UNITED JEWISH ORGANIZATIONS v. CAREY

MR. JUSTICE WHITE announced the judgment of the Court and filed an opinion in which MR. JUSTICE STEVENS joined; Parts I, II, and III of which are joined by MR. JUSTICE BRENNAN and MR. JUSTICE BLACKMUN; and Parts I and IV of which are joined by MR. JUSTICE REHNQUIST.

Section 5 of the Voting Rights Act of 1965 prohibits a State or political subdivision subject to § 4 of the Act from implementing a legislative reapportionment unless it has obtained a declaratory judgment from the District Court for the District of Columbia, or a ruling from the Attorney General of the United States, that the reapportionment "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 . . . ."1 The question presented is whether, in the circumstances of this case, the use of racial criteria by the State of New York in its attempt to comply with § 5 of the Voting Rights Act and to secure the approval of the Attorney General violated the Fourteenth or Fifteenth Amendment.

I

Kings County, N. Y., together with New York (Manhattan) and Bronx Counties, became subject to §§ 4 and 5 of the Act, by virtue of a determination by the Attorney General that a literacy test was used in these three counties as of November 1, 1968, and a determination by the Director of the Census that fewer than 50% of the voting-age residents of these three counties voted in the Presidential election of 1968.2 Litigation to secure exemption from the Act was unsuccessful,3 and it became necessary for New York to secure the approval of the Attorney General or of the United States District Court for the District of Columbia for its 1972 reapportionment statute insofar as that statute concerned Kings, New York, and Bronx Counties. On January 31, 1974, the provisions of the statute districting these counties for congressional, state senate, and state assembly seats were submitted to the Attorney General. In accordance with the regulations governing his § 5 review, the Attorney General considered submissions from interested parties criticizing and defending the plan.4 Those submissions included assertions that voting in these counties was racially polarized and that the district lines had been created with the purpose or effect of diluting the voting strength of nonwhites (blacks and Puerto Ricans).5 On April 1, 1974, the Attorney General concluded that, as to certain districts in Kings County covering the Bedford-Stuyvesant area of Brooklyn, the State had not met the burden placed on it by § 5 and the regulations thereunder to demonstrate that the redistricting had neither the purpose nor the effect of abridging the right to vote by reason of race or color.6

Under § 5, the State could have challenged the Attorney General's objections to the redistricting plan by filing a declaratory judgment action in a three-judge court in the District of Columbia. Instead, the State sought to meet what it understood to be the Attorney General's objections and to secure his approval in order that the 1974 primary and general elections could go forward under the 1972 statute.7 A revised plan, submitted to the Attorney General on May 31, 1974, in its essentials did not change the number of districts with nonwhite majorities, but did change the size of the nonwhite majorities in most of those districts. Under the 1972 plan, Kings County had three state senate district with nonwhite majorities of approximately 91%, 61%, and 53%; under the revised 1974 plan, there were again three districts with nonwhite majorities, but now all three were between 70% and 75% nonwhite.8 As for state assembly districts, both the 1972 and the 1974 plans provided for seven districts with nonwhite majorities. However, under the 1972 plan, there were four between 85% and 95% nonwhite, and three were approximately 76%, 61%, and 52%, respectively; under the 1974 plan, the two smallest nonwhite majorities were increased to 65% and 67.5%, and the two largest nonwhite majorities were decreased from greater than 90% to between 80% and 90%.9 The report of the legislative committee on reapportionment stated that these changes were made "to overcome Justice Department objections" by creating more "substantial nonwhite majorities" in two assembly districts and two senate districts.10

One of the communities affected by these revisions in the Kings County reapportionment plan was the Williamsburgh area, where about 30,000 Hasidic Jews live. Under the 1972 plan, the Hasidic community was located entirely in one assembly district (61% nonwhite) and one senate district (37% nonwhite); in order to create substantial nonwhite majorities in these districts, the 1974 revisions split the Hasidic community between two senate and two assembly districts. A staff member of the legislative reapportionment committee testified that in the course of meetings and telephone conversations with Justice Department officials, he "got the feeling . . . that 65 percent would be probably an approved figure" for the nonwhite population in the assembly district in which the Hasidic community was located, a district approximately 61% nonwhite under the 1972 plan.11 To attain the 65% figure, a portion of the white population, including part of the Hasidic community, was reassigned to an adjoining district.

Shortly after the State submitted this revised redistricting plan for Kings County to the Attorney General, petitioners sued on behalf of the Hasidic Jewish community of Williamsburgh, alleging that the 1974 plan "would dilute the value of each plaintiff's franchise by halving its effectiveness," solely for the purpose of achieving a racial quota and therefore in violation of the Fourteenth Amendment. Petitioners also alleged that they were assigned to electoral districts solely on the basis of race, and that this racial assignment diluted their voting power in violation of the Fifteenth Amendment. Petitioners sought an injunction restraining New York officials from enforcing the new redistricting plan and a declaratory judgment that the Attorney General of the United States had used unconstitutional and improper standards in objecting to the 1972 plan.

On June 20, 1974, the District Court held a hearing on petitioners' motion for a preliminary injunction. On July 1, 1974, the Attorney General informed the State of New York that he did not object to the implementation of the revised plan. The Attorney General moved to be dismissed as a party on the ground that the relief sought against him could be obtained only in the District Court for the District of Columbia and only by a State or political subdivision subject to the Voting Rights Act; the State and the intervenor NAACP moved to dismiss the complaint on the ground that it failed to state a claim upon which relief could be granted. The District Court granted the motions to dismiss the complaint, reasoning that petitioners enjoyed no constitutional right in reapportionment to separate community recognition as Hasidic Jews, that the redistricting did not disenfranchise petitioners, and that racial considerations were permissible to correct past discrimination.12 United Jewish Organizations v. Wilson, 377 F.Supp. 1164, 1165-1166 (EDNY 1974).

A divided Court of Appeals affirmed. 510 F.2d 512 (CA2 1975). The majority first held that the Attorney General had to be dismissed as a party because the court had no jurisdiction to review his objection to the 1972 plan.13 After agreeing with the District Court that petitioners had no constitutional right to separate community recognition in reapportionment —a holding not challenged by petitioners here14—the Court of Appeals went on to address petitioners' claims as white voters that the 1974 plan denied them equal protection of the laws and abridged their right to vote on the basis of race. The court noted that the 1974 plan left approximately 70% of the senate and assembly districts in Kings County with white majorities; given that only 65% of the population of the county was white, the 1974 plan would not underrepresent the white population, assuming that voting followed racial lines. Id., at 523, and n. 21. Petitioners thus could not claim that the plan canceled out the voting strength of whites as a racial group, under this Court's decisions in White v. Regester, 412 U.S. 755 (1973), and Whitcomb v. Chavis, 403 U.S. 124 (1971). The court then observed that the case did not present the question whether a legislature, "starting afresh," could draw lines on a racial basis so as to bolster nonwhite voting strength, but rather the "narrower" question whether a State could use racial considerations in drawing lines in an effort to secure the Attorney General's approval under the Voting Rights Act. 510 F. 2d, at 524. The court thought this question answered by this Court's decision in Allen v. State Board of Elections, 393 U.S. 544, 569 (1969), where a change from district to at-large voting for county supervisors was held to be covered by § 5 of the Act. The court below reasoned that the Act contemplated that the Attorney General and the state legislature would have "to think in racial terms"; because the Act "necessarily deals with race or color, corrective action under it must do the same." 510 F. 2d, at 525. (Emphasis in original; footnote omitted.) The court held that

"so long as a districting, even though based on racial considerations, is in conformity with the unchallenged directive of and has the approval of the Attorney General of the United States under the Act, at least absent a clear showing that the resultant legislative reapportionment is unfairly prejudicial to white or nonwhite, that districting is not subject to challenge." Ibid.15

We granted certiorari, 423 U.S. 945 (1975). We affirm.

II

Petitioners argue that the New York Legislature, although seeking to comply with the Voting Rights Act as construed by the Attorney General, has violated the Fourteenth and Fifteenth Amendments by deliberately revising its reapportionment plan along racial lines.16 In rejecting petitioners' claims, we address four propositions: First, that whatever might be true in other contexts, the use of racial criteria in districting and apportionment is never permissible; second, that even if racial considerations may be used to redraw district lines in order to remedy the residual effects of past unconstitutional reapportionments, there are no findings here of prior discriminations that would require or justify as a remedy that white voters be reassigned in order to increase the size of black majorities in certain districts; third, that the use of a "racial quota" in redistricting is never acceptable; and fourth, that even if the foregoing general propositions are infirm, what New York actually did in this case was unconstitutional, particularly its use of a 65% nonwhite racial quota for certain districts. The first three arguments, as we now explain, are foreclosed by our cases construing and sustaining the constitutionality of the Voting Rights Act; the fourth we address in Parts III and IV.

It is apparent from the face of the Act, from its legislative history, and from our cases that the Act itself was broadly remedial in the sense that it was "designed by Congress to banish the blight of racial discrimination in voting . . . ." South Carolina v. Katzenbach, 383 U.S. 301, 308 (1966). It is also plain, however, that after "repeatedly try[ing] to cope with the problem by facilitating case-by-case litigation against voting discrimination," id., at 313, Congress became dissatisfied with this approach, which required judicial findings of unconstitutional discrimination in specific situations and judicially approved remedies to cure that discrimination. Instead, Congress devised more stringent measures, one of which, § 5, required the covered States to seek the approval of either the Attorney General or of a three-judge court in the District of Columbia whenever they sought to implement new voting procedures. Under § 4, a State became subject to § 5 whenever it was administratively determined that certain conditions which experience had proved were indicative of racial discrimination in voting had existed in the area—in the case of New York, as already indicated, supra, at 148, that a literacy test was in use in certain counties in 1968 and that fewer than 50% of the voting-age residents in these counties voted in the Presidential election that year. At that point, New York could have escaped coverage by demonstrating to the appropriate court that the test had not been used to discriminate within the past 10 years, which New York was unable to do. See n. 3, supra.

Given this coverage of the counties involved, it is evident that the Act's prohibition against instituting new voting procedures without the approval of the Attorney General or the three-judge District Court is not dependent upon proving past unconstitutional apportionments and that in operation the Act is aimed at preventing the use of new procedures until their capacity for discrimination has been examined by the Attorney General or by a court. Although recognizing that the "stringent new remedies," including § 5, were "an uncommon exercise of congressional power," we nevertheless sustained the Act as a "permissibly decisive" response to "the extraordinary stratagem of contriving new rules of various kinds for the sole purpose of perpetrating voting discrimination in the face of adverse federal court decrees." South Carolina v. Katzenbach, supra, at 334-335 (footnote omitted).

It is also clear that under § 5, new or revised reapportionment plans are among those voting procedures, standards, or practices that may not be adopted by a covered State without the Attorney General's or a three-judge court's ruling that the plan "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." In Allen v. State Board of Elections, on which the Court of Appeals relied below, 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] 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.17 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.18

As the Court of Appeals understood the Act and our decision in Allen, compliance with the Act in reapportionment cases would often necessitate the use of racial considerations in drawing district lines. That the Court of Appeals correctly read the Act has become clearer from later cases.

In Beer v. United States, 425 U.S. 130 (1976), the Court considered the question of what criteria a legislative reapportionment must satisfy under § 5 of the Voting Rights Act to demonstrate that it does not have the "effect" of denying or abridging the right to vote on account of race. Beer established that the Voting Rights Act does not permit the implementation of a reapportionment 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. This test was satisfied where the reapportionment increased the percentage of districts where members of racial minorities protected by the Act were in the majority. See ibid. But if this test were not met, clearance by the Attorney General or the District Court for the District of Columbia could not be given, and the reapportionment could not be implemented.

The reapportionment at issue in Beer was approved by this Court, because New Orleans had created one councilmanic district with a majority of black voters where none existed before. But had there been districts with black majorities under the previous law and had New Orleans in fact decreased the number of majority black districts, it would have had to modify its plan in order to implement its reapportionment by carving out a large enough black majority in however many additional districts would be necessary to satisfy the Beer test. There was division on the Court as to what a State must show to satisfy § 5; but all eight Justices who participated in the decision implicitly accepted the proposition that a State may revise its reapportionment plan to comply with § 5 by increasing the percentage of black voters in a particular district until it has produced a clear majority. See 425 U. S., at 141-142; id., at 144 (WHITE, J., dissenting); id., at 158-161 (MARSHALL, J., dissenting). Indeed, the plan eventually approved by this Court in Beer was drawn with the purpose of avoiding dilution of the black vote by attaining at least a 54% majority of black voters in one district while preventing a 90% concentration. See App. in Beer v. United States, O. T. 1975, No. 73-1869, pp. 341-342.

The Court has taken a similar approach in applying § 5 to the extension of city boundaries through annexation. Where the annexation has the effect of reducing the percentage of blacks in the city, the proscribed "effect" on voting rights can be avoided by a post-annexation districting plan which "fairly reflects the strength of the Negro community as it exists after the annexation" and which "would afford [it] representation reasonably equivalent to [its] political strength in the enlarged community." City of Richmond v. United States, 422 U.S. 358, 370-371 (1975). Accord, City of Petersburg v. United States, 354 F.Supp. 1021 (DC 1972), summarily aff'd, 410 U.S. 962 (1973). In City of Richmond, the Court approved an annexation which reduced the proportion of blacks in the city from 52% to 42%, because the post-annexation ward system created four out of nine wards with substantial black majorities of 64%. Had the redistricting failed to "fairly [reflect] the strength of the Negro community," however, it would follow from the Court's decision that the Constitution would permit the city to modify its plan by deliberately creating black majorities in a sufficient number of wards to satisfy statutory requirements.

Implicit in Beer and City of Richmond, then, is the proposition that the Constitution does not prevent a State subject to the Voting Rights Act from deliberately creating or preserving black majorities in particular districts in order to ensure that its reapportionment plan complies with § 5. That proposition must be rejected and § 5 held unconstitutional to that extent if we are to accept petitioners' view that racial criteria may never be used in redistricting or that they may be used, if at all, only as a specific remedy for past unconstitutional apportionments. We are unwilling to overturn our prior cases, however. Section 5 and its authorization for racial redistricting where appropriate to avoid abridging the right to vote on account of race or color are constitutional. Contrary to petitioners' first argument, neither the Fourteenth nor the Fifteenth Amendment mandates any per se rule against using racial factors in districting and apportionment. Nor is petitioners' second argument valid. The permissible use of racial criteria is not confined to eliminating the effects of past discriminatory districting or apportionment.19

Moreover, in the process of drawing black majority districts in order to comply with § 5, the State must decide how substantial those majorities must be in order to satisfy the Voting Rights Act. The figure used in drawing the Beer plan, for example, was 54% of registered voters.20 At a minimum and by definition, a "black majority district" must be more than 50% black. But whatever the specific percentage, the State will inevitably arrive at it as a necessary means to ensure the opportunity for the election of a black representative and to obtain approval of its reapportionment plan. Unless we adopted an unconstitutional construction of § 5 in Beer and City of Richmond, a reapportionment cannot violate the Fourteenth or Fifteenth Amendment merely because a State uses specific numerical quotas in establishing a certain number of black majority districts. Our cases under § 5 stand for at least this much.

III

Having rejected these three broad objections to the use of racial criteria in redistricting under the Voting Rights Act, we turn to the fourth question, which is whether the racial criteria New York used in this case—the revision of the 1972 plan to create 65% nonwhite majorities in two additional senate and two additional assembly districts—were constitutionally infirm. We hold they are not, on two separate grounds. The first is addressed in this Part III, the second in Part IV.

The first ground is that petitioners have not shown, or offered to prove, that New York did more than the Attorney General was authorized to require it to do under the nonretrogression principle of Beer, a principle that, as we have already indicated, this Court has accepted as constitutionally valid. Under Beer, the acceptability of New York's 1972 reapportionment for purposes of § 5 depends on the change in nonwhite voting strength in comparison with the previous apportionment, which occurred in 1966. Yet there is no evidence in the record to show whether the 1972 plan increased or decreased the number of senate or assembly districts with substantial nonwhite majorities of 65%. For all that petitioners have alleged or proved, the 1974 revisions may have accomplished nothing more than the restoration of nonwhite voting strength to 1966 levels.21 To be successful in their constitutional challenge to the racial criteria used in New York's revised plan, petitioners must show at a minimum that minority voting strength was increased under the 1974 plan in comparison with the 1966 apportionment; otherwise the challenge amounts to a constitutional attack on compliance with the statutory rule of nonretrogression.

In the absence of any evidence regarding nonwhite voting strength under the 1966 apportionment, the creation of substantial nonwhite majorities in approximately 30% of the senate and assembly districts in Kings County was reasonably related to the constitutionally valid statutory mandate of maintaining nonwhite voting strength. The percentage of districts with nonwhite majorities was less than the percentage of nonwhites in the county as a whole (35%). The size of the nonwhite majorities in those districts reflected the need to take account of the substantial difference between the nonwhite percentage of the total population in a district and the nonwhite percentage of the voting-age population.22 Because, as the Court said in Beer, the inquiry under § 5 focuses ultimately on "the position of racial minorities with respect to their effective exercise of the electoral franchise," 425 U. S., at 141, the percentage of eligible voters by district is of great importance to that inquiry.23 In the redistricting plan approved in Beer, for example, only one of the two districts with a black population majority also had a black majority of registered voters. Id., at 142. We think it was reasonable for the Attorney General to conclude in this case that a substantial nonwhite population majority—in the vicinity of 65%— would be required to achieve a nonwhite majority of eligible voters.

Petitioners have not shown that New York did more than accede to a position taken by the Attorney General that was authorized by our constitutionally permissible construction of § 5. New York adopted the 1974 plan because it sought to comply with the Voting Rights Act. This has been its primary defense of the plan, which was sustained on that basis by the Court of Appeals. Because the Court of Appeals was essentially correct, its judgment may be affirmed without addressing the additional argument by New York and by the United States that, wholly aside from New York's obligation under the Voting Rights Act to preserve minority voting strength in Kings County, the Constitution permits it to draw district lines deliberately in such a way that the percentage of districts with a nonwhite majority roughly approximates the percentage of nonwhites in the county.

IV

This additional argument, however, affords a second, and independent, ground for sustaining the particulars of the 1974 plan for Kings County. Whether or not the plan was authorized by or was in compliance with § 5 of the Voting Rights Act, New York was free to do what it did as long as it did not violate the Constitution, particularly the Fourteenth and Fifteenth Amendments; and we are convinced that neither Amendment was infringed.

There is no doubt that in preparing the 1974 legislation, the State deliberately used race in a purposeful manner. But its plan represented no racial slur or stigma with respect to whites or any other race, and we discern no discrimination violative of the Fourteenth Amendment nor any abridgment of the right to vote on account of race within the meaning of the Fifteenth Amendment.

It is true that New York deliberately increased the nonwhite majorities in certain districts in order to enhance the opportunity for election of nonwhite representatives from those districts. Nevertheless, there was no fencing out of the white population from participation in the political processes of the county, and the plan did not minimize or unfairly cancel out white voting strength. Compare White v. Regester, 412 U. S., at 765-767, and Gomillion v. Lightfoot, 364 U.S. 339 (1960), with Gaffney v. Cummings, 412 U.S. 735, 751-754 (1973). Petitioners have not objected to the impact of the 1974 plan on the representation of white voters in the county or in the State as a whole. As the Court of Appeals observed, the plan left white majorities in approximately 70% of the assembly and senate districts in Kings County, which had a countywide population that was 65% white. Thus, even if voting in the county occurred strictly according to race, whites would not be underrepresented relative to their share of the population.

In individual districts where nonwhite majorities were increased to approximately 65%, it became more likely, given racial bloc voting, that black candidates would be elected instead of their white opponents, and it became less likely that white voters would be represented by a member of their own race; but as long as whites in Kings County, as a group, were provided with fair representation, we cannot conclude that there was a cognizable discrimination against whites or an abridgment of their right to vote on the grounds of race.24 Furthermore, the individual voter in the district with a nonwhite majority has no constitutional complaint merely because his candidate has lost out at the polls and his district is represented by a person for whom he did not vote. Some candidate, along with his supporters, always loses. See Whitcomb v. Chavis, 403 U. S., at 153-160.

Where it occurs, voting for or against a candidate because of his race is an unfortunate practice. But it is not rare; and in any district where it regularly happens, it is unlikely that any candidate will be elected who is a member of the race that is in the minority in that district. However disagreeable this result may be, there is no authority for the proposition that the candidates who are found racially unacceptable by the majority, and the minority voters supporting those candidates, have had their Fourteenth or Fifteenth Amendment rights infringed by this process. Their position is similar to that of the Democratic or Republican minority that is submerged year after year by the adherents to the majority party who tend to vote a straight party line.

It does not follow however, that the State is powerless to minimize the consequences of racial discrimination by voters when it is regularly practiced at the polls. In Gaffney v. Cummings, the Court upheld a districting plan "drawn with the conscious intent to . . . achieve a rough approximation of the statewide political strengths of the Democratic and Republican Parties." 412 U. S., at 752. We there recognized that districting plans would be vulnerable under our cases if "racial or political groups have been fenced out of the political process and their voting strength invidiously minimized," id., at 754 (emphasis added); but that was not the case there, and no such purpose or effect may be ascribed to New York's 1974 plan. Rather, that plan can be viewed as seeking to alleviate the consequences of racial voting at the polls and to achieve a fair allocation of political power between white and nonwhite voters in Kings County.

In this respect New York's revision of certain district lines is little different in kind from the decision by a State in which a racial minority is unable to elect representatives from multimember districts to change to single-member districting for the purpose of increasing minority representation. This change might substantially increase minority representation at the expense of white voters, who previously elected all of the legislators but who with single-member districts could elect no more than their proportional share. If this intentional reduction of white voting power would be constitutionally permissible, as we think it would be, we think it also permissible for a State, employing sound districting principles such as compactness and population equality, to attempt to prevent racial minorities from being repeatedly outvoted by creating districts that will afford fair representation to the members of those racial groups who are sufficiently numerous and whose residential patterns afford the opportunity of creating districts in which they will be in the majority.

As the Court said in Gaffney:

"[C]ourts have [no] constitutional warrant to invalidate a state plan, otherwise within tolerable population limits, because it undertakes, not to minimize or eliminate the political strength of any group or party, but to recognize it and, through districting, provide a rough sort of proportional representation in the legislative halls of the State." Ibid.

New York was well within this rule when, under the circumstances present in Kings County, it amended its 1972 plan.25

The judgment is

Affirmed.

MR. JUSTICE MARSHALL took no part in the consideration or decision of this case.

MR. JUSTICE BRENNAN, concurring in part.

I join Parts I, II, and III of Mr. JUSTICE WHITE'S opinion. Part II effectively demonstrates that prior cases firmly establish the Attorney General's expansive authority to oversee legislative redistricting under § 5 of the Voting Rights Act. See, e. g., Georgia v. United States, 411 U.S. 526, 532 (1973); Allen v. State Board of Elections, 393 U.S. 544, 566, 569 (1969). Part III establishes to my satisfaction that as a method of securing compliance with the Voting Rights Act, the 65% rule applied to Brooklyn in this instance was not arbitrarily or casually selected. Yet, because this case carries us further down the road of race-centered remedial devices than we have heretofore traveled—with the serious questions of fairness that attend such matters—I offer this further explanation of my position.

The one starkly clear fact of this case is that an overt racial number was employed to effect petitioners' assignment to voting districts. In brief, following the Attorney General's refusal to certify the 1972 reapportionment under his § 5 powers, unnamed Justice Department officials made known that satisfaction of the Voting Rights Act in Brooklyn would necessitate creation by the state legislature of 10 state assembly and senate districts with threshold nonwhite populations of 65%. Prompted by the necessity of preventing interference with the upcoming 1974 election, state officials complied. Thus, the Justice Department's unofficial instruction to state officials effectively resulted in an explicit process of assignment to voting districts pursuant to race. The result of this process was a countywide pattern of districting closely approximating proportional representation. While it is true that this demographic outcome did not "underrepresent the white population" throughout the county, ante, at 154—indeed, the very definition of proportional representation precludes either underrepresentation or overrepresentation —these particular petitioners filed suit to complain that they have been subjected to a process of classification on the basis of race that adversely altered their status.

If we were presented here with a classification of voters motivated by racial animus, City of Richmond v. United States, 422 U.S. 358, 378 (1975); Wright v. Rockefeller, 376 U.S. 52, 58 (1964); Gomillion v. Lightfoot, 364 U.S. 339, 347 (1960), or with a classification that effectively downgraded minority participation in the franchise, Georgia v. United States, supra, at 534; Whitcomb v. Chavis, 403 U.S. 124, 144 (1971), we promptly would characterize the resort to race as "suspect" and prohibit its use. Under such circumstances, the tainted apportionment process would not necessarily be saved by its proportional outcome, for the segregation of voters into "separate but equal" blocs still might well have the intent or effect of diluting the voting power of minority voters. See, e. g., City of Richmond v. United States, supra, at 378; Wright v. Rockefeller, supra, at 53-54; infra, at 172-173. It follows, therefore, that if the racial redistricting involved here, imposed with the avowed intention of clustering together 10 viable nonwhite majorities at the expense of preexisting white groupings, is not similarly to be prohibited, the distinctiveness that avoids this prohibition must arise from either or both of two considerations: the permissibility of affording preferential treatment to disadvantaged nonwhites generally, or the particularized application of the Voting Rights Act in this instance.

The first and broader of the two plausible distinctions rests upon the general propriety of so-called benign discrimination: The challenged race assignment may be permissible because it is cast in a remedial context with respect to a disadvantaged class rather than in a setting that aims to demean or insult any racial group. Even in the absence of the Voting Rights Act, this preferential policy plausibly could find expression in a state decision to overcome nonwhite disadvantages in voter registration or turnout through redefinition of electoral districts—perhaps, as here, through the application of a numerical rule—in order to achieve a proportional distribution of voting power. Such a decision, in my view, raises particularly sensitive issues of doctrine and policy. Unlike Part IV of MR. JUSTICE WHITE'S opinion,1 I am wholly content to leave this thorny question until another day, for I am convinced that the existence of the Voting Rights Act makes such a decision unnecessary and alone suffices to support an affirmance of the judgment before us.

I begin with the settled principle that not every remedial use of race is forbidden. For example, we have authorized and even required race-conscious remedies in a variety of corrective settings. See, e. g., Swann v. Charlotte-Mecklenburg Bd. of Education, 402 U.S. 1, 25 (1971); United States v. Montgomery County Bd. of Education, 395 U.S. 225 (1969); Franks v. Bowman Transp. Co., 424 U.S. 747, 772-774 (1976); ante, at 160. Once it is established that circumstances exist where race may be taken into account in fashioning affirmative policies,2 we must identify those circumstances, and, further, determine how substantial a reliance may be placed upon race. If resort to the 65% rule involved here is not to be sanctioned, that must be because the benign use of such a binding numerical criterion (under the Voting Rights Act) generates problems of constitutional dimension that are not relevant to other, previously tolerated race-conscious remedies. As a focus for consideration of what these problems might or might not be, it is instructive to consider some of the objections frequently raised to the use of overt preferential race-assignment practices.

First, a purportedly preferential race assignment may in fact disguise a policy that perpetuates disadvantageous treatment of the plan's supposed beneficiaries. Accordingly, courts might face considerable difficulty in ascertaining whether a given race classification truly furthers benign rather than illicit objectives. An effort to achieve proportional representation, for example, might be aimed at aiding a group's participation in the political processes by guaranteeing safe political offices, or, on the other hand, might be a "contrivance to segregate" the group, Wright v. Rockefeller, supra, at 58, thereby frustrating its potentially successful efforts at coalition building across racial lines. Compare, e. g., the positions of the black plaintiffs in Wright, supra, at 53-54, with the black intervenors, 376 U. S., at 62 (Douglas, J., dissenting). Indeed, even the present case is not entirely free of complaints that the remedial redistricting in Brooklyn is not truly benign. Puerto Rican groups, for example, who have been joined with black groups to establish the "nonwhite" category, protested to the Attorney General that their political strength under the 1974 reapportionment actually is weaker than under the invalidated 1972 districting. App. 295. A black group similarly complained of the loss of a "safe" seat because of the inadequacy of the 65% target figure. Id., at 296-297. These particular objections, as the Attorney General argued in his memorandum endorsing the 1974 reapportionment, may be ill-advised and unpersuasive. Nevertheless, they illustrate the risk that what is presented as an instance of benign race assignment in fact may prove to be otherwise. This concern, of course, does not undercut the theoretical legitimacy or usefulness of preferential policies. At the minimum, however, it does suggest the need for careful consideration of the operation of any racial device, even one cloaked in preferential garb. And if judicial detection of truly benign policies proves impossible or excessively crude, that alone might warrant invalidating any race-drawn line.

Second, even in the pursuit of remedial objectives, an explicit policy of assignment by race may serve to stimulate our society's latent race consciousness, suggesting the utility and propriety of basing decisions on a factor that ideally bears no relationship to an individual's worth or needs. See, e. g., Kaplan, Equal Justice in an Unequal World: Equality for the Negro—The Problem of Special Treatment, 61 Nw. U. L. Rev. 363, 379-380 (1966). Furthermore, even preferential treatment may act to stigmatize its recipient groups, for although intended to correct systemic or institutional inequities, such a policy may imply to some the recipients' inferiority and especial need for protection.3 Again, these matters would not necessarily speak against the wisdom or permissibility of selective, benign racial classifications. But they demonstrate that the considerations that historically led us to treat race as a constitutionally "suspect" method of classifying individuals are not entirely vitiated in a preferential context.

Third, especially when interpreting the broad principles embraced by the Equal Protection Clause, we cannot well ignore the social reality that even a benign policy of assignment by race is viewed as unjust by many in our society, especially by those individuals who are adversely affected by a given classification. This impression of injustice may be heightened by the natural consequence of our governing processes that the most "discrete and insular" of whites often will be called upon to bear the immediate, direct costs of benign discrimination. See, e. g., Kaplan, supra, at 373-374; cf. Ely, The Constitutionality of Reverse Racial Discrimination, 41 U. Chi. L. Rev. 723, 737-738 (1974). Perhaps not surprisingly, there are indications that this case affords an example of just such decisionmaking in operation. For example, the respondent-intervenors take pains to emphasize that the mandated 65% rule could have been attained through redistricting strategies that did not slice the Hasidic community in half. State authorities, however, chose to localize the burdens of race reassignment upon the petitioners rather than to redistribute a more varied and diffused range of whites into predominatly nonwhite districts. Brief for Respondent-Intervenors 29-31. I am in no position to determine the accuracy of this appraisal, but the impression of unfairness is magnified when a coherent group like the Hasidim disproportionately bears the adverse consequences of a race-assignment policy.

In my view, if and when a decisionmaker embarks on a policy of benign racial sorting, he must weigh the concerns that I have discussed against the need for effective social policies promoting racial justice in a society beset by deep-rooted racial inequities. But I believe that Congress here adequately struck that balance in enacting the carefully conceived remedial scheme embodied in the Voting Rights Act. However the Court ultimately decides the constitutional legitimacy of "reverse discrimination" pure and simple, I am convinced that the application of the Voting Rights Act substantially minimizes the objections to preferential treatment, and legitimates the use of even overt, numerical racial devices in electoral redistricting.

The participation of the Attorney General, for example, largely relieves the judiciary of the need to grapple with the difficulties of distinguishing benign from malign discrimination. Under § 5 of the Act, the Attorney General in effect is constituted champion of the interests of minority voters, and accompanying implementing regulations ensure the availability of materials and submissions necessary to discern the true effect of a proposed reapportionment plan. See 28 CFR § 51.19 (1976). This initial right of review, coupled with the factfinding competence of the Justice Department, substantially reduces the likelihood that a complicated reapportionment plan that silently furthers malign racial policies would escape detection by appropriate officials. As a practical matter, therefore, I am prepared to accord considerable deference to the judgment of the Attorney General that a particular districting scheme complies with the remedial objectives furthered by the Voting Rights Act.

Similarly, the history of the Voting Rights Act provides reassurance that, in the face of the potential for reinvigorating racial partisanship, the congressional decision to authorize the use of race-oriented remedies in this context was the product of substantial and careful deliberations. Enacted following "voluminous legislative" consideration, South Carolina v. Katzenbach, 383 U.S. 301, 309 (1966), the Voting Rights Act represents an unequivocal and well-defined congressional consensus on the national need for "sterner and more elaborate measures," ibid., to secure the promise of the Fourteenth and Fifteenth Amendments with respect to exercise of the franchise. Insofar as the drawing of district lines is a process that intrinsically involves numerical calculations, and insofar as state officials charged with the task of defining electoral constituencies are unlikely simply to close their eyes to considerations such as race and national origin,4 the resort to a numerical racial criterion as a method of achieving compliance with the aims of the Voting Rights Act is, in my view, consistent with that consensus. Whatever may be the indirect and undesirable counter-educational costs of employing such far-reaching racial devices, Congress had to confront these considerations before opting for an activist race-conscious remedial role supervised by federal officials. The "insidious and pervasive" evil of voting rights violations, 383 U. S., at 309, and the "specially informed legislative competence" in this area, Katzenbach v. Morgan, 384 U.S. 641, 656 (1966); cf. Morton v. Mancari, 417 U.S. 535, 555 (1974), argue in support of the legitimacy of the federal decision to permit a broad range of raceconscious remedial techniques, including, as here, outright assignment by race.

This leaves, of course, the objection expressed by a variety of participants in this litigation: that this reapportionment worked the injustice of localizing the direct burdens of racial assignment upon a morally undifferentiated group of whites,5 and, indeed, a group that plausibly is peculiarly vulnerable to such injustice. This argument has both normative and emotional appeal, but for a variety of reasons I am convinced that the Voting Rights Act drains it of vitality.

First, it is important to recall that the Attorney General's oversight focuses upon jurisdictions whose prior practices exhibited the purpose or effect of infringing the right to vote on account of race, thereby triggering § 4 of the Act, 42 U. S. C. § 1973b (1970 ed. and Supp. V). This direct nexus to localities with a history of discriminatory practices or effects enhances the legitimacy of the Attorney General's remedial authority6 over individuals within those communities who benefited (as whites) from those earlier discriminatory voting patterns. Moreover, the obvious remedial nature of the Act and its enactment by an elected Congress that hardly can be viewed as dominated by nonwhite representatives belie the possibility that the decisionmaker intended a racial insult or injury to those whites who are adversely affected by the operation of the Act's provisions.7 Finally, petitioners have not been deprived of their right to vote, a consideration that minimizes the detrimental impact of the remedial racial policies governing the § 5 reapportionment. True, petitioners are denied the opportunity to vote as a group in accordance with the earlier districting configuration, but they do not press any legal claim to a group voice as Hasidim. Brief for Petitioners 6 n. 6. In terms of their voting interests, then, the burden that they claim to suffer must be attributable solely to their relegation to increased nonwhite-dominated districts. Yet, to the extent that white and nonwhite interests and sentiments are polarized in Brooklyn, the petitioners still are indirectly "protected" by the remaining white assembly and senate districts within the county, carefully preserved in accordance with the white proportion of the total county population. While these considerations obviously do not satisfy petitioners, I am persuaded that they reinforce the legitimacy of this remedy.

Since I find nothing in the first three parts of Mr. JUSTICE WHITE'S opinion that is inconsistent with the views expressed herein, I join those parts.

MR. JUSTICE STEWART, with whom MR. JUSTICE POWELL joins, concurring in the judgment.

The question presented for decision in this case is whether New York's use of racial criteria in redistricting Kings County violated the Fourteenth or Fifteenth Amendment. The petitioners' contention is essentially that racial awareness in legislative reapportionment is unconstitutional per se. Acceptance of their position would mark an egregious departure from the way this Court has in the past analyzed the constitutionality of claimed discrimination in dealing with the elective franchise on the basis of race.

The petitioners have made no showing that a racial criterion was used as a basis for denying them their right to vote, in contravention of the Fifteenth Amendment. See Gomillion v. Lightfoot, 364 U.S. 339. They have made no showing that the redistricting scheme was employed as part of a "contrivance to segregate"; to minimize or cancel out the voting strength of a minority class or interest; or otherwise to impair or burden the opportunity of affected persons to participate in the political process. See Wright v. Rockefeller, 376 U.S. 52, 58; White v. Regester, 412 U.S. 755; Louisiana v. United States, 380 U.S. 145; Fortson v. Dorsey, 379 U.S. 433.

Under the Fourteenth Amendment the question is whether the reapportionment plan represents purposeful discrimination against white voters. Washington v. Davis, 426 U.S. 229. Disproportionate impact may afford some evidence that an invidious purpose was present. Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U.S. 252, 266. But the record here does not support a finding that the redistricting plan undervalued the political power of white voters relative to their numbers in Kings County. Cf. City of Richmond v. United States, 422 U.S. 358. That the legislature was aware of race when it drew the district lines might also suggest a discriminatory purpose. Such awareness is not, however, the equivalent of discriminatory intent. The clear purpose with which the New York Legislature acted—in response to the position of the United States Department of Justice under the Voting Rights Act—forecloses any finding that it acted with the invidious purpose of discriminating against white voters.*

Having failed to show that the legislative reapportionment plan had either the purpose or the effect of discriminating against them on the basis of their race, the petitioners have offered no basis for affording them the constitutional relief they seek. Accordingly, I join the judgment of the Court.

MR. CHIEF JUSTICE BURGER, dissenting.

The question presented in this difficult case is whether New York violated the rights of the petitioners under the Fourteenth and Fifteenth Amendments by direct reliance on fixed racial percentages in its 1974 redistricting of Kings County. For purposes of analysis I will treat this in two steps: (1) Is the state legislative action constitutionally permissible absent any special considerations raised by the Federal Voting Rights Act; and (2) does New York's obligation to comply with the Voting Rights Act permit it to use these means to achieve a federal statutory objective?

(1)

I begin with this Court's holding in Gomillion v. Lightfoot, 364 U.S. 339 (1960), the first case to strike down a state attempt at racial gerrymandering. If Gomillion teaches anything, I had thought it was that drawing of political boundary lines with the sole, explicit objective of reaching a predetermined racial result cannot ordinarily be squared with the Constitution. The record before us reveals—and it is not disputed —that this is precisely what took place here. In drawing up the 1974 reapportionment scheme, the New York Legislature did not consider racial composition as merely one of several political characteristics; on the contrary, race appears to have been the one and only criterion applied.

The principal opinion notes that after the 1972 apportionment plan was rejected, New York officials conferred with the Justice Department as to what plan could obtain the Attorney General's approval. One New York official testified that he "`got the feeling [from a Justice Department spokesman]. . . that 65 percent would be probably an approved figure.'" Ante, at 152. Further testimony by that same official is revealing:

This official also testified that apportionment solutions which would have kept the Hasidic community within a single district, but would have resulted in a 63.4% nonwhite concentration, were rejected for fear that, falling short of "exactly 65 percent," they "would not be acceptable" to the Justice Department. Id., at 115.

The words "racial quota" are emotionally loaded and must be used with caution. Yet this undisputed testimony shows that the 65% figure was viewed by the legislative reapportionment committee as so firm a criterion that even a fractional deviation was deemed impermissible. I cannot see how this can be characterized otherwise than a strict quota approach and I must therefore view today's holding as casting doubt on the clear-cut principles established in Gomillion.

(2)

My second inquiry is whether the action of the State of New York becomes constitutionally permissible because it was taken to comply with the remedial provisions of the federal Voting Rights Act.

In South Carolina v. Katzenbach, 383 U.S. 301 (1966) the Court, while recognizing that the "stringent new remedies" were "an uncommon exercise of Congressional power" id., at 334-335, upheld the Act as a "permissibly decisive" response to "the extraordinary stratagem of . . . perpetrating voting discrimination in the face of adverse federal court decrees." Ibid. In Allen v. State Board of Elections, 393 U.S. 544, 569 (1969), the Court sustained an application of § 5 to a change from a district to an at-large election of county supervisors because of a potential for "dilution" of minority voting power which could "nullify [the] ability to elect the candidate of [one's] choice." In Allen and Katzenbach the Court acknowledged that the Voting Rights Act contemplated that the Attorney General and the affected state legislatures would be obliged to think in racial terms. In Perkins v. Matthews, 400 U.S. 379, 397 (1971) (concurring in judgment), and again in Georgia v. United States, 411 U.S. 526, 541 (1973) (dissenting opinion), I expressed doubt as to the correctness of Allen but acquiesced in the judgments on the basis of stare decisis.

The present case, however, presents a quite different situation. Faced with the straightforward obligation to redistrict so as to avoid "a retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise," Beer v. United States, 425 U.S. 130, 141 (1976), the state legislature mechanically adhered to a plan designed to maintain—without tolerance for even a 1.6% deviation—a "nonwhite" population of 65% within several of the new districts. There is no indication whatever that use of this rigid figure was in any way related—much less necessary —to fulfilling the State's obligation under the Voting Rights Act as defined in Beer.

The plurality opinion acknowledges our recent Beer holding by noting that "there is no evidence in the record to show whether the 1972 plan increased or decreased the number of senate or assembly districts with substantial nonwhite majorities of 65%," and by speculating that "the 1974 revisions may have accomplished nothing more than the restoration of non-white voting strength to 1966 levels." Ante, at 163. It then proceeds to assume that the 1974 reapportionment was undertaken in compliance with Beer. The lack of evidence on this subject is, of course, not surprising, since petitioners' case was dismissed at the pleading stage. If this kind of racial redistricting is to be upheld, however, it should, at the very least, be done on the basis of record facts, not suppositions. If the Court seriously considers the issue in doubt, I should think that a remand for further factual determinations would be the proper course of action.1 On the present sparse record, however, I cannot find support in the Voting Rights Act for the arbitrary process followed by the New York Legislature.

The record is devoid of any evidence that the 65% figure was a reasoned response to the problem of past discrimination.2 It is, rather, clear that under the time pressure of upcoming elections, and "in an atmosphere of hasty dickering," 510 F.2d 512, 525, 526 (CA2 1975) (Frankel, J., dissenting), the New York Legislature simply accepted the standard formula from the Department of Justice and treated it as mandatory. Moreover, the formula appears to be based upon factually unsupportable assumptions. For example, it would make no sense to assure nonwhites a majority of 65% in a voting district unless it were assumed that nonwhites and whites vote in racial blocs, and that the blocs vote adversely to, or independently of, one another. Not only is the record in this case devoid of any evidence that such bloc voting has taken or will take place in Kings County, but such evidence as there is points in the opposite direction: We are informed that four out of the five "safe" (65%+) nonwhite districts established by the 1974 plan have since elected white representatives. Brief for Respondent-Intervenors 48.

The assumption that "whites" and "nonwhites" in the county form homogeneous entities for voting purposes is entirely without foundation. The "whites" category consists of a veritable galaxy of national origins, ethnic backgrounds, and religious denominations. It simply cannot be assumed that the legislative interests of all "whites" are even substantially identical. In similar fashion, those described as "nonwhites" include, in addition to Negroes, a substantial portion of Puerto Ricans. Memorandum of Decision, U. S. Dept. of Justice Nos. V6541-47, July 1, 1974, p. 13 (App. 294).3 The Puerto Rican population, for whose protection the Voting Rights Act was "triggered" in Kings County, see n. 2, supra, has expressly disavowed any identity of interest with the Negroes, and, in fact, objected to the 1974 redistricting scheme because it did not establish a Puerto Rican controlled district within the county.

(3)

Although reference to racial composition of a political unit may, under certain circumstances, serve as "a starting point in the process of shaping a remedy," Swann v. Charlotte-Mecklenburg Bd. of Education, 402 U.S. 1, 25 (1971), rigid adherence to quotas, especially in a case like this, deprives citizens such as petitioners of the opportunity to have the legislature make a determination free from unnecessary bias for or against any racial, ethnic, or religious group. I do not quarrel with the proposition that the New York Legislature may choose to take ethnic or community union into consideration in drawing its district lines. Indeed, petitioners are members of an ethnic community which, without deliberate purpose so far as shown on this record, has long been within a single assembly and senate district. While petitioners certainly have no constitutional right to remain unified within a single political district, they do have, in my view, the constitutional right not to be carved up so as to create a voting bloc composed of some other ethnic or racial group through the kind of racial gerrymandering the Court condemned in Gomillion v. Lightfoot.

If districts have been drawn in a racially biased manner in the past (which the record does not show to have been the case here) the proper remedy is to reapportion along neutral lines. Manipulating the racial composition of electoral districts to assure one minority or another its "deserved" representation will not promote the goal of a racially neutral legislature. On the contrary, such racial gerrymandering puts the imprimatur of the State on the concept that race is a proper consideration in the electoral process. "The vice lies . . . in . . . placing . . . the power of the State behind a racial classification that induces racial prejudice at the polls." Anderson v. Martin, 375 U.S. 399, 402 (1964).

The result reached by the Court today in the name of the Voting Rights Act is ironic. The use of a mathematical formula tends to sustain the existence of ghettos by promoting the notion that political clout is to be gained or maintained by marshaling particular racial, ethnic, or religious groups in enclaves. It suggests to the voter that only a candidate of the same race, religion, or ethnic origin can properly represent that voter's interests, and that such candidate can be elected only from a district with a sufficient minority concentration. The device employed by the State of New York, and endorsed by the Court today, moves us one step farther away from a truly homogeneous society. This retreat from the ideal of the American "melting pot" is curiously out of step with recent political history—and indeed with what the Court has said and done for more than a decade. The notion that Americans vote in firm blocs has been repudiated in the election of minority members as mayors and legislators in numerous American cities and districts overwhelmingly white. Since I cannot square the mechanical racial gerrymandering in this case with the mandate of the Constitution, I respectfully dissent from the affirmance of the judgment of the Court of Appeals.