Coalition for Economic Equity v. Wilson

SCHROEDER, Circuit Judge, with whom PREGERSON, NORRIS and TASHIMA, Circuit Judges, join, dissenting from denial of rehearing en bane.

I respectfully dissent from the court’s decision not to review this case en banc. En banc review was warranted in this case for two reasons.

First, the case is extraordinarily important. In upholding the constitutionality of California Proposition 209, which forbids the state to offer any preferential treatment in employment, education, or contracting on the basis of race or gender, the case has put equal protection law in a state of turmoil. At stake is whether the voters may bar the state’s institutions from using narrowly tailored race- or gender-based remedies which serve compelling state interests, including redressing the effects of past racial and gender-based discrimination by those very institutions.

Second, the decision is contrary to controlling Supreme Court precedent. Proposition 209 puts only race- or gender-based remedies for past discrimination, and no other state preferences, beyond the reach of the legislative or any other process, save another state initiative. The Supreme Court has squarely held that a state violates the Constitution when it attempts to put legislative remedies which benefit minorities at a remote level of government beyond the ordinary legislative process. See Washington v. Seattle School District, No. 1, 458 U.S. 457, 102 S.Ct. 3187, 73 L.Ed.2d 896 (1982); Hunter v. Erickson, 393 U.S. 385, 89 S.Ct. 557, 21 L.Ed.2d 616 (1969).

*712The panel opinion, Coalition for Economic Equity v. Wilson, 110 F.3d 1431 (9th Cir.1997) (“CEE II”), treats the case as if the state were merely repealing past remedies that it had determined were not constitutionally required. See id. at 1443 (citing Crawford v. Board of Educ., 458 U.S. 527, 538, 102 S.Ct. 3211, 3218, 73 L.Ed.2d 948 (1982)). Proposition 209, however, is not merely a repeal of past remedies; it is a bar to future remedies to curb injustice. See Coalition for Economic Equity v. Wilson, 946 F.Supp. 1480, 1508 (N.D.Cal.1996) (“CEE I”) (“Proposition 209, by its terms, not only repeals all existing state and local affirmative action programs, but also prohibits the adoption of such programs in the future. In so doing, Proposition 209 displaces authority with respect to a race and gender issue to ‘a new and remote level of government,’ and thus reorders the political process to the detriment of women and minorities.”) (quoting Seattle Sch. Dist., 458 U.S. at 483, 102 S.Ct. at 3202).

The panel is also incorrect when it characterizes the ordinance as “adopting the Equal Protection Clause.” CEE II, 110 F.3d at 1446 (quoting Crawford, 458 U.S. at 535, 102 S.Ct. at 3217). Unlike the amendment in Crawford, Prop. 209 does not merely conform California law to the federal constitution. See Crawford, 458 U.S. at 535, 102 S.Ct. at 3217. Rather, it is specifically designed to bar remedies that are permissible under the Equal Protection Clause. See CEE I, 946 F.Supp. at 1489. Otherwise, the Proposition would be a nullity. See id.

The concept of judicial review is a constitutional check on the excesses of majority rule. See Marbury v. Madison, 5 U.S. (1 Cranch) 137, 175-77, 2 L.Ed. 60 (1803); The Federalist No. 78 (Hamilton), at 466-70 (Clinton Rossiter ed., 1961). The responsible exercise of judicial review requires courage and vigilance. I would affirm District Judge Henderson’s fine opinion in CEE I.

NORRIS, Circuit Judge, with whom SCHROEDER, PREGERSON and TASHIMA, Circuit Judges, join, respecting the denial of rehearing en banc.

In Hunter v. Erickson, 393 U.S. 385, 89 S.Ct. 557, 21 L.Ed.2d 616 (1969), Justice John Harlan stated a constitutional principle for the protection of racial minorities from discriminatory treatment within the political process. Justice Harlan wrote that “a provision that has the clear purpose of making it more difficult for racial ... minorities to achieve legislation that is in their interest ... is discriminatory on its face.” Id. at 395, 89 S.Ct. at 563 (Harlan, J., concurring). In Washington v. Seattle School District No. 1, 458 U.S. 457, 102 S.Ct. 3187, 73 L.Ed.2d 896 (1982), the full Court adopted Justice Harlan’s reasoning, holding that whenever the state “differentiates between the treatment of problems involving racial matters and that afforded other problems in the same area [of government],” it has utilized an impermissible “racial classification” in restructuring the political process. Id. at 470, 480, 102 S.Ct. at 3195, 3200 (citing Justice Harlan’s Hunter concurrence). It is the core holding of these cases that the state may not “place special burdens on the ability of minority groups to achieve beneficial legislation ... by lodging decisionmaking authority over the question at a new and remote level of government.” Id. at 467, 483, 102 S.Ct. at 3193, 3202. Because that is precisely what Proposition 209 does, our court has no legitimate choice but to declare it unconstitutional. Unless and until the Supreme Court overrules Hunter and Seattle, California simply does not have the constitutional authority to place minority groups at a disadvantage in the political process.

Proposition 209 strips the state legislature and all political subdivisions such as city councils, county boards of supervisors, local school boards, and the Board of Regents of the University of California, of all authority to adopt racial preferences in the future. It “lodg[es] decisionmaking authority over [affirmative action] at a new and remote level of government” — the entire electorate of California. The measure thereby deprives the proponents of affirmative action — and only the proponents of affirmative action — of the ordinary benefits of representative government. The AARP can go to a city council tomorrow and request special incentives to attract older Americans to the municipal *713workforce; the VFW can lobby the state legislature for special employment or housing incentives on behalf of veterans; Marian Wright Edelman can lobby for preferential health benefits for children at all levels of state and local government. Advocates of racial preferences, in contrast, may no longer appeal to a city council or the Board of Regents, or even to the state legislature. They must take their case directly to the voters of California.

It is hard to imagine a more onerous burden in the political process than mounting a statewide initiative campaign. As the district court recognized, “substantial funds are required to organize and fund the statewide campaign” that is necessary to get an initiative passed. See Coalition for Economic Equity v. Wilson, 946 F.Supp. 1480, 1498-99 (N.D.Cal.1996). California has a population of some thirty-two million people. The cost of delivering a message to that entire electorate is staggering. The Proposition 209 campaign itself cost over three million dollars, with most of the funds going to “television, radio, print advertising, and direct mail.” Id. Proposition 209 now places that obstruction before every effort to enact affirmative action programs. It effectively shuts off all debate over racial preferences in the halls of representative government, depriving minorities even of the opportunity to persuade their elected and appointed officials that affirmative action programs make sense as a matter of social policy when they are narrowly tailored to achieve compelling remedial objectives.1

The Coalition panel does not dispute that affirmative action programs confer benefits primarily upon minorities, nor does it dispute that Proposition 209 places special obstacles in the way of those programs. “We accept without question,” the panel writes, “the district court’s findings that Proposition 209 burdens members of insular minorities within the majority that enacted it who otherwise would seek to obtain race-based and gender-based preferential treatment from local entities.” Coalition for Economic Equity v. Wilson [CEE v. Wilson], 110 F.3d 1431, 1442 (9th Cir.1997). Notwithstanding its recognition that Proposition 209 places special burdens on minority groups in the political process, the panel holds that the measure is constitutional. It reaches this result, I submit, in violation of its duty to follow controlling Supreme Court precedent.

I

The panel justifies its result by drawing a distinction between affirmative action programs and antidiscrimination laws that has no place in Hunter-Seattle analysis. It rationalizes this by characterizing all affirmative action programs as inherently discriminatory — even those that pass constitutional muster under the Fourteenth Amendment. The Equal Protection Clause is aimed at securing equality, the panel complains, and so affirmative action programs are not worthy of protection under the Hunter-Seattle doctrine because — in the panel’s opinion— they do not secure “equal treatment.” CEE v. Wilson, 110 F.3d at 1445. The panel uses its distinction to argue that Hunter and Seattle are categorically inapplicable to race-based preferences, and, in doing so, ignores the Supreme Court’s clear mandate that Hunter and Seattle shall apply to all constitutional legislation that “inures primarily to the benefit of the minority.” Seattle, 458 U.S. at 472, 102 S.Ct. at 3196.

In resting its decision on the view that affirmative action programs do not secure “equality,” the panel injects into Hunter-Seattle analysis a test that looks to the personal views of individual judges about the relative merits of affirmative action programs and antidiscrimination laws. There is absolutely no reason to believe that the Su*714preme Court intended any such result. The relevant inquiry under Hunter and Seattle is simply to ask whether legislation is beneficial to minorities. See Hunter, 393 U.S. at 392-93, 89 S.Ct. at 561-62; id. at 395, 89 S.Ct. at 563 (Harlan, J., concurring); Seattle, 458 U.S. at 472, 102 S.Ct. at 3196. The Court used language that is broad and unqualified. Nowhere has the Court suggested that the Hunter-Seattle doctrine permits judges to rely upon their own subjective impressions as to whether a particular measure aimed at benefiting minorities is also an effective means of securing equality, or whether the social costs associated with that measure are worth the potential benefits.

Indeed, the panel’s distinction between antidiscrimination laws and race-based remedial programs is squarely rejected by Seattle. There, the Court followed Hunter in striking down a Washington statute that forbade local school districts from using busing to remedy de facto school segregation. Of course, the very busing program that gave rise to the case was itself a controversial, race-based remedial program. Under the Seattle busing plan, students were assigned to schools — and bussed accordingly — on the basis of their race. See Seattle, 458 U.S. at 461, 473-74, 102 S.Ct. at 3190, 3196-97. The Seattle Court held that this race-based remedial program was entitled to the same constitutional protection that the Hunter Court afforded to antidiscrimination laws. Both types of law were “beneficial” to minorities, and both held a common goal: to “overcome the ‘special condition’ of prejudice” and remedy the injustices that have resulted from slavery and its legacy, racial discrimination. See Seattle, 458 U.S. at 486, 102 S.Ct. at 3203 (quoting U.S. v. Carolene Prods., 304 U.S. 144, 153 n. 4, 58 S.Ct. 778, 783 n. 4, 82 L.Ed. 1234 (1938)). The Court further held that the political controversy that enveloped the Seattle busing plan was not a proper factor for a court to consider in rendering a constitutional decision. Rather, it instructed, “in the absence of a constitutional violation, the desirability and efficacy of school desegregation are matters to be resolved through the political process. For present purposes, it is enough that minorities may consider busing for integration to be ‘legislation that is in their interest.’ ” Seattle, 458 U.S. at 474, 102 S.Ct. at 3197 (quoting Justice Harlan’s concurrence in Hunter) (emphasis added).

The panel’s distinction, in contrast, has the effect of transforming constitutional analysis into an arena for political debate over the “desirability and efficacy” of affirmative action. The panel “distinguishes” affirmative action programs, even though the panel agrees that such programs constitute legislation that is beneficial to minorities, because it thinks that affirmative action does not in fact promote “equality.” It deploys the full rhetorical arsenal developed by the opponents of affirmative action, characterizing such programs as “inherently invidious” instances of reverse discrimination, CEE v. Wilson, 110 F.3d at 1445 n. 16, and proclaiming that all affirmative action programs constitute disfavored legislation, whether or not they are constitutional. See id. at 1445. The proponents of affirmative action, responding on the level of social policy, would no doubt argue that such programs do in fact secure equality because they level the playing field by remedying the inequalities that are the product of the long history of state-sponsored discrimination that followed the Civil War. This is the rhetorical framework within which the debate over affirmative action has unfolded in the halls of American government. As a political manifesto, the panel’s opinion could comfortably assume a place on one side of that debate. As a document that purports to apply the Constitution as interpreted by the Supreme Court, however, the opinion fails at its appointed task.

The question of the constitutionality of Proposition 209 has nothing to do with the wisdom of affirmative action programs. The Hunter-Seattle doctrine does not call for judges to determine whether constitutionally permissible, beneficial programs are also wise and just means of securing equality. Rather, the “equality” that the Hunter-Seattle doctrine secures is the equality of opportunity that the Constitution guarantees to minority groups to use the channels of representative government to “achieve legislation that is in their interest.” Hunter, 393 U.S. at 395, 89 S.Ct. at 563 (Harlan, J., concurring).

*715II

The panel offers a number of other, subsidiary arguments in its attempt to distinguish Hunter and Seattle. These arguments are no more convincing than the panel’s attempt to inject a subjective “equality” test into the Hunter-Seattle doctrine.

A

The panel first appeals to the commonsense notion that legislatures should not be “constitutionally required ” by courts to pursue any particular legislative policy or program.

To hold that a democratically enacted affirmative action program is constitutionally permissible because the people have demonstrated a compelling state interest is hardly to hold that the program is constitutionally required. The Fourteenth Amendment, lest we lose sight of the forest for the trees, does not require what it barely permits.

CEE v. Wilson, 110 F.3d at 1446.

As the panel must have understood, the district court did not hold that any particular legislation, including race-based preference programs, was “constitutionally required.” The Hunter-Seattle doctrine does not ask whether any particular legislation is ever constitutionally required. The Supreme Court did not hold that either the antidiscrimination laws in Hunter or the remedial busing program in Seattle was constitutionally required. What the Court did hold is that states may not place political obstacles in the way of laws that favor minorities, nor may they remove such laws wholesale from the political process. All that is constitutionally required is that minorities have the opportunity, on equal terms, to seek “legislation in [their] behalf’ within existing channels of government. Hunter, 393 U.S. at 392-93, 89 S.Ct. at 561-62.

B

The panel next suggests that Proposition 209 does not implicate the Hunter-Seattle doctrine because, in fact, it does not selectively restructure the political process. The panel argues that by forbidding race-related programs in employment, education, and contracting at every level of representative government — rather than doing so at only one level, as in Seattle — the State has “addressed] in neutral fashion race-related and gender-related matters” and so has not offended Hunter or Seattle. CEE v. Wilson, 110 F.3d at 1444.

Once again, the panel makes an argument that finds no support in law or reason. The Hunter-Seattle doctrine holds that it works an injury upon minorities to subject them to a -structural disadvantage in the political process. It therefore forbids the State from “differentiating] between the treatment of problems involving racial matters and that afforded other problems in the same area.” Seattle, 458 U.S. at 480, 102 S.Ct. at 3200 (citations omitted). In Seattle, as the panel correctly observes, the State of Washington had inflicted this injury upon minorities at only one level of government: local school districts. What the panel suggests is that if the State inflicts a Hunter-Seattle injury at every level of representative government and “differentiates between the treatment of problems involving racial matters and that afforded other problems” in local school boards and city councils and the state legislature and state agencies such as the University of California Board of Regents, then the constitutional error is somehow cured. See CEE v. Wilson, 110 F.3d at 1443-44. Neither Hunter nor Seattle — nor common sense, for that matter — supports the proposition that expanding the levels at which the State disadvantages minorities will render that action any less constitutionally suspect.2

*716C

Next the panel would have us believe that Hunter and Seattle are inapposite because Proposition 209 burdens majority, not minority, interests.3 In putting forward this remarkable argument, the panel seizes upon a statement in Hunter that “[t]he majority needs no protection against discrimination.” Hunter, 393 U.S. at 391, 89 S.Ct. at 560; CEE v. Wilson, 110 F.3d at 1441. It then puts its own spin on the word “majority” by turning women and racial minorities into one undifferentiated “group.” See id. at 1441-43. In other words, the panel transforms racial minorities into a numerical majority by lumping them together with women. Presto! Proposition 209 burdens the interests of the majority, not the minority. To borrow the panel’s (uneited) paraphrase of Justice Scalia, “if merely stating this alleged ‘equal protection’ [argument] does not suffice to refute it, our constitutional jurisprudence” has gone very far astray. See id. at 1439 (quoting but not citing Romer v. Evans, — U.S. -, -, 116 S.Ct. 1620, 1630, 134 L.Ed.2d 855 (1996) (Scalia, J., dissenting)).4

D

Finally, the panel turns to Crawford v. Board of Education of the City of Los Angeles, 458 U.S. 527, 102 S.Ct. 3211, 73 L.Ed.2d 948 (1982), in its attempt to distinguish Hunter and Seattle. See CEE v. Wilson, 110 F.3d at 1443, 1445-46. For the same reason that the Hunter-Seattle doctrine controls this case, however, Crawford has no bearing upon the constitutionality of Proposition 209.

Crawford’s focus is on the difference between the “mere repeal” of existing, minority-oriented legislation and the systematic obstruction of such legislation in the future. It holds that a “mere repeal,” unlike systematic obstruction, does not violate equal protection. In Crawford, the people of California amended their Constitution to do away with a substantive right that the state’s Equal Protection Clause had conferred above and beyond what was required by the Federal Constitution: the right to be free from de facto segregation in the public schools. The new amendment effectively repealed this right by rendering it unenforceable in state court. See id. at 529-35, 102 S.Ct. at 3213-17. In rejecting a Hunter-Seattle challenge to that amendment, the Supreme Court held that “the Equal Protection Clause is not violated by the mere repeal of race-related legislation or policies.” Id. at 538, 102 S.Ct. at 3218. Thus, Crawford permits the states to repeal any existing affirmative action programs, even though the repeal might operate to the detriment of minorities, so long as they do not prospectively,. disadvantage minorities in the political process.

Proposition 209, of course, is not a mere repeal of existing affirmative action programs. Had the initiative simply repealed all such programs throughout the state and left minority groups free to lobby the institutions of representative government for their reenactment, then Crawford might well control this case. Proposition 209 does much more, however: It forbids all institutions of representative government from enacting race- or gender-based affirmative action programs — and only those programs— in the future. A measure that places permanent political obstacles in the way of minority-oriented legislation violates the Hunter-Seattle doctrine and cannot fit itself within the “mere repeal” doctrine of Crawford.

Ill

When a federal court disregards Supreme Court precedent, it jeopardizes the founda*717tion upon which the entire judiciary rests in our constitutional system of government. The power of the federal courts “lies ... in [their] legitimacy, a product of substance and perception that shows itself in the people’s acceptance of the Judiciary as fit to determine what the Nation’s law means and to declare what it demands.” Planned Parenthood of Southeastern Pa. v. Casey, 505 U.S. 833, 865, 112 S.Ct. 2791, 2814, 120 L.Ed.2d 674 (1992) (Joint Opinion of O’Connor, Kennedy, and Souter, JJ.). Faithful adherence to precedent does not always come easily. “Some cost will be paid by anyone who approves or implements a constitutional decision where it is unpopular, or who refuses to work to undermine the decision or to force its reversal.” Id. at 867, 112 S.Ct. at 2815. Even the broad mantle of protection that Article III offers cannot completely shield federal judges from these heavy costs. We must sometimes implement precedent that comes into conflict with our most deeply held personal convictions. In those instances, for judges as much as anyone else, “[a]n extra price will be paid by those who themselves disapprove of the decision’s results when viewed outside of constitutional terms.” Id. at 867-68, 112 S.Ct. at 2815. I would never minimize or belittle the burden that judges must bear when they are presented with a fundamental issue of conscience and called upon by binding Supreme Court precedent to render a decision whose substantive result violates their deeply held convictions.

It is the responsibility of all federal judges, however, to “struggle to accept [that burden].” Id. at 868, 112 S.Ct. at 2815. We must do so “because [we] respect the rule of law.” Id. It is the particular responsibility of the judges of lower federal courts, moreover, to abide by the precedent that binds us without regard to changes in the political climate of the Nation — or even changes in the sentiment of the Supreme Court — that we think might cast a favorable eye upon a different result. “If a precedent of [the Supreme] Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to [the Supreme] Court the prerogative of overruling its own decisions.” Rodriguez de Quijas v. Shearson/American Express, 490 U.S. 477, 484, 109 S.Ct. 1917, 1921-22, 104 L.Ed.2d 526 (1989). The federal courts of appeals are neither final nor infallible,5 and our duty is more constrained than we might sometimes wish it to be.

I submit, with all due respect, that the Coalition panel has neglected this duty in favor of a path of conservative judicial activism. Its failure to hold Proposition 209 unconstitutional on the authority of Hunter and Seattle is unsupportable as a matter of law. I deeply regret that a majority of our active judges — some, perhaps, assuming that the Supreme Court will correct the panel’s error — has failed to vote to rehear this case en banc.

HAWKINS, Circuit Judge,

commenting on the denial of rehearing en banc.

The panel was faced with the task of filtering a citizen initiative through its understanding and interpretation of the Constitution. Apparently because the panel believed there was no recent authority directly on point, it chose to do what courts often must do: interpret and apply existing authority to the legal challenge presented. At the end of the day, that is the same task in which the district court was engaged. I have no doubt that both the district court and my colleagues on the panel acted in the utmost good faith in doing so. But therein lies an interesting and important question of jurisprudence. It has nothing to do with the merits of affirmative action; it has everything to do with the way in which courts subject to review by higher authority carry out their duties. It is about the proper role of an inferior court faced with contrary, but apparently controlling, precedent that it honestly and earnestly believes will not be followed (or will be distinguished) by higher authority. Is it the role (the duty, if you will) of a court in that circumstance to attempt to accurately predict what the higher authority will do? Or is its *718duty to faithfully follow existing precedent? If an inferior court in such a circumstance is free to predict what the higher court will do, then this panel probably got the issue right.1 If, however, the duty of a lower court is to faithfully apply existing authority, then I have seen no persuasive argument that contradicts Judge Norris’ Carotene Products-Hunter-Seattle analysis.2

The Supreme Court may well tell us that this case is not governed by that line' of authority or that an exception to the application of those cases should be made for this particular measure. Until that happens, it is not our role to predict — however accurate our predictions might turn out to be.3

. As the Supreme Court has made clear, a race-based preference program may be a constitutional means of remedying discrimination.

The unhappy persistence of both the practice and the lingering effects of racial discrimination against minority groups in this country is an unfortunate reality, and the government is not disqualified from acting in response to it.... When race-based action is necessary to further a compelling interest, such action is within constitutional constraints if it satisfies the 'narrow tailoring' test this Court has set out in previous cases.

Adarand Constructors v. Pena, 515 U.S. 200, 237, 115 S.Ct. 2097, 2118, 132 L.Ed.2d 158 (1995).

. In making this extraordinary argument, the panel appears to rely exclusively upon language that it quotes out of context from a footnote in Seattle that discusses the distribution of authority between state and local governments in Washington. See Seattle, 458 U.S. at 479 n. 22, 102 S.Ct. at 3200 n.22 (“We also note that the State has not attempted to reserve to itself exclusive power to deal with racial issues generally.”); CEE v. Wilson, 110 F.3d at 1444. This attempt is unavailing. Seattle clearly held that “when the political process or the decisionmaking mechanism used to address racially conscious legislation— and only such legislation — is singled out for peculiar and disadvantageous treatment, the government action plainly rests on distinctions *716based on race.” Seattle, 458 U.S. at 485-86, 102 S.Ct. at 3203.

. Even the Attorney General of the State of California, in defending Proposition 209, was not prepared to make this argument. It is advanced only by intervenor Californians Against Discrimination and Preferences, Inc. See CEE v. Wilson, 110 F.3d at 1441.

. Curiously, after devoting three pages of its opinion to the enthusiastic development of the argument that Proposition 209 burdens majority, not minority, interests, the panel disclaims any reliance upon it, instead "accepting] without questioning the District Court’s findings that Proposition 209 burdens members of insular minorities within the majority that enacted it who otherwise would seek to obtain race-based and gender-based preferential treatment from local entities.” CEE v. Wilson, 110 F.3d at 1441-43, 1442.

. See Brown v. Allen, 344 U.S. 443, 540, 73 S.Ct. 397, 427, 97 L.Ed. 469 (1953) (Jackson, J„ concurring in judgment) ("We are not final because we are infallible, but we are infallible only because we are final.”).

. For an analysis of the theory of predictionism, see Richard A. Posner, The Problems of Jurisprudence 221-28 (1990). For a case in which a circuit court engaged in such precedent-defying predictionism and was criticized by the Supreme Court for doing so, see Rodriguez de Quijas v. Shearson/American Express, 490 U.S. 477, 484, 109 S.Ct. 1917, 1921-22, 104 L.Ed.2d 526 (1989) (Kennedy, J.), aff'g Rodriguez De Quijas v. Shearson/Lehman Bros., 845 F.2d 1296, 1298-99 (5th Cir.1988).

. Justice Stevens, dissenting in Rodriguez, described the Fifth Circuit’s decision as "an indefensible brand of judicial activism.” 490 U.S. at 486, 109 S.Ct. at 1923 (1989) (Stevens, J., dissenting).

."If a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to this Court the prerogative of overruling its own decisions.” Rodriguez, 490 U.S. at 484, 109 S.Ct. at 1921-22 (Kennedy, J.). Accord Agostini v. Felton, - U.S. -, -, 117 S.Ct. 1997, 2017, 138 L.Ed.2d 391 (1997) (O’Connor, J.).