with whom Mr. Justice Douglas and Mr. Justice Marshall join, concurring in part and dissenting in part.
I join in the reversal of the District Court’s judgment insofar as that judgment upheld the constitutionality of Act 194 and the provisions of Act 195 respecting instructional materials and equipment, but dissent from Part III and the affirmance of the judgment upholding the constitutionality of the textbook provisions of Act 195.
A three-factor test by which to determine the compatibility with the Establishment Clause of state subsidies of sectarian educational institutions has evolved over 50 years of this Court’s stewardship in the field. The law in question must, first, reflect a clearly secular legislative purpose; second, have a primary effect1 that neither *374advances nor inhibits religion; and, third, avoid excessive government entanglement with religion. But four years ago, the Court, albeit without express recognition of the fact, added a significant fourth factor to the test: “A broader base of entanglement of yet a different character is presented by the divisive political potential of these state programs.” Lemon v. Kurtzman, 403 U. S. 602, 622 (1971). The evaluation of this factor in determining compatibility of a state subsidy law with the Establishment Clause is essential, said the Court, because:
“In a community where ... a large number of pupils are served by church-related schools, it can be assumed that state assistance will entail considerable political activity. Partisans of parochial schools, understandably concerned with rising costs and sincerely dedicated to both the religious and secular educational missions of their schools, will inevitably champion this cause and promote political action to achieve their goals. Those who oppose state aid, whether for constitutional, religious, or fiscal reasons, will inevitably respond and employ all of the usual political campaign techniques to' prevail. Candidates will be forced to declare and voters to choose. It would be unrealistic to ignore the fact that many people confronted with issues of this kind will find their votes aligned with their faith.
“Ordinarily political debate and division, however vigorous or even partisan, are normal and healthy manifestations of our democratic system of government, but political division along religious lines was one of the principal evils against which the First Amendment was intended to protect. . . . The potential divisiveness of such conflict is a threat to the normal political process. ... It conflicts with our whole history and tradition to permit questions of *375the Religion Clauses to assume such importance in our legislatures and in our elections that they could divert attention from the myriad issues and problems that confront every level of government. . . .
“. . . Here we are confronted with successive and very likely permanent annual appropriations that benefit relatively jew religious groups. Political fragmentation and divisiveness on religious lines are thus likely to be intensified.
“The potential for political divisiveness related to religious belief and practice is aggravated . . . by the need for continuing annual appropriations and the likelihood of larger and larger demands as costs and populations grow. . . .” Id., at 622-623. (Emphasis added.)
This factor was key in Kurtzman’s determination that Pennsylvania and Rhode Island statutes providing state aid to church-related elementary and secondary schools violated the Establishment Clause. The Pennsylvania statute provided financial support by way of reimbursement for the cost of teachers’ salaries, textbooks, and instructional materials in specified secular subjects. The Rhode Island statute provided a program under which the State paid directly to teachers in nonpublic schools a supplement of 15% of their annual salary.
Committee for Public Education & Religious Liberty v. Nyquist, 413 U. S. 756 (1973), decided two years later, emphasized the importance to be attached by judges to this fourth factor: “One factor of recurring significance in this weighing process is the potentially divisive political effect of an aid program.” Id., at 795. The Court held that the factor applied “with peculiar force to the New York statute now before us.” Id., at 796. That statute created three aid programs. The first provided for direct money grants to be used for maintenance and *376repair of facilities to ensure the students’ welfare, health, and safety. The second established a tuition-reimbursement plan for parents of children attending nonpublic elementary schools. The third provided tax relief for parents not qualifying for tuition reimbursements. Stating that “while the prospect of [political] divisiveness may not alone warrant the invalidation of state laws that otherwise survive the careful scrutiny required by the decisions of this Court, it is certainly a 'warning signal’ not to be ignored,” id., at 797-798, the Court held that “in light of all relevant considerations,” each of the New York-programs had a “ ‘primary effect that advances religion’ and offends the constitutional prohibition against laws ‘respecting an establishment of religion.’ ” Id., at 798.
The Court today also relies on the factor of divisive political potential but only as support for its holding that Act 194 is an unconstitutional law “respecting an establishment of religion,” stating:
“In addition, Act 194, like the statutes considered in [Kurtzman and Nyquist] creates a serious potential for divisive conflict over the issue of aid to religion — ‘entanglement in the broader sense of continuing political strife.’ . . . The recurrent nature of the appropriation process guarantees annual reconsideration of Act 194 and the prospect of repeated confrontation between proponents and opponents of the auxiliary-services program. The Act thus provides successive opportunities for political fragmentation and division along religious lines, one of the principal evils against which the Establishment Clause was intended to protect.” Ante, at 372.
Contrary to the plain and explicit teaching of Kurtzman and Nyquist, however, and inconsistently with- its own treatment of Act 194, the plurality, in considering *377the constitutionality of Act 195 says not a single word about the political-divisiveness factor in Part III of the opinion upholding the textbook loan program created by that Act, and makes only a passing footnote reference to the factor, without evaluation of its bearing on the result, in holding that Act 195’s program for loans of instructional materials and equipment constitutes Act 195 in that respect “direct aid to Pennsylvania’s predominantly church-related, nonpublic elementary and secondary schools, even though ostensibly limited to wholly neutral, secular instructional material and equipment, [that] inescapably results in the direct and substantial advancement of religious activity . . . and thus constitutes an impermissible establishment of religion.” Ante, at 366.
I recognize that the plurality was on the horns of a dilemma. The plurality notes that the total 1972-1973 appropriation under Act 195 was $16,660,000, of which $4,670,000 was appropriated to finance the textbook program. Ante, at 365 n. 15. The plurality notes further that “aid programs like Act 195 ... are dependent on continuing annual appropriations . . . which generate increasing demands as costs and population grow . . . ,” ibid., and, indeed, that the total Act 195 appropriation was increased $900,000 to $17,560,000 for the 1973-1974 school year. Plainly then, as in Nyquist, the political-divisiveness factor applies “with peculiar force to the ... statute now before us.” But to comply with Nyquist, as is required, the plurality obviously must attach determinative weight to the factor as respects both the textbook loan and instructional materials and equipment loan provisions, since both are inextricably intertwined in Act 195.2 For in light of the massive appropriations in*378volved, the plurality would be hard put to explain how the factor weighs determinatively against the validity of the instructional materials loan provisions, and not also against the validity of the textbook loan provisions. The plurality therefore would extricate itself from the horns of the dilemma by simply ignoring the factor in the weighing process.
But however much this evasion may be tolerable in the case of the instructional materials loan provisions, since these are invalidated on other grounds, responsibility for evaluating the weight to be accorded the factor cannot be evaded, in the case of the textbook loan provisions, by relying, as the plurality does, upon its agreement with the District Court that the textbook loan program is indistinguishable from the New York textbook loan program upheld in Board of Education v. Allen, 392 U. S. 236 (1968). For Allen, which I joined, was decided before Kurtzman ordained that the political-divisiveness factor must be involved in the weighing process, and understandably neither the parties to Allen nor the Court addressed that factor in that case. But whether or not Allen can withstand overruling in light of Kurtzman and Nyquist, which I question, it is clear that Kurtzman — which, I repeat, applied the factor to a Pennsylvania program that included reimbursement for the cost of textbooks — requires that the plurality weigh the factor in the instant case. Further, giving the factor the weight that Kurtzman and Nyquist require, compels, in my view *379the conclusion that the textbook loan program of Act 195, equally with the program for loan of instructional materials and equipment, violates the Establishment Clause. The plurality’s answer is that a difference in result is justified because Act 195 distinguishes between recipients of the loans: textbooks are lent to students, while instructional material and equipment are lent directly to the schools. That answer will not withstand analysis.
First, it is pure fantasy to treat the textbook program as a loan to students. It is true that, like the New York statute in Allen, Act 195 in terms talks of loans by the State of acceptable secular textbooks directly to students attending nonpublic schools. But even the 'plurality acknowledges that “the administrative practice is to have student requests for the books filed initially with the nonpublic school and to have the school authorities prepare collective summaries of these requests which they forward to the appropriate public officials. . . .” Ante, at 361. Further, “the nonpublic schools are permitted to store on their premises the textbooks being lent to the students.” Ante, at 361 n. 9. Even if these practices were also followed under the New York statute, the regulations implementing Act 195 make clear, as the record in Allen did not, that the nonpublic school in Pennsylvania is something more than a conduit between the State and pupil. The Commonwealth has promulgated “Guidelines for the Administration of Acts 194 and 195” to implement the statutes. These regulations, unlike those upheld in Allen, constitute a much more intrusive, and detailed involvement of the State and its processes into the administration of. nonpublic schools. The whole business is handled by the schools and public authorities, and neither parents nor students have a say. The guidelines make crystal clear that the nonpublic school, not its pupils, is the motivating force behind the textbook *380loan, and that virtually the entire loan transaction is to be, and is in fact, conducted between officials of the nonpublic school, on the one hand, and officers of the State, on the other.
For example, § 4.3 of the Guidelines requires that on or before March 1 of each year, an official of each nonpublic school submit to the Pennsylvania Department of Education a loan request for the desired textbooks. The requests must be submitted on standardized forms “distributed by the Department of Education ... to each nonpublic school or the appropriate chief administrator.” Section 4.6 of the Guidelines provides that the “[textbooks requested will be shipped directly to the appropriate nonpublic school.” Thus, although in terms the form provided by the Commonwealth for parents of nonpublic school students states that the parents of these pupils request the loan of textbooks directly from the State, the form is not returnable to the State, but to the nonpublic school, which tabulates the requests and submits its total to the State. Then, after the submission by the nonpublic school is approved by the appropriate state official, the books are transported not to the children whose parents ostensibly made the request, but directly to the nonpublic school, where they are physically retained when not in use in the classroom.
Indeed, the Guidelines make no attempt to mask the true nature of the loan transaction. In explicit words §4.10 describes the transaction: “Textbooks loaned to the nonpublic schools: (a) shall be maintained on an inventory by the nonpublic school.” (Emphasis added.) Section 4.11 provides: “It is presumed that textbooks on loan to nonpublic schools after a period of time will be lost, missing, obsolete or worn out. This information should be communicated to the Department of Education. After a period of six years, textbooks shall be *381declared unserviceable and the disposal of such shall be at the discretion of the Secretary of Education.” (Emphasis added.) Thus, the loan of the textbooks is treated by the regulations as what it in fact is: a loan from the State directly to the nonpublic school. Finally, § 4.12 completely removes any possible doubt. It provides:
"The nonpublic school or the agency which it is a member shall be responsible for maintaining on file certificates of requests from parents of children for all textbook materials loaned to them under this act. The file must be open to inspection by the appropriate authority. A letter certifying the certificates on file shall accompany all loan requests.”
Plainly, then, whatever may have been the case under the New York statute sustained in Allen, the loan ostensibly to students is, under Act 195, a loan in fact to the schools. In this regard, it should be observed that sophisticated attempts to avoid the Constitution are just as invalid as simple-minded ones. Lane v. Wilson, 307 U. S. 268, 275 (1939).
Second, in any event, Allen itself made clear that, far from providing a per se immunity from examination of the substance of the State’s program, even if the fact were, and it is not, that textbooks are loaned to the children rather than to the schools, that is only one among the factors to be weighed in determining the compatibility of the program with the Establishment Clause. Committee for Public Education & Religious Liberty v. Nyquist, 413 U. S., at 781. And, clearly, in the context of application of the factor of political divisiveness, it is wholly irrelevant whether the loan is to the children or to the school. A divisive political potential exists because aid programs, like Act 195, are dependent on continuing *382annual appropriations, and Act 195’s textbook loan program, even if we accepted it as a form of loans to students, involves increasingly massive sums now approaching $5,000,000 annually.3 It would blind reality to treat massive aid to nonpublic schools, under the guise of loans to the students, as not creating “a serious potential for divisive conflict over the issue of aid to religion.” Ante, at 372.4 The focus of the textbook loan program in terms of massive financial support for religious schools that creates the potential divisiveness is no less real than it is in the case of Act 195’s instructional materials provisions and Act 194’s invalidated program for auxiliary services. Act 195 is intended solely as a financial aid program to relieve the desperate financial plight of nonpublic, primarily parochial, schools. The plurality suggests that it is immaterial that Act 195 has that cast, in contrast with New York’s statute in Allen which authorized loans to students attending both public and nonpublic schools. Ante, at 360 n. 8. On the contrary, Act 195’s limitation of its financial support to aid to nonpublic school children exacerbates the potential for political divisiveness.*3835 “In this situation, where the underlying issue is the deeply emotional one of Church-State relationships, the potential for seriously divisive political consequences needs no elaboration.” Committee for Public Education & Religious Liberty v. Nyquist, supra, at 797.
Finally, the textbook loan provisions of Act 195, even if ostensibly limiting loans to nonpublic school children, violate the Establishment Clause for reasons independent of the political-divisiveness factor. As I have said, unlike the New York statute in Allen which extended assistance to all students, whether attending public or nonpublic schools, Act 195 extends textbook assistance only to a special class of students, children who attend nonpublic schools which are, as the plurality notes, primarily religiously oriented. The Act in that respect contains the same fatal defect as the New Jersey statute held violative of the Establishment Clause in Public Funds for Public Schools v. Marburger, 358 F. Supp. 29 (NJ 1973), aff’d, 417 U. S. 961 (1974). The statute there involved was N. J. Stat. Ann. § 18A: 58-63 which furnished state aid, in amounts up to $10 for elementary school students and up to $20 for high school students, to the parents of nonpublic school students as reimbursement for the cost of *384“secular, nonideological textbooks, instructional materials and supplies.” We affirmed the holding of the three-judge court that “because the language of [the statute] limits the assistance provided therein only to parents of children who attend nonpublic, predominately religiously-affiliated schools and not to parents of all school children, we are satisfied that its primary effect is to advance religion and that it is thereby unconstitutional.” 358 F. Supp., at 36. Marburger thus establishes that the plurality’s reliance today upon Allen is clearly misplaced.
Indeed, that reliance is also misplaced in light of its own holding today invalidating the provisions of Act 195 respecting the loan of instructional materials and equipment. I have no doubt that such materials and equipment are tools that substantially enhance the quality of the secular education provided by the religiously oriented schools. But surely the heart tools of that education are the textbooks that are prescribed for use and kept at the schools, albeit formally at the request of the students. Thus, what the Court says of the instructional materials and equipment, ante, at 365-366, may be said perhaps even more accurately of the textbooks:
“But faced with the substantial amounts of direct support authorized by Act 195, it 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 Act 195 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.”
In sum, I join the Court’s opinion as to Parts I, II, IV, *385and V, except that I would go further in Part IV and rest the invalidation of the provisions of Act 195 for loans of instructional materials and equipment also upon the political-divisiveness factor. I dissent from Part III.
The Court emphasized in Committee for Public Education & Religious Liberty v. Nyquist, 413 U. S. 756, 783-784, n. 39 (1973), that “primary effect” did not connote a requirement that the Court render an ultimate judgment on the effect of the statute in question. The Court stated:
“Appellees, focusing on the term 'principal or primary effect’ which this Court has utilized in expressing the second prong of the three-part test, . . . have argued that the Court must decide in these cases whether the ‘primary’ effect of New York’s tuition grant program is to subsidize religion or to promote these legitimate secular objectives. . . . We do not think that such metaphysical judgments are either possible or necessary. Our cases simply do not support the notion that a law found to have a ‘primary’ effect to promote some legitimate end under the State’s police power is immune from further examination to ascertain whether it also has the direct and immediate effect of advancing religion. . . .”
Kurtzman supports this conclusion:
“We have already noted that modem governmental programs have self-perpetuating and self-expanding propensities. These internal *378pressures are only enhanced when the schemes involve institutions whose legitimate needs are growing and whose interests have substantial political support. Nor can we fail to see that in constitutional adjudication some steps, which when taken were thought to approach ‘the verge,’ have become the platform for yet further steps. A certain momentum develops in constitutional theory and it can be a ‘downhill thrust’ easily set in motion but difficult to retard or stop.” 403 U. S. 602, 624 (1971).
I concede that I failed to apprehend the significance of the political-divisiveness factor in writing my separate opinion in Kurtzman, 403 U. S., at 642-661.
The Court stated in Nyqvist, 413 U. S., at 797 n. 56:
“The self-perpetuating tendencies of any form of government aid to religion have been a matter of concern running throughout our Establishment Clause cases. In Schempp, the Court emphasized that it was ‘no defense to urge that the religious practices here may be relatively minor encroachments on the First Amendment,’ for what today is a 'trickling stream’ may be a torrent tomorrow. 374 U. S., at 225. See also Lemon v. Kurtzman, 403 U. S., at 624-625. But, to borrow the words from Mr. Justice Rutledge’s forceful dissent in Everson, it is not alone the potential expandability of state tax aid that renders such aid invalid. Not even ‘three pence’ could be assessed: ‘Not the amount but “the principle of assessment was wrong.” ’ 330 U. S., at 40-41 (quoting from Madison’s Memorial and Remonstrance).”
Paraphrasing the Court’s observation in Nyquist, supra, at 783: “There has been no endeavor ‘to guarantee the separation between secular and religious educational functions and to ensure that State financial aid supports only the former.’ Lemon v. Kurtzman, supra, at 613. Indeed, it is precisely the function of [Act 195] to provide assistance to private schools, the great majority of which are sectarian. By [relieving parents of their textbook bill] the State seeks to relieve their financial burdens sufficiently to assure that they continue to have the option to send their children to religion-oriented schools. And while the other purposes for that aid — to perpetuate a pluralistic educational environment and to protect the fiscal integrity of overburdened public schools — are certainly unexceptionable, the effect of the aid is unmistakably to provide desired financial support for nonpublic, sectarian institutions.”