JUSTICE O'CONNOR, with whom JUSTICE BREYER joins, concurring in the judgment.
In 1965, Congress passed the Elementary and Secondary Education Act, 79 Stat. 27 (1965 Act). Under Title I, Congress provided monetary grants to States to address the needs of educationally deprived children of low-income families. Under Title II, Congress provided further monetary grants to States for the acquisition of library resources, textbooks, and other instructional materials for use by children and teachers in public and private elementary and secondary schools. Since 1965, Congress has reauthorized the Title I and Title II programs several times. Three Terms ago, we held in Agostini v. Felton, 521 U. S. 203 (1997), that Title I, as applied in New York City, did not violate the Establishment Clause. I believe that Agostini likewise controls the constitutional inquiry respecting Title II presented here, and requires the reversal of the Court of Appeals' judgment that the program is unconstitutional as applied in Jefferson Parish, Louisiana. To the extent our decisions in Meek v. Pit tenger, 421 U. S. 349 (1975), and Wolman v. Walter, 433 U. S. 229 (1977), are inconsistent with the Court's judgment today, I agree that those decisions should be overruled. I therefore concur in the judgment.
I
I write separately because, in my view, the plurality announces a rule of unprecedented breadth for the evaluation of Establishment Clause challenges to government school aid programs. Reduced to its essentials, the plurality's rule states that government aid to religious schools does not have the effect of advancing religion so long as the aid is offered on a neutral basis and the aid is secular in content. The plurality also rejects the distinction between direct and indirect aid, and holds that the actual diversion of secular aid by a religious school to the advancement of its religious mission is permissible. Although the expansive scope of the plurality's rule is troubling, two specific aspects of the opinion compel me to write separately. First, the plurality's treatment of neutrality comes close to assigning that factor singular importance in the future adjudication of Establishment Clause challenges to government school aid programs. Second, the plurality's approval of actual diversion of government aid to religious indoctrination is in tension with our O'CONNOR, J., concurring in judgment
precedents and, in any event, unnecessary to decide the instant case.
The clearest example of the plurality's near-absolute position with respect to neutrality is found in its following statement:"If the religious, irreligious, and areligious are all alike eligible for governmental aid, no one would conclude that any indoctrination that any particular recipient conducts has been done at the behest of the government. For attribution of indoctrination is a relative question. If the government is offering assistance to recipients who provide, so to speak, a broad range of indoctrination, the government itself is not thought responsible for any particular indoctrination. To put the point differently, if the government, seeking to further some legitimate secular purpose, offers aid on the same terms, without regard to religion, to all who adequately further that purpose, then it is fair to say that any aid going to a religious recipient only has the effect of furthering that secular purpose." Ante, at 809-810 (citation omitted).
I agree with JUSTICE SOUTER that the plurality, by taking such a stance, "appears to take evenhandedness neutrality and in practical terms promote it to a single and sufficient test for the establishment constitutionality of school aid." Post, at 900 (dissenting opinion).
I do not quarrel with the plurality's recognition that neutrality is an important reason for upholding government-aid programs against Establishment Clause challenges. Our cases have described neutrality in precisely this manner, and we have emphasized a program's neutrality repeatedly in our decisions approving various forms of school aid. See, e. g., Agostini, supra, at 228, 231-232; Zobrest v. Catalina Foothills School Dist., 509 U. S. 1, 10 (1993); Witters v. Washington Dept. of Servs. for Blind, 474 U. S. 481, 487-488 (1986); id., at 493 (O'CONNOR, J., concurring in part and concurring in judgment); Mueller v. Allen, 463 U. S. 388, 397-399 (1983). Nevertheless, we have never held that a government-aid program passes constitutional muster solely because of the neutral criteria it employs as a basis for distributing aid. For example, in Agostini, neutrality was only one of several factors we considered in determining that New York City's Title I program did not have the impermissible effect of advancing religion. See 521 U. S., at 226-228 (noting lack of evidence of inculcation of religion by Title I instructors, legal requirement that Title I services be supplemental to regular curricula, and that no Title I funds reached religious schools' coffers). Indeed, given that the aid in Agostini had secular content and was distributed on the basis of wholly neutral criteria, our consideration of additional factors demonstrates that the plurality's rule does not accurately describe our recent Establishment Clause jurisprudence. See also Zobrest, supra, at 10, 12-13 (noting that no government funds reached religious school's coffers, aid did not relieve school of expense it otherwise would have assumed, and aid was not distributed to school but to the child).
JUSTICE SOUTER provides a comprehensive review of our Establishment Clause cases on government aid to religious institutions that is useful for its explanation of the various ways in which we have used the term "neutrality" in our decisions. See post, at 878-883. Even if we at one time used the term "neutrality" in a descriptive sense to refer to those aid programs characterized by the requisite equipoise between support of religion and antagonism to religion, JusTICE SOUTER'S discussion convincingly demonstrates that the evolution in the meaning of the term in our jurisprudence is cause to hesitate before equating the neutrality of recent decisions with the neutrality of old. As I have previously explained, neutrality is important, but it is by no means the only "axiom in the history and precedent of the Establishment Clause." Rosenberger v. Rector and Visitors of Univ. of Va., 515 U. S. 819, 846 (1995) (concurring opinion). Thus, O'CONNOR, J., concurring in judgment
I agree with JUSTICE SOUTER'S conclusion that our "most recent use of 'neutrality' to refer to generality or evenhandedness of distribution… is relevant in judging whether a benefit scheme so characterized should be seen as aiding a sectarian school's religious mission, but this neutrality is not alone sufficient to qualify the aid as constitutional." Post, at 883-884.
I also disagree with the plurality's conclusion that actual diversion of government aid to religious indoctrination is consistent with the Establishment Clause. See ante, at 820825. Although "[o]ur cases have permitted some government funding of secular functions performed by sectarian organizations," our decisions "provide no precedent for the use of public funds to finance religious activities." Rosenberger, supra, at 847 (O'CONNOR, J., concurring). At least two of the decisions at the heart of today's case demonstrate that we have long been concerned that secular government aid not be diverted to the advancement of religion. In both Agostini, our most recent school aid case, and Board of Ed. of Central School Dist. No.1 v. Allen, 392 U. S. 236 (1968), we rested our approval of the relevant programs in part on the fact that the aid had not been used to advance the religious missions of the recipient schools. See Agostini, supra, at 226-227 ("[N]o evidence has ever shown that any New York City Title I instructor teaching on parochial school premises attempted to inculcate religion in students"); Allen, supra, at 248 ("Nothing in this record supports the proposition that all textbooks, whether they deal with mathematics, physics, foreign languages, history, or literature, are used by the parochial schools to teach religion"). Of course, our focus on the lack of such evidence would have been entirely unnecessary if we had believed that the Establishment Clause permits the actual diversion of secular government aid to religious indoctrination. Our decision in Bowen v. Kendrick, 487 U. S. 589 (1988), also demonstrates that actual diversion is constitutionally impermissible. After conclud ing that the government-aid program in question was constitutional on its face, we remanded the case so that the District Court could determine, after further factual development, whether aid recipients had used the government aid to support their religious objectives. See id., at 621-622; id., at 624 (KENNEDY, J., concurring) ("[T]he only purpose of further inquiring whether any particular grantee institution is pervasively sectarian is as a preliminary step to demonstrating that the funds are in fact being used to further religion"). The remand would have been unnecessary if, as the plurality contends, actual diversion were irrelevant under the Establishment Clause.
The plurality bases its holding that actual diversion is permissible on Witters and Zobrest. Ante, at 820-821. Those decisions, however, rested on a significant factual premise missing from this case, as well as from the majority of cases thus far considered by the Court involving Establishment Clause challenges to school aid programs. Specifically, we decided Witters and Zobrest on the understanding that the aid was provided directly to the individual student who, in turn, made the choice of where to put that aid to use. See Witters, 474 U. S., at 488; Zobrest, 509 U. S., at 10, 12. Accordingly, our approval of the aid in both cases relied to a significant extent on the fact that "[a]ny aid… that ultimately flows to religious institutions does so only as a result of the genuinely independent and private choices of aid recipients." Witters, supra, at 487; see Zobrest, supra, at 10 ("[A] government-paid interpreter will be present in a sectarian school only as a result of the private decision of individual parents"). This characteristic of both programs made them less like a direct subsidy, which would be impermissible under the Establishment Clause, and more akin to the government issuing a paycheck to an employee who, in turn, donates a portion of that check to a religious institution. See, e. g., Witters, supra, at 486-487; see also Rosenberger, supra, at 848 (O'CONNOR, J., concurring) (discussing Witters). O'CONNOR, J., concurring in judgment
Recognizing this distinction, the plurality nevertheless finds Witters and Zobrest-to the extent those decisions might permit the use of government aid for religious purposes-relevant in any case involving a neutral, per-capitaaid program. See ante, at 830-831. Like JUSTICE SOUTER, I do not believe that we should treat a per-capita-aid program the same as the true private-choice programs considered in Witters and Zobrest. See post, at 902. First, when the government provides aid directly to the student beneficiary, that student can attend a religious school and yet retain control over whether the secular government aid will be applied toward the religious education. The fact that aid flows to the religious school and is used for the advancement of religion is therefore wholly dependent on the student's private decision. See Rosenberger, 515 U. S., at 848 (O'CONNOR, J., concurring) (discussing importance of private choice in Witters); Witters, 474 U. S., at 488 ("[T]he fact that aid goes to individuals means that the decision to support religious education is made by the individual, not by the State"); id., at 493 (O'CONNOR, J., concurring in part and concurring in judgment) ("The aid to religion at issue here is the result of petitioner's private choice"). It is for this reason that in Agostini we relied on Witters and Zobrest to reject the rule "that all government aid that directly assists the educational function of religious schools is invalid," 521 U. S., at 225, yet also rested our approval of New York City's Title I program in part on the lack of evidence of actual diversion, id., at 226-227.
Second, I believe the distinction between a per capita school aid program and a true private-choice program is significant for purposes of endorsement. See, e. g., Lynch v. Donnelly, 465 U. S. 668, 692 (1984) (O'CONNOR, J., concurring). In terms of public perception, a government program of direct aid to religious schools based on the number of students attending each school differs meaningfully from the government distributing aid directly to individual students who, in turn, decide to use the aid at the same religious schools. In the former example, if the religious school uses the aid to inculcate religion in its students, it is reasonable to say that the government has communicated a message of endorsement. Because the religious indoctrination is supported by government assistance, the reasonable observer would naturally perceive the aid program as government support for the advancement of religion. That the amount of aid received by the school is based on the school's enrollment does not separate the government from the endorsement of the religious message. The aid formula does notand could not-indicate to a reasonable observer that the inculcation of religion is endorsed only by the individuals attending the religious school, who each affirmatively choose to direct the secular government aid to the school and its religious mission. No such choices have been made. In contrast, when government aid supports a school's religious mission only because of independent decisions made by numerous individuals to guide their secular aid to that school, "[n]o reasonable observer is likely to draw from the facts… an inference that the State itself is endorsing a religious practice or belief." Witters, supra, at 493 (O'CONNOR, J., concurring in part and concurring in judgment). Rather, endorsement of the religious message is reasonably attributed to the individuals who select the path of the aid.
Finally, the distinction between a per-capita-aid program and a true private-choice program is important when considering aid that consists of direct monetary subsidies. This Court has "recognized special Establishment Clause dangers where the government makes direct money payments to sectarian institutions." Rosenberger, 515 U. S., at 842; see also ibid. (collecting cases). If, as the plurality contends, a per-capita-aid program is identical in relevant constitutional respects to a true private-choice program, then there is no reason that, under the plurality's reasoning, the government should be precluded from providing direct money payments O'CONNOR, J., concurring in judgment
to religious organizations (including churches) based on the number of persons belonging to each organization. And, because actual diversion is permissible under the plurality's holding, the participating religious organizations (including churches) could use that aid to support religious indoctrination. To be sure, the plurality does not actually hold that its theory extends to direct money payments. See ante, at 818-820. That omission, however, is of little comfort. In its logic-as well as its specific advisory language, see ante, at 819-820, n. 8-the plurality opinion foreshadows the approval of direct monetary subsidies to religious organizations, even when they use the money to advance their religious objectives.
Our school aid cases often pose difficult questions at the intersection of the neutrality and no-aid principles and therefore defy simple categorization under either rule. As I explained in Rosenberger, "[r]esolution instead depends on the hard task of judging-sifting through the details and determining whether the challenged program offends the Establishment Clause. Such judgment requires courts to draw lines, sometimes quite fine, based on the particular facts of each case." 515 U. S., at 847 (concurring opinion). Agostini represents our most recent attempt to devise a general framework for approaching questions concerning neutral school aid programs. Agostini also concerned an Establishment Clause challenge to a school aid program closely related to the one at issue here. For these reasons, as well as my disagreement with the plurality's approach, I would decide to day's case by applying the criteria set forth in Agostini.
II
In Agostini, after reexammmg our jurisprudence since School Dist. of Grand Rapids v. Ball, 473 U. S. 373 (1985), we explained that the general principles used to determine whether government aid violates the Establishment Clause have remained largely unchanged. 521 U. S., at 222. Thus, we still ask "whether the government acted with the purpose of advancing or inhibiting religion" and "whether the aid has the 'effect' of advancing or inhibiting religion." Id., at 222223. We also concluded in Agostini, however, that the specific criteria used to determine whether government aid has an impermissible effect had changed. Id., at 223. Looking to our recently decided cases, we articulated three primary criteria to guide the determination whether a governmentaid program impermissibly advances religion: (1) whether the aid results in governmental indoctrination, (2) whether the aid program defines its recipients by reference to religion, and (3) whether the aid creates an excessive entanglement between government and religion. Id., at 234. Finally, we noted that the same criteria could be reviewed to determine whether a government-aid program constitutes an endorsement of religion. Id., at 235.
Respondents neither question the secular purpose of the Chapter 2 (Title II) program nor contend that it creates an excessive entanglement. (Due to its denomination as Chapter 2 of the Education Consolidation and Improvement Act of 1981, 95 Stat. 469, the parties refer to the 1965 Act's Title II program, as modified by subsequent legislation, as "Chapter 2." For ease of reference, I will do the same.) Accordingly, for purposes of deciding whether Chapter 2, as applied in Jefferson Parish, Louisiana, violates the Establishment Clause, we need ask only whether the program results in governmental indoctrination or defines its recipients by reference to religion.
Taking the second inquiry first, it is clear that Chapter 2 does not define aid recipients by reference to religion. In Agostini, we explained that scrutiny of the manner in which a government-aid program identifies its recipients is important because "the criteria might themselves have the effect of advancing religion by creating a financial incentive to undertake religious indoctrination." 521 U. S., at 231. We then clarified that this financial incentive is not present O'CONNOR, J., concurring in judgment
"where the aid is allocated on the basis of neutral, secular criteria that neither favor nor disfavor religion, and is made available to both religious and secular beneficiaries on a nondiscriminatory basis." Ibid. Under Chapter 2, the Secretary of Education allocates funds to the States based on each State's share of the Nation's school-age population. 20 U. S. C. § 7311(b). The state educational agency (SEA) of each recipient State, in turn, must distribute the State's Chapter 2 funds to local educational agencies (LEA's) "according to the relative enrollments in public and private, nonprofit schools within the school districts of such agencies," adjusted to take into account those LEA's "which have the greatest numbers or percentages of children whose education imposes a higher than average cost per child." § 7312(a). The LEA must then expend those funds on "innovative assistance programs" designed to improve student achievement. § 7351(b). The statute generally requires that an LEA ensure the "equitable participation" of children enrolled in private nonprofit elementary and secondary schools, § 7372(a)(1), and specifically mandates that all LEA expenditures on behalf of children enrolled in private schools "be equal (consistent with the number of children to be served) to expenditures for programs… for children enrolled in the public schools of the [LEA]," § 7372(b). As these statutory provisions make clear, Chapter 2 uses wholly neutral and secular criteria to allocate aid to students enrolled in religious and secular schools alike. As a result, it creates no financial incentive to undertake religious indoctrination.
Agostini next requires us to ask whether Chapter 2 "result[s] in governmental indoctrination." 521 U. S., at 234. Because this is a more complex inquiry under our case law, it is useful first to review briefly the basis for our decision in Agostini that N ew York City's Title I program did not result in governmental indoctrination. Under that program, public-school teachers provided Title I instruction to eligible students on private school premises during regular school hours. Twelve years earlier, in Aguilar v. Felton, 473 U. S. 402 (1985), we had held the same New York City program unconstitutional. In Ball, a companion case to Aguilar, we also held that a similar program in Grand Rapids, Michigan, violated the Constitution. Our decisions in Aguilar and Ball were both based on a presumption, drawn in large part from Meek, see 421 U. S., at 367-373, that public-school instructors who teach secular classes on the campuses of religious schools will inevitably inculcate religion in their students.
In Agostini, we recognized that "[o]ur more recent cases [had] undermined the assumptions upon which Ball and Aguilar relied." 521 U. S., at 222. First, we explained that the Court had since abandoned "the presumption erected in Meek and Ball that the placement of public employees on parochial school grounds inevitably results in the impermissible effect of state-sponsored indoctrination or constitutes a symbolic union between government and religion." Id., at 223. Rather, relying on Zobrest, we explained that in the absence of evidence showing that teachers were actually using the Title I aid to inculcate religion, we would presume that the instructors would comply with the program's secular restrictions. See Agostini, 521 U. S., at 223-224, 226227. The Title I services were required by statute to be " 'secular, neutral, and nonideological.'" Id., at 210 (quoting 20 U. S. C. § 6321(a)(2)).
Second, we noted that the Court had "departed from the rule relied on in Ball that all government aid that directly assists the educational function of religious schools is invalid." Agostini, supra, at 225. Relying on Witters and Zobrest, we noted that our cases had taken a more forgiving view of neutral government programs that make aid available generally without regard to the religious or nonreligious character of the recipient school. See Agostini, 521 U. S., at 225-226. With respect to the specific Title I pro O'CONNOR, J., concurring in judgment
gram at issue, we noted several factors that precluded us from finding an impermissible financing of religious indoctrination: the aid was "provided to students at whatever school they choose to attend," the services were "by law supplemental to the regular curricula" of the benefited schools, "[n]o Title I funds ever reach the coffers of religious schools," and there was no evidence of Title I instructors having "attempted to inculcate religion in students." Id., at 226-228. Relying on the same factors, we also concluded that the New York City program could not "reasonably be viewed as an endorsement of religion." Id., at 235. Although we found it relevant that Title I services could not be provided on a schoolwide basis, we also explained that this fact was likely a sufficient rather than a necessary condition of the program's constitutionality. We were not "willing to conclude that the constitutionality of an aid program depends on the number of sectarian school students who happen to receive the otherwise neutral aid." Id., at 229.
The Chapter 2 program at issue here bears the same hallmarks of the New York City Title I program that we found important in Agostini. First, as explained above, Chapter 2 aid is distributed on the basis of neutral, secular criteria. The aid is available to assist students regardless of whether they attend public or private nonprofit religious schools. Second, the statute requires participating SEA's and LEA's to use and allocate Chapter 2 funds only to supplement the funds otherwise available to a religious school. 20 U. S. C. § 7371(b). Chapter 2 funds must in no case be used to supplant funds from non-Federal sources. Ibid. Third, no Chapter 2 funds ever reach the coffers of a religious school. Like the Title I program considered in Agostini, all Chapter 2 funds are controlled by public agencies-the SEA's and LEA's. § 7372(c)(1). The LEA's purchase instructional and educational materials and then lend those materials to public and private schools. See §§ 7351(a), (b)(2). With respect to lending to private schools under Chapter 2, the statute specifically provides that the relevant public agency must retain title to the materials and equipment. § 7372(c)(1). Together with the supplantation restriction, this provision ensures that religious schools reap no financial benefit by virtue of receiving loans of materials and equipment. Finally, the statute provides that all Chapter 2 materials and equipment must be "secular, neutral, and nonideological." § 7372(a)(1). That restriction is reinforced by a further statutory prohibition on "the making of any payment… for religious worship or instruction." § 8897. Although respondents claim that Chapter 2 aid has been diverted to religious instruction, that evidence is de minimis, as I explain at greater length below. See infra, at 864-867.
III
Respondents contend that Agostini is distinguishable, pointing to the distinct character of the aid program considered there. See Brief for Respondents 44-47. In Agostini, federal funds paid for public-school teachers to provide secular instruction to eligible children on the premises of their religious schools. Here, in contrast, federal funds pay for instructional materials and equipment that LEA's lend to religious schools for use by those schools' own teachers in their classes. Because we held similar programs unconstitutional in Meek and Wolman, respondents contend that those decisions, and not Agostini, are controlling. See, e. g., Brief for Respondents 11,22-25. Like respondents, JUSTICE SOUTER also relies on Meek and Wolman in finding the character of the Chapter 2 aid constitutionally problematic. See post, at 893,903.
At the time they were decided, Meek and Wolman created an inexplicable rift within our Establishment Clause jurisprudence concerning government aid to schools. Seven years before our decision in Meek, we held in Allen that a New York statute that authorized the lending of textbooks to students attending religious schools did not violate the O'CONNOR, J., concurring in judgment
Establishment Clause. 392 U. S., at 238. We explained that the statute "merely [made] available to all children the benefits of a general program to lend school books free of charge," that the State retained ownership of the textbooks, and that religious schools received no financial benefit from the program. Id., at 243-244. We specifically rejected the contrary argument that the statute violated the Establishment Clause because textbooks are critical to the teaching process, which in a religious school is employed to inculcate religion. Id., at 245-248.
In Meek and Wolman, we adhered to Allen, holding that the textbook lending programs at issue in each case did not violate the Establishment Clause. See Meek, 421 U. S., at 359-362 (plurality opinion); Wolman, 433 U. S., at 236238 (plurality opinion). At the same time, however, we held in both cases that the lending of instructional materials and equipment to religious schools was unconstitutional. See Meek, supra, at 362-366; Wolman, supra, at 248-251. We reasoned that, because the religious schools receiving the materials and equipment were pervasively sectarian, any assistance in support of the schools' educational missions would inevitably have the impermissible effect of advancing religion. For example, in Meek we explained:"[I]t would simply ignore reality to attempt to separate secular educational functions from the predominantly religious role performed by many of Pennsylvania's church-related elementary and secondary schools and to then characterize [the statute] as channeling aid to the secular without providing direct aid to the sectarian. Even though earmarked for secular purposes, 'when it flows to an institution in which religion is so pervasive that a substantial portion of its functions are subsumed in the religious mission,' state aid has the impermissible primary effect of advancing religion." 421 U. S., at 365-366 (quoting Hunt v. McNair, 413 U. S. 734, 743 (1973)). Thus, we held that the aid program "necessarily results in aid to the sectarian school enterprise as a whole," and "inescapably results in the direct and substantial advancement of religious activity." Meek, supra, at 366 (emphases added). Similarly, in Wolman, we concluded that, "[i]n view of the impossibility of separating the secular education function from the sectarian, the state aid inevitably flows in part in support of the religious role of the schools." 433 U. S., at 250 (emphasis added).
For whatever reason, the Court was not willing to extend this presumption of inevitable religious indoctrination to school aid when it instead consisted of textbooks lent free of charge. For example, in Meek, despite identifying the religious schools' secular educational functions and religious missions as inextricably intertwined, 421 U. S., at 366, the Court upheld the textbook lending program because "the record in the case…, like the record in Allen, contains no suggestion that religious textbooks will be lent or that the books provided will be used for anything other than purely secular purposes," id., at 361-362 (citation omitted). Accordingly, while the Court was willing to apply an irrebuttable presumption that secular instructional materials and equipment would be diverted to use for religious indoctrination, it required evidence that religious schools were diverting secular textbooks to religious instruction.
The inconsistency between the two strands of the Court's jurisprudence did not go unnoticed, as Justices on both sides of the Meek and Wolman decisions relied on the contradiction to support their respective arguments. See, e. g., Meek, 421 U. S., at 384 (Brennan, J., concurring in part and diss