly has endorsed the broad construction this Court has given the Act in general and § 5 in particular.1 Significantly, when Congress considered the 1982 amendments to the Voting Rights Act, it made no effort to curtail the application of § 5 to changes in size, in the face of the longstanding practice of submitting such changes for preclearance, and on the heels of this Court's recognition just two years earlier that it was "not disputed" that a change in the size of a governing body was covered under § 5. See City of Rome, 446 U. S., at 161. Similarly, the Attorney General, whose construction of the Act "is entitled to considerable deference," NAACP v. Hampton County Election Comm'n, 470 U. S., at 178-179, for years has required §5 preclearance of the expansion or reduction of a governing

1 See Georgia v. United States, 411 U. S. 526, 533 (1973) ("After extensive deliberations in 1970 on bills to extend the Voting Rights Act, during which the Allen case was repeatedly discussed, the Act was extended for five years, without any substantive modification of § 5") (footnote omitted); Dougherty County Ed. of Ed. v. White, 439 U. S. 32, 39 (1978) ("Again in 1975, both the House and Senate Judiciary Committees, in recommending extension of the Act, noted with approval the 'broad interpretations to the scope of Section 5' in Allen and Perkins v. Matthews [400 U. S. 379 (1971)]"); NAACP v. Hampton County Election Comm'n, 470 U. S. 166, 176 (1985) (in the 1982 extension of the Act, "Congress specifically endorsed a broad construction" of § 5).


950

body.2 It is not surprising that no party to this case argued that the size of a governing authority is not a "standard, practice, or procedure."

In light of this consistent and expansive interpretation of the Act by this Court, Congress, and the Attorney General, the Act's "all-inclusive" definition of "standard, practice, or procedure" cannot be read to exclude threshold coverage of challenges to the size of a governing authority. As five Members of the Court today agree, the size of a governing authority is a "standard, practice, or procedure" with respect to voting for purposes of § 2 as well as § 5 of the Voting Rights Act.

II

Although five Justices agree that the size of a governing body is a "standard, practice, or procedure" under § 2, a like number of Justices conclude, under varying rationales, that Voting Rights plaintiffs nonetheless cannot bring size challenges under § 2. This conclusion is inconsistent with our precedent giving the Act "'the broadest possible scope' in combating racial discrimination," Chisom, 501 U. S., at 403, quoting Allen, 393 U. S., at 567, and with the vote-dilution

2 See Hearings on S. 1992 before the Subcommittee on the Constitution of the Senate Committee on the Judiciary, 97th Cong., 2d Sess., 1748 (1982) (noting Attorney General's objection in 1971 to proposed reduction in the size of a school board); id., at 1751 (1971 objection to expansion of a parish council); id., at 1782 (1980 objection to decrease in number of city council members); id., at 1384-1385 (the Voting Rights Act afforded protection against "[s]hifts from ward to at-large elections, from plurality win to majority vote, from slating to numbered posts, annexations and changes in the size of electoral bodies," that "could ... deprive minority voters of fair and effective procedures for electing candidates of their choice") (statement of Drew S. Days III, former Assistant Attorney General for Civil Rights) (emphasis added).

Since covered jurisdictions routinely have submitted changes in the size of their legislative bodies for preclearance, it is not surprising that petitioners concede that a change in the size of the Bleckley County Commission would be subject to § 5 preclearance. Tr. of Oral Arg. 4, 13.


951

analysis prescribed in Thornburg v. Gingles, 478 U. S. 30 (1986).

To prevail in a vote-dilution challenge, minority voters must show that they "possess the potential to elect representatives in the absence of the challenged structure or practice." Id., at 50, n. 17 (second emphasis supplied).3 There is widespread agreement, see ante, at 880 (opinion of KENNEDY, J., and REHNQUIST, C. J.); ante, at 887 (opinion of O'CONNOR, J.), that minority voters' potential "in the absence of" the allegedly dilutive mechanism must be measured against the benchmark of an alternative structure or practice that is reasonable and workable under the facts of the specific case.4

By all objective measures, the proposed five-member Bleckley County Commission presents a reasonable, workable benchmark against which to measure the practice of electing a sole commissioner. First, the Georgia Legislature specifically authorized a five-member commission for Bleckley County. 1985 Ga. Laws 4406. Moreover, a fivemember commission is the most common form of governing authority in Georgia. See Georgia Dept. of Community Af-

3 Although Gingles dealt with the use of multimember districts, the analysis it prescribes is applicable in certain other vote-dilution contexts, such as a claim of "vote fragmentation" through single-member districts, see Growe v. Emison, 507 U. S. 25, 37-42 (1993), or the case before us.

4 As the United States explains, the minority group must be permitted to establish that, under "a proposed alternative voting arrangement that is reasonable in the legal and factual context of a particular case," it could constitute a majority. Brief for United States as Amicus Curiae 8. The Court of Appeals followed this approach, concluding that "it is appropriate to consider the size and geographical compactness of the minority group within a restructured form of the challenged system when the existing structure is being challenged as dilutive" (emphasis in original). 955 F.2d 1563, 1569 (CAll 1992). See also Carrollton Branch of NAACP v. Stallings, 829 F.2d 1547 (CAll 1987) (remand of challenge to solecommissioner system with instructions to consider size and geographic compactness within proposed three- and five-member commission forms of government).


952

fairs, County Government Information Catalog (1989) (Table 1.A: Form of Government) (76 of Georgia's 159 counties had five commissioners, including 25 counties smaller than Bleckley County). Bleckley County, as one of a small and dwindling number of counties in Georgia still employing a sole commissioner, markedly departs from practices elsewhere in Georgia. This marked "depart[ure] ... from practices elsewhere in the jurisdiction ... bears on the fairness of [the sole commissioner's] impact." S. Rep. No. 97-417, p. 29, n. 117 (1982). Finally, the county itself has moved from a single superintendent of education to a school board with five members elected from single-member districts, providing a workable and readily available model for commission districts. Thus, the proposed five-member baseline is reasonable and workable.

In this case, identifying an appropriate baseline against which to measure dilution is not difficult. In other cases, it may be harder. But the need to make difficult judgments does not "justify a judicially created limitation on the coverage of the broadly worded statute, as enacted and amended by Congress." Chisom, 501 U. S., at 403. Vote dilution is inherently a relative concept, requiring a highly "flexible, fact-intensive" inquiry, Gingles, 478 U. S., at 46, and calling for an exercise of the "court's overall judgment, based on the totality of circumstances and guided by those relevant factors in the particular case," as mandated by Congress. S. Rep. No. 97-417, at 29, n. 118. Certainly judges who engage in the complex task of evaluating reapportionment plans and examining district lines will be able to determine whether a proposed baseline is an appropriate one against which to measure a claim of vote dilution based on the size of a county commission.

There are, to be sure, significant constraints on size challenges. Minority plaintiffs, who bear the burden of demonstrating dilution, also bear the burden of demonstrating that


953

their proposed benchmark is reasonable and workable. One indication of a benchmark's reasonableness is its grounding in history, custom, or practice. This consideration will discourage size challenges to traditional single-member executive offices, such as governors and mayors, or even sheriffs or clerks of court. By tradition and practice, these executive positions are occupied by one person, so plaintiffs could rarely point to an objectively reasonable alternative size that has any foundation in the past or present. Cf. The Federalist No. 69, p. 415 (C. Rossiter ed. 1961) (A. Hamilton) ("[T]he executive authority, with few exceptions, is to be vested in a single magistrate"). The sole commissioner, by contrast, holds plenary legislative, as well as executive, power. Ga. Code Ann. § 36-5-22.1 (1993). A one-member legislature, far from being the norm, is an anomaly. Accordingly, the Eleventh Circuit, while permitting § 2 challenges to the practice of electing a sole commissioner, has held that this provision cannot be used to alter the practice of electing a single person to offices such as lieutenant governor, sheriff, probate judge, and tax collector. See Dillard v. Crenshaw County, 831 F.2d 246, 251 (1987); United States v. Dallas County Comm'n, 850 F.2d 1430, 1432, n. 1 (1988), cert. denied, 490 U. S. 1030 (1989).5

Additionally, every successful vote-dilution challenge will be based on the "totality of the circumstances," often including the lingering effects of past discrimination. S. Rep. No. 97-417, at 28-30. Not every racial or language minority that constitutes 5% of the population has a claim to have a governing authority expanded to 20 members in order to give them an opportunity to elect a representative. Instead, the voters would have to prove that a 20-

5 Of course, this is not to suggest that single-member executive offices are not within the scope of §2, see Houston Lawyers' Assn. v. Attorney General of Tex., 501 U. S. 419, 425-428 (1991), but only that they are not generally susceptible to size challenges under § 2.


954

member governing authority was a reasonable benchmarkwhich, of course, respondents could not do here-and that their claim satisfied the three Gingles preconditions, 478 U. S., at 49, and was warranted under the totality of the circumstances.6

6 The Senate Report accompanying the 1982 amendments to the Act directed that the vote-dilution inquiry include an examination of the factors identified in White v. Regester, 412 U. S. 755 (1973), and refined and developed in Zimmer v. McKeithen, 485 F.2d 1297 (CA5 1973) (en banc), aff'd, 424 U. S. 636 (1976) (per curiam). This nonexclusive list of factors, now known variously as the Regester-Zimmer factors or "Senate Report factors," includes "the extent of any history of official discrimination ... that touched the right of the members of the minority group to register, to vote, or otherwise to participate in the democratic process; ... the extent to which the state or political subdivision has used unusually large election districts, majority vote requirements, anti-single shot provisions, or other voting practices or procedures that may enhance the opportunity for discrimination against the minority group; ... [and] the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as education, employment and health, which hinder their ability to participate effectively in the political process." S. Rep. No. 97-417, pp. 28-29 (1982).

In this case, for example, the District Court found that, until the passage of federal civil rights laws, Bleckley County "enforced racial segregation in all aspects of local government-courthouse, jails, public housing, governmental services-and deprived its black citizens of the opportunity to participate in local government." Hall v. Holder, 757 F. Supp. 1560, 1562 (MD Ga. 1991). Until the passage of the Voting Rights Act of 1965, "black citizens were virtually prohibited from registering to vote in Bleckley County." Id., at 1563. Until 1984, there were no African-American voting registrars and no voter registration in places where MricanAmericans normally congregated. Ibid. From 1978 until 1986, the respondent probate judge appointed 224 poll managers, all white, and 509 poll clerks, 479 of whom were white. Ibid. Since 1964, the election of Bleckley County's sole commissioner has been subject to a majority-vote requirement. Although official segregation is no longer imposed, its vestiges remain, as "more black than white residents of Bleckley County continue to endure a depressed socio-economic status," id., at 1562, which "hinders the ability of and deters black residents of Bleckley County from running for public office, voting and otherwise participating in the political


955

With these limitations, successful vote-dilution challenges to the size of a governing authority always will be based not on abstract manipulation of numbers, but on a "searching practical evaluation of the 'past and present reality.'" S. Rep. No. 97-417, at 30, quoting White v. Regester, 412 U. S. 755, 770 (1973). These limitations protect against a proliferation of vote-dilution challenges premised on eccentric or impracticable alternative methods of redistricting.

III

The Voting Rights Act of 1965 was bold and ambitious legislation, designed to eradicate the vestiges of past discrimination and to make members of racial and language minorities full participants in American political life. Nearly 30 years after the passage of this landmark civil rights legislation, its goals remain unfulfilled. Today, the most blatant forms of discrimination-including poll taxes, literacy tests, and "white" primaries-have been eliminated. But subtler, more complex means of infringing minority voting strength-including submergence or dispersion of minority voters-are still present and indeed prevalent. We have recognized over the years that seemingly innocuous and even well-intentioned election practices may impede minority voters' ability not only to vote, but to have their votes count. It is clear that the practice of electing a single-member county commission can be one such dilutive practice. It is equally clear that a five-member commission is an appropriate benchmark against which to measure the alleged dilutive effects of Bleckley County's practice of electing a sole commissioner. I respectfully dissent.

process," id., at 1563. The "barriers to active participation in the political process are ... compounded by the fact that Bleckley County now has only one voting precinct for the entire 219 square-mile area." Id., at 1563, n. 3. That single polling place is located at an all-white civic club. 955 F.2d 1563, 1566 (CAll 1992).


956

JUSTICE GINSBURG, dissenting.

I join the dissenting opinion by JUSTICE BLACKMUN and the separate opinion of JUSTICE STEVENS, and add a further observation about the responsibility Congress has given to the judiciary.

Section 2 of the Voting Rights Act of 1965 calls for an inquiry into "[t]he extent to which members of a protected class have been elected to office," but simultaneously disclaims any "right to have members of a protected class elected in numbers equal to their proportion in the population." 42 U. S. C. § 1973(b). "There is an inherent tension between what Congress wished to do and what it wished to avoid"-between Congress' "inten[t] to allow vote dilution claims to be brought under § 2" and its intent to avoid "creat[ing] a right to proportional representation for minority voters." Thornburg v. Gingles, 478 U. S. 30, 84 (1986) (O'CONNOR, J., joined by Burger, C. J., and Powell and REHNQUIST, JJ., concurring in judgment). Tension of this kind is hardly unique to the Voting Rights Act, for when Congress acts on issues on which its constituents are divided, sometimes bitterly, the give-and-take of legislative compromise can yield statutory language that fails to reconcile conflicting goals and purposes.

Title VII of the Civil Rights Act of 1964, for example, is similarly janus faced, prohibiting discrimination against historically disadvantaged groups, see 42 U. S. C. §§ 2000e-2(a), (d), without "diminish[ing] traditional management prerogatives," Steelworkers v. Weber, 443 U. S. 193, 207 (1979), in regard to employment decisions. See 42 U. S. C. § 2000e2(j) (no requirement that employer "grant preferential treatment to any individual or to any group because of ... race, color, religion, sex, or national origin"); see also Johnson v. Transportation Agency, Santa Clara Cty., 480 U. S. 616, 649 (1987) (O'CONNOR, J., concurring in judgment) (noting two "conflicting concerns" built into Title VII: "Congress' intent


957

to root out invidious discrimination against any person on the basis of race or gender, and its goal of eliminating the lasting effects of discrimination against minorities") (emphasis in original) (citation omitted).

When courts are confronted with congressionally crafted compromises of this kind, it is "not an easy task" to remain "faithful to the balance Congress struck." Thornburg v. Gingles, 478 U. S., at 84 (O'CONNOR, J., joined by Burger, C. J., and Powell and REHNQUIST, JJ., concurring in judgment). The statute's broad remedial purposes, as well as the constraints on the courts' remedial powers, need to be carefully considered in light of the particular circumstances of each case to arrive at an appropriate resolution of the competing congressional concerns. However difficult this task may prove to be, it is one that courts must undertake because it is their mission to effectuate Congress' multiple purposes as best they can. See Chisom v. Roemer, 501 U. S. 380, 403 (1991) ("Even if serious problems lie ahead in applying the 'totality of the circumstances' [inquiry under § 2(b) of the Voting Rights Act], that task, difficult as it may prove to be, cannot justify a judicially created limitation on the coverage of the broadly worded statute[.]").

Separate opinion of JUSTICE STEVENS, in which JUSTICE BLACKMUN, JUSTICE SOUTER, and JUSTICE GINSBURG join.

JUSTICE THOMAS has written a separate opinion proposing that the terms "standard, practice, or procedure" as used in the Voting Rights Act of 1965 should henceforth be construed to refer only to practices that affect minority citizens' access to the ballot. Specifically, JUSTICE THOMAS would no longer interpret the Act to forbid practices that dilute minority voting strength. To the extent that his opinion advances policy arguments in favor of that interpretation of the statute, it should be addressed to Congress, which has ample power to amend the statute. To the extent that the opinion


958

Opinion of STEVENS, J.

suggests that federal judges have an obligation to subscribe to the proposed narrow reading of statutory language, it is appropriate to supplement JUSTICE THOMAS' writing with a few words of history.

I

JUSTICE THOMAS notes that the first generation of Voting Rights Act cases focused on access to the ballot. Ante, at 894-895. By doing so, he suggests that the early pattern of enforcement is an indication of the original meaning of the statute. In this regard, it is important to note that the Court's first case addressing a voting practice other than access to the ballot arose under the Fifteenth Amendment. In Gomillion v. Lightfoot, 364 U. S. 339 (1960), the Court held that a change in the boundaries of the city of Tuskegee, Alabama, violated the Fifteenth Amendment. In his opinion for the Court, Justice Frankfurter wrote:

"The opposite conclusion, urged upon us by respondents, would sanction the achievement by a State of any impairment of voting rights whatever so long as it was cloaked in the garb of the realignment of political subdivisions." Id., at 345.

"A statute which is alleged to have worked unconstitutional deprivations of petitioners' rights is not immune to attack simply because the mechanism employed by the legislature is a redefinition of municipal boundaries. According to the allegations here made, the Alabama Legislature has not merely redrawn the Tuskegee city limits with incidental inconvenience to the petitioners; it is more accurate to say that it has deprived the petitioners of the municipal franchise and consequent rights and to that end it has incidentally changed the city's boundaries. While in form this is merely an act redefining metes and bounds, if the allegations are established, the inescapable human effect of this essay in geometry and geography is to despoil colored citizens,


959

and only colored citizens, of their theretofore enjoyed voting rights." Id., at 347.1

Because Gomillion was decided only a few years before the Voting Rights Act of 1965 was passed, and because coverage under the Voting Rights Act is generally coextensive with or broader than coverage under the Fifteenth Amendment, see Katzenbach v. Morgan, 384 U. S. 641 (1966); Mobile v. Bolden, 446 U. S. 55, 60-61 (1980) (plurality opinion), it is surely not unreasonable to infer that Congress intended the Act to reach the kind of voting practice that was at issue in that case. Nevertheless, the text of the Act would also have supported the opposite inference, because the language of the Fifteenth Amendment would seem to forbid any denial or abridgment of the right to vote, whereas §§ 2 and 5 of the Voting Rights Act refer only to "voting qualification[s,] ... prerequisite[s] to voting, ... standard [s], practice[s], [and] procedure[s]."

During the years between 1965 and 1969 the question whether the Voting Rights Act should be narrowly construed to cover nothing more than impediments to access to the ballot was an unresolved issue. What JUSTICE THOMAS describes as "a fundamental shift in the focal point of the Act," ante, at 895, occurred in 1969 when the Court unequivocally rejected the narrow reading, relying heavily on a broad

1 In most of his opinion, JUSTICE THOMAS seems to use the phrase "access to the ballot" to refer to the voter's ability to cast a vote. In an attempt to characterize the Gomillion gerrymander as a practice that interfered with access to the ballot, however, he seems to take the position that the redrawing of the boundaries of a governmental unit is a practice that affects access to the ballot because some voters' ballots could not thereafter be cast for the same offices as before. See ante, at 920, n. 20. Under such reasoning the substitution of an appointive office for an elective office, see Bunton v. Patterson, decided with Allen v. State Bd. of Elections, 393 U. S. 544, 550-551 (1969), or a change in district boundaries that prevented voters from casting ballots for the reelection of their incumbent congressional Representatives, would also be covered practices.


960

Opinion of STEVENS, J.

definition of the term "voting" as including" 'all action necessary to make a vote effective.'" Allen v. State Bd. of Elections, 393 U. S. 544, 565-566.

Despite Allen's purported deviation from the Act's true meaning, Congress one year later reenacted § 5 without in any way changing the operative words. During the next five years, the Court consistently adhered to Allen, see Perkins v. Matthews, 400 U. S. 379 (1971); Georgia v. United States, 411 U. S. 526 (1973), and in 1975, Congress again reenacted § 5 without change.

When, in the late seventies, some parties advocated a narrow reading of the Act, the Court pointed to these congressional reenactments as solid evidence that Allen, even if not correctly decided in 1969, would now be clearly correct. In United States v. Sheffield Bd. of Comm'rs, 435 U. S. 110, 132133 (1978), the Court noted:

"In 1970, Congress was clearly fully aware of this Court's interpretation of § 5 as reaching voter changes other than those affecting the registration process and plainly contemplated that the Act would continue to be so construed. See, e. g., Hearings on H. R. 4249 et al. before Subcommittee No.5 of the House Committee on the Judiciary, 91st Cong., 1st Sess., 1, 4, 18,83, 130-131, 133,147-149,154-155,182-184,402-454(1969); Hearings on S. 818 et al. before the Subcommittee on Constitutional Rights of the Senate Committee on the Judiciary, 91st Cong., 1st and 2d Sess., 48, 195-196, 369-370, 397398, 426-427, 469 (1970) ....

"The congressional history is even clearer with respect to the 1975 extension .... " 2

2 See also United Jewish Organizations of Williamsburgh, Inc. v. Carey, 430 U. S. 144, 157-159 (1977) (opinion of White, J.): "In Allen v. State Board of Elections[, 393 U. S. 544 (1969),] ... we held that a change from district to at-large voting for county supervisors had to be submitted for federal approval under § 5, because of the potential for a 'dilution' of minority voting power which could 'nullify [its] ability to elect the candidate of [its]


961

As the Court in that case also noted, when Congress reenacts a statute with knowledge of its prior interpretation, that interpretation is binding on the Court.

"Whatever one might think of the other arguments advanced, the legislative background of the 1975 reenactment is conclusive of the question before us. When a Congress that re-enacts a statute voices its approval of an administrative or other interpretation thereof, Congress is treated as having adopted that interpretation, and this Court is bound thereby. See, e. g., Don E. Williams Co. v. Commissioner, 429 U. S. 569,576-577 (1977); Albemarle Paper Co. v. Moody, 422 U. S. 405, 414 n. 8 (1975); H. Hart & A. Sacks, The Legal Process: Basic Problems in the Making and Application of Law 1404 (tent. ed. 1958); cf. Zenith Radio Corp. v. Hazeltine Research, 401 U. S. 321, 336 n. 7 (1971); Girouard v. United States, 328 U. S. 61, 69-70 (1946). Don E. Williams Co. v. Commissioner, supra, is instructive. As here, there had been a longstanding administrative interpretation of a statute when Congress re-enacted it, and there, as here, the legislative history of the reenactment showed that Congress agreed with that inter-

choice ... .' 393 U. S., at 569. When it renewed the Voting Rights Act in 1970 and again in 1975, Congress was well aware of the application of § 5 to redistricting. In its 1970 extension, Congress relied on findings by the United States Commission on Civil Rights that the newly gained voting strength of minorities was in danger of being diluted by redistricting plans that divided minority communities among predominantly white districts. In 1975, Congress was unmistakably cognizant of this new phase in the effort to eliminate voting discrimination. Former Attorney General Katzenbach testified that § 5 'has had its broadest impact ... in the areas of redistricting and reapportionment,' and the Senate and House Reports recommending the extension of the Act referred specifically to the Attorney General's role in screening redistricting plans to protect the opportunities for nonwhites to be elected to public office" (footnote omitted).


962

Opinion of STEVENS, J.

pretation, leading this Court to conclude that Congress had ratified it. 429 U. S., at 574-577." Id., at 134-135.

If the 1970 and 1975 reenactments had left any doubt as to congressional intent, that doubt would be set aside by the 1982 amendments to § 2. Between 1975 and 1982, the Court continued to interpret the Voting Rights Act in the broad manner set out by Allen. See City of Rome v. United States, 446 U. S. 156 (1980); Dougherty County Bd. of Ed. v. White, 439 U. S. 32 (1978); United Jewish Organizations of Williamsburgh, Inc. v. Carey, 430 U. S. 144 (1977); Richmond v. United States, 422 U. S. 358 (1975). In Mobile v. Bolden, 446 U. S. 55 (1980), a plurality of this Court concluded that violations of both the Voting Rights Act and the Fifteenth Amendment required discriminatory purpose. The case involved a claim that at-large voting diluted minority voting strength. In his opinion for the plurality in Bolden, Justice Stewart expressly relied upon Gomillion v. Lightfoot's holding "that allegations of a racially motivated gerrymander of municipal boundaries stated a claim under the Fifteenth Amendment." 446 U. S., at 62; see also id., at 85-86 (STEVENS, J., concurring in judgment). The only reason Gomillion did not control the outcome in Bolden was that an "invidious purpose" had been alleged in the earlier case but not in Bolden. 446 U. S., at 63.3 The congressional response to Bolden is familiar history. In the 1982 amendment to § 2 of the Voting Rights Act, Congress substituted a "results" test for an intent requirement. Pub. L. 97-205, § 3, 96 Stat. 134; see 42 U. S. C. § 1973. It is crystal

3 The idea that the Court in Bolden cast doubt on whether the Voting Rights Act reached diluting practices is flatly refuted by another decision handed down the very same day as the Bolden decision. In City of Rome v. United States, 446 U. S. 156, 186-187 (1980), the Court held that § 5 required preclearance of annexations potentially diluting minority voting strength. Even the dissenters did not suggest that vote dilution claims were now questionable.


963

clear that Congress intended the 1982 amendment to cover nonaccess claims like those in Bolden and Gomillion.4

II

JUSTICE THOMAS' narrow interpretation of the words "voting qualification ... standard, practice, or procedure," if adopted, would require us to overrule Allen and the cases that have adhered to its reading of the critical statutory language. The radical character of that suggested interpretation is illustrated by the following passage from an opinion decided only nine years after Allen:

"The Court's decisions over the past 10 years have given § 5 the broad scope suggested by the language of the Act. We first construed it in Allen v. State Board of Elections, [393 U. S. 544 (1969)]. There our examination of the Act's objectives and original legislative history led us to interpret § 5 to give it 'the broadest possible scope,' 393 U. S., at 567, and to require prior federal scrutiny of 'any state enactment which altered the election law in a covered State in even a minor way.' Id., at 566. In so construing § 5, we unanimously rejectedas the plain terms of the Act would themselves have seemingly required-the argument of an appellee that § 5 should apply only to enactments affecting who may register to vote. 393 U. S., at 564. Our decisions have required federal preclearance of laws changing the location of polling places, see Perkins v. Matthews, 400 U. S.

4 We recently confirmed that interpretation of the 1982 amendment, stating: "Moreover, there is no question that the terms 'standard, practice, or procedure' are broad enough to encompass the use of multimember districts to minimize a racial minority's ability to influence the outcome of an election covered by § 2." Chisom v. Roemer, 501 U. S. 380, 390 (1991). Though disagreeing with the Court's holding that the statute covered judicial elections, even the dissenters in that case agreed that the amended § 2 "extends to vote dilution claims for the elections of representatives .... " Id., at 405.


964

Opinion of STEVENS, J.

379 (1971), laws adopting at-large systems of election, ibid.; Fairley v. Patterson (decided with Allen, supra); laws providing for the appointment of previously elected officials, Bunton v. Patterson (decided with Allen, supra); laws regulating candidacy, Whitley v. Williams (decided with Allen, supra); laws changing voting procedures, Allen, supra; annexations, City of Richmond v. United States, 422 U. S. 358 (1975); City of Petersburg v. United States, 410 U. S. 962 (1973), summarily aff'g 354 F. Supp. 1021 (DC 1972); Perkins v. Matthews, supra; and reapportionment and redistricting, Beer v. United States, 425 U. S. 130 (1976); Georgia v. United States, 411 U. S. 526 (1973); see United Jewish Organizations v. Carey, 430 U. S. 144 (1977). In each case, federal scrutiny of the proposed change was required because the change had the potential to deny or dilute the rights conferred by § 4(a)." United States v. Sheffield Bd. ofComm'rs, 435 U. S., at 122-123 (footnote omitted).

The Allen interpretation of the Act has also been followed in a host of cases decided in later years, among them Houston Lawyers' Assn. v. Attorney General of Tex., 501 U. S. 419 (1991); Pleasant Grove v. United States, 479 U. S. 462 (1987); Thornburg v. Gingles, 478 U. S. 30 (1986); Port Arthur v. United States, 459 U. S. 159 (1982); City of Rome v. United States, 446 U. S. 156 (1980); Dougherty County Bd. of Ed. v. White, 439 U. S. 32 (1978). In addition, JUSTICE THOMAS' interpretation would call into question the numerous other cases since 1978 that have assumed the broad coverage of the Voting Rights Act that JUSTICE THOMAS would now have us reject. Chisom v. Roemer, 501 U. S. 380 (1991); Clark v. Roemer, 500 U. S. 646 (1991); McCain v. Lybrand, 465 U. S. 236 (1984); Hathorn v. Lovorn, 457 U. S. 255 (1982); Blanding v. DuBose, 454 U. S. 393 (1982); McDaniel v. Sanchez, 452 U. S. 130 (1981); Berry v. Doles, 438 U. S. 190 (1978); see also Presley v. Etowah County Comm'n, 502 U. S. 491 (1992); Voinovich v. Quilter, 507 U. S. 146 (1993); Growe v. Emison,


965

507 U. S. 25 (1993); City of Lockhart v. United States, 460 U. S. 125 (1983).

The large number of decisions that we would have to overrule or reconsider, as well as the congressional reenactments discussed above, suggests that JUSTICE THOMAS' radical reinterpretation of the Voting Rights Act is barred by the well-established principle that stare decisis has special force in the statutory arena. Ankenbrandt v. Richards, 504 U. S. 689, 700 (1992); Patterson v. McLean Credit Union, 491 U. S. 164, 171-172 (1989); Illinois Brick Co. v. Illinois, 431 U. S. 720, 736-737 (1977).

JUSTICE THOMAS attempts to minimize the radical implications of his interpretation of the phrase "voting qualification ... standard, practice, or procedure" by noting that this case involves only the interpretation of § 2 of the Voting Rights Act. Section 5, he hints, might be interpreted differently. Even limiting the reinterpretation to § 2 cases, however, would require overruling a sizable number of this Court's precedents. Houston Lawyers' Assn. v. Attorney General of Tex., 501 U. S. 419 (1991); Chisom v. Roemer, 501 U. S. 380 (1991); Thornburg v. Gingles, 478 U. S. 30 (1986); see also Voinovich v. Quilter, 507 U. S. 146 (1993); Growe v. Emison, 507 U. S. 25 (1993). In addition, a distinction between §§ 2 and 5 is difficult to square with the language of the statute. Sections 2 and 5 contain exactly the same words: "voting qualification ... standard, practice, or procedure." If anything, the wording of § 5 is narrower, because it adds the limiting phrase "with respect to voting" after the word "procedure." Moreover, when Congress amended the Voting Rights Act in 1982 in response to Bolden, it amended § 2. As noted above, in those amendments Congress clearly endorsed the application of the Voting Rights Act to vote dilution claims. While a distinction between §§ 2 and 5 might be supportable on policy grounds, it is an odd distinction for devotees of "plain language" interpretation.


966

Opinion of STEVENS, J.

Throughout his opinion, JUSTICE THOMAS argues that this case is an exception to stare decisis, because Allen and its progeny have "immersed the federal courts in a hopeless project of weighing questions of political theory." Ante, at 892. There is no question that the Voting Rights Act has required the courts to resolve difficult questions, but that is no reason to deviate from an interpretation that Congress has thrice approved. Statutes frequently require courts to make policy judgments. The Sherman Act, for example, requires courts to delve deeply into the theory of economic organization. Similarly, Title VII of the Civil Rights Act has required the courts to formulate a theory of equal opportunity. Our work would certainly be much easier if every case could be resolved by consulting a dictionary, but when Congress has legislated in general terms, judges may not invoke judicial modesty to avoid difficult questions.

III

When a statute has been authoritatively, repeatedly, and consistently construed for more than a quarter century, and when Congress has reenacted and extended the statute several times with full awareness of that construction, judges have an especially clear obligation to obey settled law. Whether JUSTICE THOMAS is correct that the Court's settled construction of the Voting Rights Act has been "a disastrous misadventure," ante, at 893, should not affect the decision in this case. It is therefore inappropriate for me to comment on the portions of his opinion that are best described as an argument that the statute be repealed or amended in important respects.