US SUPREME COURT DECISIONS

III

Relying solely on Vieth, Justice Kennedy maintains that even if legislation is enacted based solely on a desire to harm a politically unpopular minority, this fact is insufficient to establish unconstitutional partisan gerrymandering absent proof that the legislation did in fact burden “the complainants’ representative rights.” Ante, at 11. This conclusion—which clearly goes to the merits, rather than the manageability, of a partisan gerrymandering claim—is not only inconsistent with the constitutional requirement that state action must be supported by a legitimate interest, but also provides an insufficient response to appellants’ claim on the merits.

Justice Kennedy argues that adopting “the modified sole-intent test” could “encourage partisan excess at the outset of the decade, when a legislature redistricts pursuant to its decennial constitutional duty and is then immune from the charge of sole-motivation.” Ante, at 12–13. But this would be a problem of the Court’s own making. As the decision in Cox v. Larios, 542 U. S. 947, demonstrates, there are, in fact, readily manageable judicial standards that would allow injured parties to challenge excessive (and unconstitutional) partisan gerrymandering undertaken in response to the release of the decennial census data. [Footnote 5] See also Vieth, 541 U. S., at 328–339 (Stevens, J., dissenting); id., at 347–353 (Souter, J., joined by Ginsburg, J., dissenting); id., at 365–367 (Breyer, J., dissenting). Justice Kennedy’s concern about a heightened incentive to engage in such excessive partisan gerrymandering would be avoided if the Court were willing to enforce those standards.

In any event, Justice Kennedy’s additional requirement that there be proof that the gerrymander did in fact burden the complainants’ representative rights is clearly satisfied by the record in this litigation. Indeed, the Court’s accurate exposition of the reasons why the changes to District 23 diluted the voting rights of Latinos who remain in that district simultaneously explains why those changes also disadvantaged Democratic voters and thus demonstrates that the effects of a political gerrymander can be evaluated pursuant to judicially manageable standards.

In my judgment the record amply supports the conclusion that Plan 1374C not only burdens the minority party in District 23, but also imposes a severe statewide burden on the ability of Democratic voters and politicians to influence the political process. [Footnote 6].

In arguing that Plan 1374C does not impose an unconstitutional burden on Democratic voters and candidates, the State takes the position that the plan has resulted in an equitable distribution of political power between the State’s two principal political parties. The State emphasizes that in the 2004 elections—held pursuant to Plan 1374C—Republicans won 21 of 32, or 66%, of the congressional seats. That same year, Republicans carried 58% of the vote in statewide elections. Admittedly, these numbers do suggest that the State’s congressional delegation was “roughly proportional” to the parties’ share of the statewide vote, Brief for State Appellees 44, particularly in light of the fact that our electoral system tends to produce a “seat bonus” in which a party that wins a majority of the vote generally wins an even larger majority of the seats, see Brief for Alan Heslop et al. as Amici Curiae (describing the seat bonus phenomenon). Cf. ante, at 12 (opinion of Kennedy, J.) (arguing that, compared to the redistricting plan challenged in Vieth, “Plan 1374C can be seen as making the party balance more congruent to statewide party power”).

That Plan 1374C produced a “roughly proportional” congressional delegation in 2004 does not, however, answer the question whether the plan has a discriminatory effect against Democrats. As appellants point out, whether a districting map is biased against a political party depends upon the bias in the map itself—in other words, it depends upon the opportunities that the map offers each party, regardless of how candidates perform in a given year. And, as the State’s expert found in this litigation, Plan 1374C clearly has a discriminatory effect in terms of the opportunities it offers the two principal political parties in Texas. Indeed, that discriminatory effect is severe.

According to Professor Gaddie, the State’s expert, Plan 1374C gives Republicans an advantage in 22 of 32 congressional seats. The plaintiffs’ expert, Professor Alford, who had been cited favorably by the Balderas Court as having applied a “neutral approach” to redistricting in that litigation, App. to Juris. Statement 207a, agreed. He added that, in his view, the only surprise from the 2004 elections was “how far things moved” toward achieving a 22-to-10 pro-Republican split “in a single election year,” id., at 226a (declaration of John R. Alford, Ph.D.). [Footnote 7] But this 22-to-10 advantage does not depend on Republicans winning the 58% share of the statewide vote that they received in 2004. Instead, according to Professor Gaddie, Republicans would be likely to carry 22 of 32 congressional seats if they won only 52% of the statewide vote. App. 216, 229. Put differently, Plan 1374C ensures that, even if the Democratic Party succeeds in convincing 10% of the people who voted for Republicans in the last statewide elections to vote for Democratic congressional candidates, [Footnote 8] which would constitute a major electoral shift, there is unlikely to be any change in the number of congressional seats that Democrats win. Moreover, Republicans would still have an overwhelming advantage if Democrats achieved full electoral parity. According to Professor Gaddie’s analysis, Republicans would be likely to carry 20 of the 32 congressional seats even if they only won 50% (or, for that matter, 49%) of the statewide vote. Id., at 216, 229–230. This demonstrates that Plan 1374C is inconsistent with the symmetry standard, a measure social scientists use to assess partisan bias, which is undoubtedly “a reliable standard” for measuring a “burden … on the complainants’ representative rights,” ante, at 11 (opinion of Kennedy, J.).

The symmetry standard “requires that the electoral system treat similarly-situated parties equally, so that each receives the same fraction of legislative seats for a particular vote percentage as the other party would receive if it had received the same percentage.” Brief for Gary King et al. as Amici Curiae 4–5. This standard is widely accepted by scholars as providing a measure of partisan fairness in electoral systems. See, e.g., Tufte, The Relationship Between Seats and Votes in Two-Party Systems, 67 Am. Pol. Sci. Rev. 540, 542–543 (1973); Gelman & King, Enhancing Democracy Through Legislative Redistricting, 88 Am. Pol. Sci. Rev. 541, 545 (1994); Thompson, Election Time: Normative Implications of Temporal Properties of the Electoral Process in the United States, 98 Am. Pol. Sci. Rev. 51, 53, and n. 7 (2004); Engstrom & Kernell, Manufactured Responsiveness: The Impact of State Electoral Laws on Unified Party Control of the Presidency and House of Representatives, 1840–1940, 49 Am. J. Pol. Sci. 531, 541 (2005). Like other models that experts use in analyzing vote dilution claims, compliance with the symmetry standard is measured by extrapolating from a sample of known data, see, e.g., Thornburg v. Gingles, 478 U. S. 30, 53 and n. 20 (1986) (discussing extreme case analysis and bivariate ecological regression analysis). In this litigation, the symmetry standard was not simply proposed by an amicus to this Court, it was also used by the expert for plaintiffs and the expert for the State in assessing the degree of partisan bias in Plans 1151C and 1374C. See App. 34–42 (report of Professor Alford); id., at 189–193, 216 (report of Professor Gaddie).

Because, as noted above, Republicans would have an advantage in a significant majority of seats even if the statewide vote were equally distributed between Republicans and Democrats, Plan 1374C constitutes a significant departure from the symmetry standard. By contrast, based on Professor Gaddie’s evaluation, the Balderas Plan, though slightly biased in favor of Republicans, provided markedly more equitable opportunities to Republicans and Democrats. For example, consistent with the symmetry standard, under Plan 1151C the parties were likely to each take 16 congressional seats if they won 50% of the statewide vote. See App. 216.

Plan 1374C then, clearly has a discriminatory impact on the opportunities that Democratic citizens have to elect candidates of their choice. Moreover, this discriminatory effect cannot be dismissed as de minimis. According to the State’s expert, if each party receives half the statewide vote, under Plan 1374C the Republicans would carry 62.5% (20) of the congressional seats, whereas the Democrats would win 37.5% (12) of those seats. In other words, at the vote distribution point where a politically neutral map would result in zero differential in the percentage of seats captured by each party, Plan 1374C is structured to create a 25% differential. When a redistricting map imposes such a significant disadvantage on a politically salient group of voters, the State should shoulder the burden of defending the map. Cf. Brown v. Thomson, 462 U. S. 835, 842–843 (1983) (holding that the implementation of a redistricting plan for state legislative districts with population deviations over 10% creates a prima facie case of discrimination under the Equal Protection Clause, thus shifting the burden to the State to defend the plan); Larios v. Cox, 300 F. Supp. 2d 1320, 1339–1340 (ND Ga.) (per curiam), summarily aff’d, 542 U. S. 947 (2004) (same, but further pointing out that the “ ‘ten percent rule’ ” is not a safe harbor, and concluding that, under the circumstances of the case before it, a state legislative districting plan was unconstitutional even though population deviations were under 10%). At the very least, once plaintiffs have established that the legislature’s sole purpose in adopting a plan was partisan—as plaintiffs have established in this action, see Part II, supra—such a severe discriminatory effect should be sufficient to meet any additional burden they have to demonstrate that the redistricting map accomplishes its discriminatory purpose. [Footnote 9].

The bias in Plan 1374C is most striking with regard to its effect on the ability of Democratic voters to elect candidates of their choice, but its discriminatory effect does not end there. Plan 1374C also lessens the influence Democratic voters are likely to be able to exert over Republican lawmakers, thus further minimizing Democrats’ capacity to play a meaningful role in the political process.

Even though it “defies political reality to suppose that members of a losing party have as much political influence over . . . government as do members of the victorious party,” Davis v. Bandemer, 478 U. S. 109, 170 (1986) (Powell, J., concurring in part and dissenting in part), the Court has recognized that “the power to influence the political process is not limited to winning elections,” id., at 132 (plurality opinion); see also Georgia v. Ashcroft, 539 U. S. 461, 482 (2003). In assessing whether members of a group whose candidate is defeated at the polls can nonetheless influence the elected representative, it is “important to consider ‘the likelihood that candidates elected without decisive minority support would be willing to take the minority’s interests into account.’ ” Id., at 482 (quoting Gingles, 478 U. S., at 100 (O’Connor, J., concurring in judgment)). One justification for majority rule is that elected officials will generally “take the minority’s interests into account,” in part because the majority recognizes that preferences shift and today’s minority could be tomorrow’s majority. See, e.g., L. Guinier, Tyranny of the Majority 77 (1994); J. Ely, Democracy and Distrust 84 (1980); cf. Letter from James Madison to Thomas Jefferson, (Oct. 24, 1787), reprinted in Republic of Letters 502 (J. Smith ed. 1995) (arguing that “[t]he great desideratum in Government is . . . to modify the sovereignty as that it may be sufficiently neutral between different parts of the Society” and thus prevent a fixed majority from oppressing the minority). Indeed, this Court has concluded that our system of representative democracy is premised on the assumption that elected officials will seek to represent their constituency as a whole, rather than any dominant faction within that constituency. See Shaw v. Reno, 509 U. S. 630, 648 (1993).

Plan 1374C undermines this crucial assumption that congressional representatives from the majority party (in this case Republicans) will seek to represent their entire constituency. “When a district obviously is created solely to effectuate the perceived common interests of one racial group, elected officials are more likely to believe that their primary obligation is to represent only the members of that group, rather than their constituency as a whole.” Ibid. Shaw’s analysis of representational harms in the racial gerrymandering context applies with at least as much force in the partisan gerrymandering context because, in addition to the possibility that a representative may believe her job is only to represent the interests of a dominant constituency, a representative may feel more beholden to the cartographers who drew her district than to the constituents who live there. See Vieth, 541 U. S., at 329–331 (Stevens, J., dissenting). In short, Plan 1374C reduces the likelihood that Republican representatives elected from gerrymandered districts will act as vigorous advocates for the needs and interests of Democrats who reside within their districts.

In addition, Plan 1374C further weakens the incentives for members of the majority party to take the interests of the minority party into account, because it locks in a Republican congressional majority of 20–22 seats, so long as Republicans achieve at least 49% of the vote. The result of this lock-in is that, according to the State’s expert, between 19 and 22 of these Republican seats are safe seats, meaning seats where one party has at least a 10% advantage over the other. See App. 227–228 (expert report of Professor Gaddie). Members of Congress elected from such safe districts need not worry much about the possibility of shifting majorities, so they have little reason to be responsive to political minorities within their district. [Footnote 10].

In sum, I think it is clear that Plan 1374C has a severe burden on the capacity of Texas Democrats to influence the political process. Far from representing an example of “one of the most significant acts a State can perform to ensure citizen participation in republican self-governance,” ante, at 9 (opinion of Kennedy, J.), the plan guarantees that the Republican-dominated membership of the Texas congressional delegation will remain constant notwithstanding significant pro-Democratic shifts in public opinion. Moreover, the harms Plan 1374C imposes on Democrats are not “hypothetical” or “counterfactual,” id., at 13, simply because, in the 2004 elections, Republicans won a share of seats roughly proportional to their statewide voting strength. By creating 19–22 safe Republican seats, Plan 1374C has already harmed Democrats because, as explained above, it significantly undermines the likelihood that Republican lawmakers from those districts will be responsive to the interests of their Democratic constituents. In addition, Democrats will surely have a more difficult time recruiting strong candidates, and mobilizing voters and resources, in these safe Republican districts. Thus, appellants have satisfied any requisite obligation to demonstrate that they have been harmed by the adoption of Plan 1374C.

Furthermore, as discussed in Part II, supra, the sole intent motivating the Texas Legislature’s decision to replace Plan 1151C with Plan 1374C was to benefit Republicans and burden Democrats. Accordingly, in terms of both its intent and effect, Plan 1374C violates the sovereign’s duty to govern impartially.

“When a State adopts rules governing its election machinery or defining electoral boundaries, those rules must serve the interests of the entire community. If they serve no purpose other than to favor one segment—whether racial, ethnic, religious, or political—that may occupy a position of strength at a particular point in time, or to disadvantage a politically weak segment of the community, they violate the constitutional guarantee of equal protection.” Karcher, 462 U. S., at 748 (Stevens, J., concurring) (citation omitted).

Accordingly, even accepting the Court’s view that a gerrymander is tolerable unless it in fact burdens the minority’s representative rights, I would hold that Plan 1374C is unconstitutional. [Footnote 11]

IV

Even if I thought that Plan 1374C were not unconstitutional in its entirety, I would hold that the cracking of District 24—which, under the Balderas Plan, was a majority-minority district that consistently elected Democratic Congressman Martin Frost—was unconstitutional. Readily manageable standards enable us to analyze both the purpose and the effect of the “granular” decisions that produced the replacements for District 24. Applying these standards, which I set forth below, I believe it is clear that the manipulation of this district for purely partisan gain violated the First and Fourteenth Amendments.

The same constitutional principles discussed above concerning the sovereign’s duty to govern impartially inform the proper analysis for claims that a particular district is an unconstitutional partisan gerrymander. We have on several occasions recognized that a multimember district is subject to challenge under the Fourteenth Amendment if it operates “ ‘to minimize or cancel out the voting strength of racial or political elements of the voting population.’ ” E.g., Gaffney v. Cummings, 412 U. S. 735, 751 (1973) (emphasis added); Burns v. Richardson, 384 U. S. 73, 88 (1966). There is no constitutionally relevant distinction between the harms inflicted by single-member district gerrymanders that minimize or cancel out the voting strength of a political element of the population and the same harms inflicted by multimember districts. In both situations, the State has interfered with the voter’s constitutional right to “engage in association for the advancement of beliefs and ideas,” NAACP v. Alabama ex rel. Patterson, 357 U. S. 449, 460 (1958).

I recognize that legislatures will always be aware of politics and that we must tolerate some consideration of political goals in the redistricting process. See Cousins v. City Council of Chicago, 466 F. 2d 830, 847 (CA7 1972) (Stevens, J., dissenting). However, I think it is equally clear that, when a plaintiff can prove that a legislature’s predominant motive in drawing a particular district was to disadvantage a politically salient group, and that the decision has the intended effect, the plaintiff’s constitutional rights have been violated. See id., at 859–860. Indeed, in Vieth, five Members of this Court explicitly recognized that extreme partisan gerrymandering violates the Constitution. See 541 U. S., at 307, 312–316 (Kennedy, J., concurring in judgment); id., at 317–318 (Stevens, J., dissenting); id., at 343, 347–352 (Souter, J., joined by Ginsburg, J., dissenting); id., at 356–357, 366–367 (Breyer, J., dissenting). The other four Justices in Vieth stated that they did not disagree with that conclusion. See id., at 292 (plurality opinion). The Vieth plurality nonetheless determined that there were no judicially manageable standards to assess partisan gerrymandering claims. Id., at 305–306. However, the following test, which shares some features of the burden-shifting standard for assessing unconstitutional partisan gerrymandering proposed by Justice Souter’s opinion in Vieth, see id., at 348–351, would provide a remedy for at least the most blatant unconstitutional partisan gerrymanders and would also be eminently manageable.

First, to have standing to challenge a district as an unconstitutional partisan gerrymander, a plaintiff would have to prove that he is either a candidate or a voter who resided in a district that was changed by a new districting plan. See id., at 327–328 (Stevens, J., dissenting) (discussing United States v. Hays, 515 U. S. 737 (1995)). See also 541 U. S., at 347–348 (Souter, J., joined by Ginsburg, J., dissenting) (citing Hays). A plaintiff with standing would then be required to prove both improper purpose and effect.

With respect to the “purpose” portion of the inquiry, I would apply the standard fashioned by the Court in its racial gerrymandering cases. Under the Court’s racial gerrymandering jurisprudence, judges must analyze whether plaintiffs have proved that race was the predominant factor motivating a districting decision such that other, race-neutral districting principles were subordinated to racial considerations. If so, strict scrutiny applies, see, e.g., Vera, 517 U. S., at 958–959 (plurality opinion), and the State must justify its districting decision by establishing that it was narrowly tailored to serve a compelling state interest, such as compliance with §2 of the Voting Rights Act, see King v. Illinois Bd. of Elections, 979 F. Supp. 619 (ND Ill. 1997), aff’d, 522 U. S. 1087 (1998); Vera, 517 U. S., at 994 (O’Connor, J., concurring). [Footnote 12] However, strict scrutiny does not apply merely because race was one motivating factor behind the drawing of a majority-minority district. Id., at 958–959 (plurality opinion); see also Easley v. Cromartie, 532 U. S. 234, 241 (2001). Applying these standards to the political gerrymandering context, I would hold that, if a plaintiff carried her burden of demonstrating that redistricters subordinated neutral districting principles to political considerations and that their predominant motive was to maximize one party’s power, she would satisfy the intent prong of the constitutional inquiry. [Footnote 13] Cf. Vieth, 541 U. S., at 349–350 (Souter, J., joined by Ginsburg, J., dissenting) (discussing the importance of a district’s departures from traditional districting principles in determining whether the district is an unconstitutional gerrymander).

With respect to the effects inquiry, a plaintiff would be required to demonstrate the following three facts: (1) her candidate of choice won election under the old plan; (2) her residence is now in a district that is a safe seat for the opposite party; and (3) her new district is less compact than the old district. The first two prongs of this effects inquiry would be designed to measure whether or not the plaintiff has been harmed, whereas the third prong would be relevant because the shape of the gerrymander has always provided crucial evidence of its character, see Karcher, 462 U. S., at 754–758, 762–763 (Stevens, J., concurring); see also Vieth, 541 U. S., at 348 (Souter, J., joined by Ginsburg, J., dissenting) (noting that compactness is a traditional districting principle, which “can be measured quantitatively”). Moreover, a safe harbor for more compact districts would allow a newly elected majority to eliminate a prior partisan gerrymander without fear of liability or even the need to devote resources to litigating whether or not the legislature had acted with an impermissible intent.

If a plaintiff with standing could meet the intent and effects prong of the test outlined above, that plaintiff would clearly have demonstrated a violation of her constitutional rights. Moreover, I do not think there can be any colorable claim that this test would not be judicially manageable.

Applying this test to the facts of this case, I think plaintiffs in new Districts 6, 24, 26, and 32—four of the districts in Plan 1374C that replaced parts of Balderas District 24—can demonstrate that their constitutional rights were violated by the cracking of Balderas District 24. First, I assume that there are plaintiffs who reside in Districts 6, 24, 26, and 32, and whose homes were previously located in Balderas District 24. [Footnote 14] Accordingly, I assume that there are plaintiffs who have standing to challenge the creation of these districts.

Second, plaintiffs could easily satisfy their burden of proving predominant partisan purpose. Indeed, in this litigation, the State has acknowledged that its predominant motivation for cracking District 24 was to achieve partisan gain. See State Post-Trial Brief 51–52 (noting that, in spite of concerns that the cracking of District 24 could lead to Voting Rights Act liability, “[t]he Legislature … chose to pursue a political goal of unseating Congressman Frost instead of following a course that might have lowered risks [of such liability]”).

The District Court agreed with the State’s analysis on this issue. In the District Court, plaintiffs claimed that the creation of District 26 violated the Equal Protection Clause because the decision to create District 26 was motivated by unconstitutional racial discrimination against black voters. The District Court rejected this argument, concluding that the State’s decision to crack Balderas District 24 was driven not by racial prejudice, but rather by the political desire to maximize Republican advantage and to “remove Congressman Frost,” which required that Frost “lose a large portion of his Democratic constituency, many of whom lived in a predominately Black area of Tarrant County.” Session, 298 F. Supp. 2d, at 471.

That an impermissible, predominantly partisan, purpose motivated the cracking of former District 24 is further demonstrated by the fact that, in my judgment, this cracking caused Plan 1374C to violate §5 of the Voting Rights Act, 42 U. S. C. §1973c. The State’s willingness to adopt a plan that violated its legal obligations under the Voting Rights Act, combined with the other indicia of partisan intent in this litigation, is compelling evidence that politics was not simply one factor in the cracking of District 24, but rather that it was an impermissible, predominant factor.

Section 5 of the Voting Rights Act “was intended ‘to insure that that [the gains thus far achieved in minority political participation] shall not be destroyed through new [discriminatory] procedures and techniques.’ ” Beer v. United States, 425 U. S. 130, 140–141 (1976) (quoting S. Rep. No. 94–295, p. 19 (1975) (alteration in Beer)). To effectuate this goal, §5 prevents covered jurisdictions, such as Texas, from making changes to their voting procedures “that would lead to a retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise.” Georgia, 539 U. S., at 477 (internal quotation marks and citations omitted). In other words, during the redistricting process, covered jurisdictions may not “leave minority voters with less chance to be effective in electing preferred candidates than they were” under the prior districting plan. See id., at 494 (Souter, J., dissenting). By cracking Balderas District 24, and by not offsetting the loss in black voters’ ability to elect preferred candi- dates elsewhere, Plan 1374C resulted in impermissible retrogression.

Under the Balderas Plan, black Americans constituted a majority of Democratic primary voters in District 24. According to the unanimous report authored by staff attorneys in the Voting Section of the Department of Justice, black voters in District 24 generally voted cohesively, and thus had the ability to elect their candidate of choice in the Democratic primary. Section 5 Recommendation Memorandum 33 (Dec. 12, 2003), available at http://www.washingtonpost.com/wp-srv/nation/documents/ texasDOJmemo.pdf (all Internet materials as visited June 21, 2006, and available in Clerk of Court’s case file). Moreover, the black community’s candidates of choice could consistently attract sufficient crossover voting from nonblacks to win the general election, even though blacks did not constitute a majority of voters in the general election. Id., at 33–34. Representative Frost, who is white, was clearly the candidate of choice of the black community in District 24, based on election returns, testimony of community leaders, and “scorecards” he received from groups dedicated to advancing the interests of African-Americans. See id., at 35.

As noted above, in Plan 1374C, “the minority community in [Balderas District] 24 [was] splintered and submerged into majority Anglo districts in the Dallas-Fort Worth area.” Id., at 67. By dismantling one district where blacks had the ability to elect candidates of their choice, [Footnote 15] and by not offsetting this loss of a district with another district where black voters had a similar opportunity, Plan 1374C was retrogressive, in violation of §5 of the Voting Rights Act. See id., at 31, 67–69.

Notwithstanding the unanimous opinion of the staff attorneys in the Voting Section of the Justice Department that Plan 1374C was retrogressive and that the Attorney General should have interposed an objection, the Attorney General elected to preclear the map, thus allowing it to take effect. We have held that, under the statutory scheme, voters may not directly challenge the Attorney General’s decision to preclear a redistricting plan, see Morris v. Gressette, 432 U. S. 491 (1977), which means that the Attorney General’s vigilant enforcement of the Act is critical, and which also means that plaintiffs could not bring a §5 challenge as part of this litigation. [Footnote 16] However, judges are frequently called upon to consider whether a redistricting plan violates §5, because a covered jurisdiction has the option of seeking to achieve preclearance by either submitting its plan to the Attorney General or filing a declaratory judgment action in the District Court for the District of Columbia, whose judgment is subject to review by this Court, see, e.g., Georgia, 539 U. S. 461. Accordingly, we have the tools to analyze whether a redistricting plan is retrogressive.

Even though the §5 issue is not directly before this Court, for the reasons stated above, I believe that the cracking of District 24 caused Plan 1374C to be retrogressive. And the fact that the legislature promulgated a retrogressive plan is relevant, because it provides additional evidence that the legislature acted with a predominantly partisan purpose. Complying with §5 is a neutral districting principle, and the legislature’s promulgation of a retrogressive redistricting plan buttresses my conclusion that the “legislature subordinated traditional [politically-]neutral districting principles . . . to [political] considerations.” Miller v. Johnson, 515 U. S. 900, 916 (1995). This evidence is particularly compelling in light of the State’s acknowledgment that “[t]he Legislature … chose to pursue a political goal of unseating Congressman Frost instead of following a course that might have lowered risks in the preclearance process.” State Post-Trial Brief 52 (citing, inter alia, trial testimony of state legislators).

In sum, the record in this litigation makes clear that the predominant motive underlying the fragmentation of Balderas District 24 was to maximize Republicans’ electoral opportunities and ensure that Congressman Frost was defeated.

Turning now to the effects test I have proposed, plaintiffs in new Districts 6, 24, 36, and 32 could easily meet the three parts of that test because: (1) under the Balderas plan, they lived in District 24 and their candidate of choice (Frost) was the winning candidate; (2) under Plan 1374C, they have been placed in districts that are safe seats for the Republican party, see App. 106 (showing that the Democratic share of the two-party vote in statewide elections from 1996 to 2002 was 40% or less in Districts 6, 24, 26, and 32); and (3) their new districts are less compact than Balderas District 24, see App. 319–320 (compactness scores for districts under the Balderas Plan and Plan 1374C). [Footnote 17].

Justice Kennedy rejects my proposed effects test, as applied in this case, because, in his view Balderas District 24 lacks “any special claim to fairness,” ante, at 36. But my analysis in no way depends on the proposition that Balderas District 24 was fair. The district was more compact than four of the districts that replaced it, and, as explained above, compactness serves important values in the districting process. This is why, in my view, a State that creates more compact districts should enjoy a safe harbor from partisan gerrymandering claims. However, the mere fact that a prior district was unfair should surely not provide a safe harbor for the creation of an even more unfair district. Conversely, a State may of course create less compact districts without violating the Constitution so long as its purpose is not to disadvantage a politically disfavored group. See supra, at 31–32 and n. 13. The reason I focus on Balderas District 24 is not because the district was fair, but because the prior district provides a clear benchmark in analyzing whether plaintiffs have been harmed.

In sum, applying the judicially manageable test set forth in this Part of my opinion reveals that the cracking of Balderas District 24 created several unconstitutional partisan gerrymanders. Even if I believed that Plan 1374C were not invalid in its entirety, I would reverse the judgment below with regard to Districts 6, 24, 26, and 32.

*  *  *

For the foregoing reasons, although I concur with the majority’s decision to invalidate District 23 under §2 of the Voting Rights Act, I respectfully dissent from the Court’s decision to affirm the judgment below with respect to plaintiffs’ partisan gerrymandering claim. I would reverse with respect to the plan as a whole, and also, more specifically, with respect to Districts 6, 24, 26, and 32.