CLAY, J., delivered the opinion of the court, in which SILER, J., joined. RYAN, J. (963-74), delivered a separate opinion concurring in part and dissenting in part.
OPINION
CLAY, Circuit Judge.Defendants and Intervenors Dr. Susan Tave Zelman, et al; Senel Taylor, et al.; and Hanna Perkins School, et al., appeal from the order entered by the United States District Court for the Northern District of Ohio, enjoining on summary judgment the Ohio Pilot Project Scholarship Program on the ground that it violates the Establishment Clause of the First Amendment. For the following reasons, we AFFIRM.
I.
In 1995, Ohio’s General Assembly adopted the Ohio Pilot Project Scholarship Program (“voucher program” or “the program”) in response to an order by the United States District Court that placed the Cleveland School District under the direct management and supervision of the State Superintendent of Public Instruction due to mismanagement by the local school board. The voucher program covers any state school district that has been the subject of a federal court order “requiring supervision and operational management of the district by the state superintendent.” Ohio Rev.Code § 3313.975(A). The program provides scholarships to children residing within the applicable district in grades kindergarten through eighth grade. See Ohio Rev.Code § 3313.975(C)(1). The program gives “preference to students from low-income families,” defining them as those whose families’ income is less than 200% of the poverty line. See Ohio Rev.Code § 3313.978(A). “Scholarships may be awarded to students who are not from low-income families only if all students from low-income families have been given first consideration for placement.” Cleveland Scholarship and Tutoring Program, Administration Procedures Manual, 1-11 (J.A. at 1358) (emphasis original). Over sixty percent of the children receiving scholarships in the program are from families with incomes at or below the poverty line.
The voucher program pays scholarships according to family income. The program requires participating private schools to cap tuition at $2500 per student per year and pays 90% of whatever tuition the school actually charges for low-income families; for other families, the State pays 75% of the school’s tuition up to a maximum of $1875. See Ohio Rev.Code. §§ 3313.976(A)(8), 3313.978(A). Each scholarship for children attending a private school is payable to the parents of the student entitled to the scholarship. Ohio Rev.Code § 3313.979. Scholarship checks are mailed to the school selected by the parents, where the parents are required to endorse the checks over to the school in order to pay tuition.
Schools wishing to be designated as program participants eligible to enroll scholarship students must register with the voucher program. Private schools located within the boundaries of the Cleveland school district which meet the State’s educational standards may participate. See Ohio Rev.Code § 3313.976(A)(1) and (3). Schools are required to follow .the program’s priority rules regarding the placement of students and may not discriminate on the basis of race, religion, or ethnic background; advocate or foster unlawful behavior; or teach hatred of any person or group on the basis of race, ethnicity, na*949tional origin, or religion. See Ohio Rev. Code § 3313.976(A)(6). Public schools in districts adjacent to the district in which the voucher program is implemented may also register for the program and “receive scholarship payments on behalf of parents,” but none of the public schools in districts adjacent to Cleveland have done so. Ohio Rev.Code § 3313.976(C). The checks for program participants at public schools are made out to the participating school district rather than to the parents. No public schools have registered for the program since its enactment.
For the 1999-2000 school year, 3,761 students enrolled in the program; 60% of the enrollees are from families at or below the poverty level. Of these, 3,632 (96%) are enrolled in sectarian schools. At one time in the course of the program, as many as 22% of the students enrolled in the program attended nonreligious schools. During the 1999-2000 school year, fifty-six schools registered to participate in the program; forty-six (82%) are church-affiliated. Program monies may be used by the participating schools for whatever purpose they deem appropriate; the voucher program does not place restrictions on the use of funds made available under the program.
The sectarian schools vary in their religious affiliation and approaches; however, the handbooks and mission statements of these schools reflect that most believe in interweaving religious beliefs with secular subjects. The sectarian schools also follow religious guidelines, including instruction in religion and mandated participation in religious services; interweaving of Christian doctrines with science and language arts classes; requiring that “all learning take place in an atmosphere of religious ideals,” St. Vincent de Paul School, Parent Handbook 11 (1999-2000); and designing educational scholarship in order “to make ... faith become living, conscious, and active through the light of instruction ... religious truths and values permeate the whole atmosphere of the school.” Saint Rocco School, Parent Student Handbook 1 (1999-2000). Other sectarian schools in the voucher program believe that “the one cardinal objective of education to which all others point is to develop devotion to God as our Creator, Redeemer, and Sanctifier,” Saint John Nottingham Lutheran School, Parent Handbook 2 (1999-2000); and to require students to “pledge allegiance to the Christian flag and to the Savior for whose Kingdom it stands, One Savior crucified, risen and coming again with life and liberty for all who believe.” Calvary Center Academy, Parent Student Handbook 24 (1999-2000).
In prior litigation, Doris Simmons Harris, one of the Plaintiffs herein, brought a state court lawsuit challenging the constitutionality of the voucher program under multiple provisions of the Ohio Constitution, and under the Establishment Clause of the United States Constitution. On May 27, 1999, the Ohio Supreme Court issued a judgment in favor of Plaintiffs, holding that the 1995 voucher program had been enacted in violation of the one-subject rule of the Ohio Constitution, and “must be stricken” from the Ohio statute books. See Simmons-Harris v. Goff, 86 Ohio St.3d 1, 711 N.E.2d 203, 216 (1999). However, a majority of the justices rejected Plaintiffs’ claims that the program violated the Establishment Clause. Id. at 207-11, 218-19. Those justices reasoned that “[t]he Nyquist holding [had been] undermined” by subsequent cases and was thus no longer good law. Id. at 208. A concurring opinion noted that “[w]ith regard to the rest of the majority opinion [the section not dealing with the one-subject rule], ... I find a number of the other assertions by the majority to be advisory in nature.” Id. at 216 (Douglas, J., joined by Resnick and Sweeney, JJ., concurring in the judgment only). Since this case, the Ohio Legislature has re-enacted the voucher program in a manner remedying the one-subject problem; however, the 1999 program is in all relevant aspects, the *950same as the original pilot scholarship program enacted by the Legislature in 1995.
On July 20, 1999, Simmons Harris, the parent of a minor child enrolled in the Cleveland City School District for the 1999-2000 school year; Marla Franklin, a teacher in the Lorain City School District; and Steven Behr, pastor of Our Savior/Nuestro Salvado Church in Lorain, Ohio, filed suit in Case No. I:99cvl740, against Defendant Dr. Susan Tave Zelman in her official capacity as Superintendent of Public Instruction for the Ohio Department of Education, seeking to enjoin a portion of the program on the ground that it violated the Establishment Clause of the First Amendment. On July 29, 1999, Sue Gatton, chair of Citizens Against Vouchers; Mary Murphy, a teacher in the Cleveland City School District; Michael Debose, a pastor in Cuyahoga County, Ohio; Cheryl Debose and Glenn Altschuld, Ohio Taxpayers; and Deidra Pearson, the parent of a child enrolled in the Cleveland City School District, filed suit in Case No. I:99cvl818 against Defendants Dr. Susan Tave Zel-man, in her official capacity as Superintendent; the State of Ohio through its General Assembly, Governor and other agents; and Saundra Berry, in her official capacity as. Program Administrator of the Ohio Pilot Scholarship Program, seeking the same injunctive relief as in Case No. I:99cvl740.
Proposed Intervenors Senel Taylor, Johnnietta McGrady, Christine Suma, Arkela Winston, and Amy Hudock, on their own behalf and as natural guardians of their respective children, filed an answer to Simmons Harris’ complaint on July 27, 1999. Proposed Intervenors Hanna Perkins School, Ivy Chambers, Carol Lambert, Our Lady of Peace School, West-park Lutheran School Association, Inc., Lutheran Memorial Association of Cleveland, and Delores Jones, filed an answer to Simmons Harris’ complaint on August 2, 1999.
On August 13, 1999, the district court held a preliminary injunction hearing in both cases, and on August 24, 1999, granted Plaintiffs the injunctive relief sought. In the same order, the district court consolidated the two cases, and found that the Ohio Supreme Court’s decision in Simmons-Harris v. Goff, 86 Ohio St.3d 1, 711 N.E.2d 203 (1999), did not preclude federal consideration of the constitutional challenge to the voucher program because the Ohio court’s decision rested on a state ground which independently supported its resolution of the case. Thereafter, on August 27, 1999, the district court granted in part Defendants’ motion for a stay of the preliminary injunction.
On August 24, 1999, the same day that the district court granted Plaintiffs’ motion for a preliminary injunction, the State and the two intervening Defendants appealed that decision to this Court. After the district court’s August 27, 1999, order granting a limited stay of its preliminary injunction, all Defendants filed revised briefs with this Court, appealing the preliminary injunction with regard to students who were new to the voucher program — i.e., the portion of the preliminary injunction not stayed by the district court’s August 27, 1999 order. While those appeals were pending, the State filed a motion for a stay of the preliminary injunction with the United States Supreme Court, which the Supreme Court granted by a vote of 5-4 on November 5, 1999, pending this Court’s final disposition of the entire appeal. See Zelman v. Simmons-Harris, 528 U.S. 983, 120 S.Ct. 443, 145 L.Ed.2d 346 (1999). Thereafter, this Court entered an order on November 15, 1999, concluding that the Supreme Court’s decision granting the State’s motion for a stay rendered moot Defendants’ pending motions for a stay in this Court. The case proceeded in the district court on an expedited basis.
On October 15, 1999, all parties stipulated that the handbooks, mission statements and brochures of the schools participating in the Cleveland scholarship program are “authentic, speak for themselves, have been made available to the parents of the scholarship students and *951are not false or misleading. Some of these documents however may not accurately reflect admission standards that had to be revised.” Plaintiffs and Defendants both filed motions for summary judgment on November 1, 1999.
On November 29, 1999, the district court denied Intervenor Taylor’s motion to have the following question certified to the Ohio Supreme Court: “Does Ohio law give pre-clusive effect to the resolution of the Establishment Clause claim in Simmons-Harris v. Goff, 86 Ohio St.3d 1, 711 N.E.2d 203 (1999)?”
The district court granted Plaintiffs’ motion for summary judgment on December 20, 1999, finding that the voucher program violated the Establishment Clause; enjoined Defendants from administering the program; and denied Defendants’ motion for summary judgment. See Simmons-Harris v. Zelman, 72 F.Supp.2d 834, 836 (N.D.Ohio 1999). The court stayed its summary judgment order with Plaintiffs’ consent pending review by this Court. Defendants and Intervenors appealed to this Court on January 12, 2000.
II.
We recognize the significance that this issue holds for many members of our society. The issue of school vouchers has been the subject of intense political and public commentary, discussion, and attention in recent years, and we would be remiss if we failed to acknowledge the seriousness of the concerns this case has raised. We do not, however, have the luxury of responding to advents in educational policy with academic discourse on practical solutions to the problem of failing schools; nor may we entertain a discussion on what might be legally acceptable in a hypothetical school district. We may only apply the controlling law to the case and statute before us.
The courts do not make educational policy; we do not sit in omnipotent judgment as to the efficacy of one scheme or program versus another. The design or specifics of a program intended to remedy the problem of failing schools and to rectify educational inequality must be reserved to the states and the school boards within them, with one caveat: the proposed program may not run afoul of the freedoms guaranteed to all citizens in the Constitution. In other words, the determinations of states and school boards cannot infringe upon the necessary separation between church and state. We therefore consider the program presented before us under the controlling precedents of the United States Supreme Court and this Court to determine whether such infringement has occurred.
This Court reviews the district court’s grant of summary judgment de novo. See Coles v. Cleveland Bd. of Educ., 171 F.3d 369, 376 (6th Cir.1999). Summary judgment is proper when “the pleadings, depositions, answers to interrogate-' ries, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(c). There is no genuine issue of material fact when the “record taken as a whole could not lead a rational trier of fact to find for the nonmoving party.” Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 587, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986).
The Establishment Clause provides that “Congress shall make no law respecting an establishment of religion-” U.S. Const., amend. I. In Lemon v. Kurtzman, 403 U.S. 602, 612-13, 91 S.Ct. 2105, 29 L.Ed.2d 745 (1971), the Supreme Court set forth the following test to determine whether a statute passes muster under the Establishment Clause: 1) the statute must have a secular legislative purpose; 2) the principal or primary effect of the statute must be one that neither advances nor inhibits religion; and 3) the statute must not foster excessive entanglement with re*952ligion. If a statute fails any portion of this test, it violates the Establishment Clause. Id. The Supreme Court has applied the Lemon test regularly in the context of schools and education. See, e.g., Kiryas Joel Sch. Dist. v. Grumet, 512 U.S. 687, 114 S.Ct. 2481, 129 L.Ed.2d 546 (1994) (drawing a separate school district for one religion violates the Establishment Clause); Stone v. Graham, 449 U.S. 39, 101 S.Ct. 192, 66 L.Ed.2d 199 (1980)(posting Ten Commandments on walls of public school violates Establishment Clause); Bowen v. Kendrick, 487 U.S. 589, 108 S.Ct. 2562, 101 L.Ed.2d 520 (1988) (holding that Establishment Clause does not prevent religious organizations from participating in federally funded program); Edwards v. Aguillard, 482 U.S. 578, 107 S.Ct. 2573, 96 L.Ed.2d 510 (1987) (overturning statute which required the teaching of Creationism in public schools); Aguilar v. Felton, 473 U.S. 402, 105 S.Ct. 3232, 87 L.Ed.2d 290 (1985) (paying public school employees to teach in parochial school violates the Establishment Clause), overruled by Agostini v. Felton, 521 U.S. 203, 117 S.Ct. 1997, 138 L.Ed.2d 391 (1997); Sch. Dist. of Grand Rapids v. Ball, 473 U.S. 373, 105 S.Ct. 3216, 87 L.Ed.2d 267 (1985) (finding a shared time program to be a violation of Establishment Clause), overruled by Agostini v. Felton, 521 U.S. 203, 117 S.Ct. 1997, 138 L.Ed.2d 391 (1997); Mueller v. Allen, 463 U.S. 388, 103 S.Ct. 3062, 77 L.Ed.2d 721 (1983)(permitting taxpayers to deduct from state income tax expenses incurred in sending children to parochial schools does not violate Establishment Clause); Roemer v. Bd. of Pub. Works, 426 U.S. 736, 96 S.Ct. 2337, 49 L.Ed.2d 179 (1976)(fmding no violation of Establishment Clause in providing government aid to both public and private universities).
The Supreme Court and individual justices have introduced variations on the Lemon test in other contexts. See Lee v. Weisman, 505 U.S. 577, 592-98, 112 S.Ct. 2649, 120 L.Ed.2d 467 (1992) (using coercion test developed by Justice Kennedy to hold that school could not provide for nonsectarian prayer to be given at graduation by school-selected clergyman); County of Allegheny v. ACLU, 492 U.S. 573, 594-602, 655-79, 109 S.Ct. 3086, 106 L.Ed.2d 472 (1989) (using two-part test developed by Justice O’Connor — Establishment Clause is violated when 1) government is excessively entangled with religion, or 2) government endorses or disapproves of religion- — as part of the Lemon test regarding government displays of objects with religious connotations; introducing Justice Kennedy’s two-part coercion test — 1) government may not coerce participation in religion, and 2) government may not directly benefit religion — in his concurrence); Wallace v. Jaffree, 472 U.S. 38, 56 & n. 42, 105 S.Ct. 2479, 86 L.Ed.2d 29 (1985) (using Justice O’Connor’s two-part test as part of the Lemon analysis and finding it appropriate to determine whether the government’s purpose is to endorse or disapprove of religion); Lynch v. Donnelly, 465 U.S. 668, 687-94, 104 S.Ct. 1355, 79 L.Ed.2d 604 (1984) (introducing Justice O’Connor’s two-part test).
In Agostini v. Felton, 521 U.S. 203, 117 S.Ct. 1997, 138 L.Ed.2d 391 (1997), the Court reaffirmed the importance of the Lemon test in Establishment Clause cases involving school aid, but noted that the entanglement prong could be considered as an aspect of the effects inquiry. Agosti-ni found “three primary criteria” used by the Court in evaluating whether government aid has the effect of advancing religion: whether the statute or programs in question “result in governmental indoctrination; define its recipients by reference to religion; or create an excessive entanglement.” Id. at 234, 117 S.Ct. 1997.
The Supreme Court has not overturned or rescinded the Lemon test even as it has used its framework to shape differing anal-yses. Although in Agostini the Court articulated the primary criteria it would utilize to determine whether a government-aid program impermissibly advanced or endorsed religion, the Court has not neees-*953sarily limited itself to considering solely those criteria. Rather, it seems evident that the Agostini Court illustrated the Lemon test’s flexibility and its evolution from the relatively rigid three-part test to an approach in which the varying components of a particular program or statute are analyzed with regard to their impact on, in the context of schools, the relevant students or communities. We therefore look to these components as aspects of the proper analysis under Lemon, but acknowledge that precedent does not limit itself to only these components should other components previously utilized by the Court be relevant, such as coercion of citizens, endorsement of religion, and direct benefit to religion. See Mitchell v. Helms, 580 U.S. 793, -, 120 S.Ct. 2580, 2556, 147 L.Ed.2d 660 (2000) (O’Connor, J., concurring) (finding that Agostini represents a general framework for approaching questions concerning neutral school-aid programs but recognizing that these type of eases depend on the particular facts of each case).
Of the cases which follow Lemon, we find the most persuasive, in that it is on point with the matter at hand, to be Committee for Public Education v. Nyquist, 413 U.S. 756 (1973). In Nyquist, a New York State statute established, among other aid, a tuition grant program which provided for partial tuition reimbursement to low-income parents whose children attended private elementary or secondary schools. See id. at 761-70. The tuition reimbursement plan applied to parents of children who attended any private school, not solely sectarian schools, and was limited to 50% of tuition paid. Id. at 764, 93 S.Ct. 2955. At the time the plan was challenged, nearly 20% of New York’s school-age children attended nonpublic schools, and approximately 85% of these schools were sectarian. Id. at 768, 93 S.Ct. 2955. The Nyquist Court noted that although “the characteristics of individual schools may vary widely from [the] profile,” institutions which qualified for assistance under the statute were ones that included religious instruction and requirements as part of their academic curriculum. See id. at 767-68, 93 S.Ct. 2955.
Following Lemon, the Nyquist Court found that the New York statute passed the first prong of the Lemon test — whether the statute had a secular purpose-because the tuition reimbursement program promoted pluralism and diversity among New York’s public and private schools, and alleviated concern that the State’s overburdened public schools would be harmed if a large number of children who had previously been attending private schools decided to return to the public schools. See 413 U.S. at 773, 93 S.Ct. 2955. The Court did “not question the propriety, and fully secular content, of New York’s interest in preserving a healthy and safe educational environment for all of its schoolchildren.” Id.
Under the second prong of the Lemon test — that the statute neither advance nor inhibit religion-the New York statute did not fare as well. The Court next found that the New York reimbursement program failed because “[i]n the absence of an effective means of guaranteeing that the state aid derived from public funds will be used exclusively for secular, neutral, and nonideological purposes, it is clear from our cases that direct aid in whatever form is invalid.” 413 U.S. at 780, 93 S.Ct. 2955. The Court opined that the fact that program grants were delivered to parents rather than schools was “only one among many factors to be considered.” Id. at 781, 93 S.Ct. 2955. The Court rested its analysis on the premise that there had been “ ‘no endeavor to guarantee the separation between secular and religious educational functions and to ensure the State financial aid supports only the former.’ ” Id. at 783, 93 S.Ct. 2955 (quoting Lemon, 403 U.S. at 613, 91 S.Ct. 2105). “By reimbursing parents for a portion of their tuition bill, the State seeks to relieve their financial burdens sufficiently to assure that they continue to have the option to *954send their children to religion-oriented schools.” Id. The Court noted that “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.” Id.
The Nyquist Court also recognized and discarded the state’s arguments that it was of controlling significance that New York’s program called for reimbursement of tuition already paid, thus ensuring that the parent is free to spend that tuition money in any manner he or she sees fit. 413 U.S. at 785-86, 98 S.Ct. 2955. “[I]f the grants are offered as an incentive to parents to send their children to sectarian schools by making unrestricted cash payments to them, the Establishment Clause is violated.... Whether the grant is labeled a reimbursement, a reward, or a subsidy, its substantive impact is still the same.” Id. at 786, 93 S.Ct. 2955. The Court also rejected the state’s argument that the plan paid for only a portion of the tuition at a sectarian school, saying “if accepted, this argument would provide the foundation for massive, direct subsidization of sectarian elementary and secondary schools.” Id. at 787, 93 S.Ct. 2955.
The cases of Everson v. Board of Education, 330 U.S. 1, 67 S.Ct. 504, 91 L.Ed. 711 (1947), and Board of Education v. Allen, 392 U.S. 236, 88 S.Ct. 1923, 20 L.Ed.2d 1060 (1968) were carefully distinguished by the Court. In Everson, the Court upheld tax deductions for parents who expended bus fare for children who attended religious schools, reasoning that the bus fare program was analogous to the provision of services such as police and fire protection, sewage disposal, highways and sidewalks for parochial schools. See 330 U.S. at 17-18, 67 S.Ct. 504. The Court found that these services, common to all citizens, are “so separate and so indisputably marked off from the religious function, that they may fairly be viewed as reflections of a neutral posture toward religious institutions.” Id. at 18, 67 S.Ct. 504 In Allen, the Court upheld a statute which allowed secular textbooks to be provided to children attending religious schools, finding that “the State claims no right to distribute religious literature,” and noting that “we cannot assume that school authorities ... are unable to distinguish between secular and religious books.” 392 U.S. at 244-45, 88 S.Ct. 1923. The Nyquist Court distinguished these two cases not only based on their neutral posture toward religion, but on the fact that in both of those cases, “the class of beneficiaries included all schoolchildren, those in public as well as those in private schools.” 413 U.S. at 782 n. 38, 93 S.Ct. 2955, (citing Tilton v. Richardson, 403 U.S. 672, 91 S.Ct. 2091, 29 L.Ed.2d 790 (1971) (making federal aid available to all institutions of higher learning)).
The Court noted that unlike in Everson and Allen, the tuition grants in Nyquist were not a neutral attempt to provide comparable benefits to all parents of schoolchildren whether enrolled in public or nonpublic schools, as the “grants to parents of private schoolchildren are given in addition to the right that they have to send their children to public schools totally at state expense.” 413 U.S. at 782 n. 38, 93 S.Ct. 2955 (internal quotation marks omitted). The Court additionally determined that this argument of neutrality, if upheld, would be overly broad, providing “a basis for approving through tuition grants the complete subsidization of all religious schools on the ground that such action is necessary if the State is fully to equalize the position of parents who elect such schools-a result wholly at variance with the Establishment Clause.” Id.
The Supreme Court has révisited many of the issues raised in Nyquist. In Agostini v. Felton, 521 U.S. 203, 237, 117 S.Ct. 1997, 138 L.Ed.2d 391 (1997), the Court stated that “[w]e do not acknowledge, and we do not hold, that other courts should *955conclude our more recent cases have, by implication, overruled an earlier precedent.” The Court continued to reaffirm “that 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 folloio the case tohich directly controls, leaving to this Court the prerogative of overruling its own decisions.” Id. (emphasis added). The Supreme Court has refrained from overruling Nyquist, and has instead distinguished various cases on the basis of their facts; this Court has accordingly followed that approach. “A single factual difference consequently can serve to entangle or free a particular governmental practice from the reach of the [Establishment] Clause’s constitutional prohibition.” Coles, 171 F.3d at 376. We therefore look to relevant case law to assist us by analogy in analyzing the factual discrepancies between this case and Ny-quist.
In Mueller v. Allen, 463 U.S. 388, 103 S.Ct. 3062, 77 L.Ed.2d 721 (1983), the Court found constitutional a Minnesota statute which allowed state taxpayers to deduct on their state income tax certain tuition, transportation and educational expenses of their children attending elementary or secondary schools. Following the Lemon test, the Court agreed that the statute had the secular purpose of defraying the cost of education for all parents, regardless of the type of school their children attend. Id. at 395, 103 S.Ct. 3062. Analyzing the effect of the statute, the Court held that the program did not have a primary effect of advancing religion because the tax deduction was a traditional area for state legislatures to codify policies which “achieve an equitable distribution of the tax’burden,” and because the deduction was available for educational expenses incurred by all parents, whether their children attended public schools, nonsectarian private schools, or church-affiliated schools. Id. at 396-98, 103 S.Ct. 3062.
The Court found it compelling that the deduction was available for all parents with school age children, stating that this aspect was “vitally different from the scheme struck down in Nyquist. There, public assistance amounting to tuition grants was provided only to parents of children in nonpublic schools.” Id. at 398, 103 S.Ct. 3062. The Court analogized the tax deduction scheme as similar to the G.I. Bill, or other forms of “public assistance made available generally without regard to the sectarian-nonsectarian, or public-nonpublic nature of the institution benefited.” Id. The Court found it significant that under the Minnesota plan, governmental aid was only channeled through the parents, rather than directly paid to the parochial institutions, and noted that it would not base the constitutionality of a statute on the consideration of yearly statistical evidence concerning which nonsectarian schools — religious or otherwise — benefit-ted from the tax deduction. Id. at 401, 103 S.Ct. 3062. The Court stated that the Establishment Clause is not a bar to “the sort of attenuated financial benefit, ultimately controlled by the private choices of individual parents, that eventually flows to parochial schools from the neutrally available tax benefit at issue in this case.” Id. at 400,103 S.Ct. 3062.
In Witters v. Washington Department of Services for the Blind, 474 U.S. 481, 106 S.Ct. 748, 88 L.Ed.2d 846 (1986), the Supreme Court found constitutional a Washington State program which provided vocational rehabilitation assistance grants to a blind individual who attended a Christian college in the hope of becoming a pastor. The Court found the statute’s purpose “unmistakably secular,” id. at 486, 106 S.Ct. 748, and found that unlike Nyquist, “any aid provided under Washington’s program that ultimately flows to religious institutions does so only as a result of the genuinely independent and private choices of aid recipients.” Id. at 488, 106 S.Ct. 748. The fact that the vocational assistance was paid directly to the student, who could expend the educational funds on “wholly *956secular education,” persuaded the Court to find that the program did not create an incentive to apply the aid to religious education. Id. “Aid recipients’ choices are made among a huge variety of possible careers, of which only a small handful are sectarian.... Nothing in the record indicates that ... any significant portion of the aid expended under the Washington program as a whole will end up flowing to religious education.” Id. The Court also found that “the mere circumstance that petitioner has chosen to use neutrally available state aid to help pay for his religious education [does not] confer any message of state endorsement of religion.” Id. at 489, 106 S.Ct. 748.
In Agostini v. Felton, the Court held that the Establishment Clause did not bar a New York program which sent public school teachers into parochial schools to provide remedial education to disadvantaged children. Recognizing that there had been significant changes in Establishment Clause jurisprudence, the Court found that it could no longer presume “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.” 521 U.S. at 223, 117 S.Ct. 1997. The Court noted that those direct aid programs where grants are “made available generally without regard to the sectarian-nonsectarian, or public-nonpublic nature of the institution benefitted [are valid departures from the general rule] that all government aid that directly aids the educational function of religious schools is invalid.” Id. at 225, 117 S.Ct. 1997. Relying on an earlier case which had found it permissible to place a public school sign language interpreter into a private parochial school under the Individuals with Disabilities Education Act, 20 U.S.C. § 1400, the Court determined that these instances involve situations where money ultimately goes to religious schools “only as a result of the genuinely independent and private choices of individuals.” Id. at 225-26, 117 S.Ct. 1997 (citing Zobrest v. Catalina Foothills Sch. Dist., 509 U.S. 1, 113 S.Ct. 2462, 125 L.Ed.2d 1 (1993)). The Court noted as a central part of its analysis that the services provided by public school employees were remedial, stating that the “services do not, therefore, relievfe] sectarian schools of costs they otherwise would have borne in educating their students.” Id. at 228, 117 S.Ct. 1997. The Court concluded that the New York program services “are available to all children who meet the Act’s eligibility requirements, no matter what their religious beliefs or where they go to school.” Id. at 232, 117 S.Ct. 1997.
The Supreme Court revisited this issue this past term. By a plurality, the Court upheld a program to loan educational materials and equipment to private religious schools which channeled federal funds through state agencies. See Mitchell v. Helms, 530 U.S. 793, 120 S.Ct. 2530, 147 L.Ed.2d 660 (2000). Writing for four justices, Justice Thomas concluded that the critical question in cases of government aid to religious schools is whether the government aid is neutral: whether it results from the genuinely independent and private choices of individual parents. See id. at 2541-44. Justice Thomas noted that the nexus between neutrality and private choice was the prominent, even the chief factor, in upholding government aid in Agostini, Zobrest, Witters, and Mueller, and found that there is a close relationship between private choice and the question of whether a program creates a financial incentive to undertake religious schooling. See id. at 2543.
The opinion goes on to state that “[i]f aid to schools, even direct aid, is neutrally available and, before reaching or benefit-ting any religious school, first passes through the hands (literally or figuratively) of numerous private citizens who are free to direct the aid elsewhere, the government has not provided any support of religion.” 530 U.S. 793, 120 S.Ct. at 2544 *957(internal citations omitted). The opinion recognizes that there exist “special Establishment Clause dangers when money is given to religious schools or entities directly rather than, as in Witters and Mueller, indirectly.” Id. at 2546 (internal citations omitted). In a footnote, Justice Thomas hypothesized “that the principles of neutrality and private choice would be adequate to address those special risks.” Id. at 2547 n. 8. He continued to find that at least in regards to Nyquist, the prohibition against direct payments was linked to “serious concerns about whether the payments were truly neutral.” Id.
Although Justice O’Connor concurred in the judgment, she wrote separately in Mitchell based upon her belief that “the plurality announces a rule of unprecedented breadth for the evaluation of Establishment Clause challenges to government school-aid programs.” 530 U.S. 793, 120 S.Ct. at 2556. Justice O’Connor’s concurring opinion shows disagreement not only with the “expansive scope of the plurality’s rule[,]” but with two specific aspects of its analysis. Id. First, she found 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.” Id. Second, she found “the plurality’s approval of actual diversion of government aid to religious indoctrination is in tension with our precedents and ... unnecessary to decide the instant case.” Id.
While agreeing with Justice Thomas that “neutrality is an important reason for upholding government-aid programs against Establishment Clause challenges,” Justice O’Connor opined that “neutrality is not alone sufficient to qualify the aid as constitutional.” 530 U.S. 793, 120 S.Ct at 2557-58. She criticized Justice Thomas’s opinion for relying on logic which would support direct government aid to religious organizations based on the number of persons belonging to each organization. “[T]he plurality opinion foreshadows the approval of direct monetary subsidies to religious organizations, even when they use the money to advance their religious objectives.” Id. at 2560. Justice O’Con-nor rejected an outright ban on any diversion of government funds to sectarian uses, but would enact a rule which requires plaintiffs to prove that the aid in question is, or has been, used for religious purposes, and found that “presumptions of religious indoctrination are normally inappropriate when evaluating neutral school-aid programs under the Establishment Clause.” Id. at 2567.
Justice O’Connor concluded that because the school-aid program in Agostini was similar to that at issue in Mitchell, the Agostini criteria should control the outcome of the case; however, she noted that the “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.” 530 U.S. 793, 120 S.Ct. at 2560.
In regard to Mitchell and its sharply divided plurality, we note that “[w]hen a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, ‘the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds.’ ” Marks v. United States, 430 U.S. 188, 193, 97 S.Ct. 990, 51 L.Ed.2d 260 (1977) (quoting Gregg v. Georgia, 428 U.S. 153, 169 n. 15, 96 S.Ct. 2909, 49 L.Ed.2d 859(1976)). Accordingly, we find that the opinion of Justice O’Connor is the narrower of the plurality, as it utilizes the standard of Agostini based on a factual similarity rather than creating a new standard centered on neutrality, thereby making its mandates controlling.
III.
We now apply the framework established by precedent to the case before us, recognizing the predominating theme in this area of law to be the need for careful *958judicial attention to the factual detail in the challenged statutory scheme.
At the outset, we note that Defendants’ argument concerning other options available to Cleveland parents such as the Community Schools is at best irrelevant. Analyzing the scholarship program choices as compared to choices or schools outside the program is asking this Court to examine the entire context of Ohio education. Such a question is not before this Court. The Defendants’ argument would rewrite the law to require that the courts look to all possible alternatives to a challenged program, thus visiting issues of legislative choice and educational policy which no plaintiff has raised.
At oral argument, Defendants asserted repeatedly that the Community Schools program should be considered coterminous with the voucher program, arguing that the programs are merely separate sections in the statute. However, the statutory record does not support this argument, except for perhaps the literal meaning that the two programs do indeed occupy separate sections in the Ohio Code. The school voucher program is enacted as a complete program in the Ohio Revised Code. See Ohio Rev.Code §§ 3313.974-3313.983. The program is enacted as a part of the chapter on Boards of Education. See Ohio Rev.Code § 3313.01 et seq. Furthermore, the school voucher program, and only the school voucher program, was challenged by Plaintiffs in this lawsuit. In contrast, the Community Schools program is codified in its own chapter. See Ohio Rev. Code §§ 3314.01 et seq. It is similarly a complete program within the Code: the statutory provisions govern all aspects of the program without reference to the voucher program. We may not view these two programs as inextricably interdependent when the plain language of the statutory scheme demonstrates the opposite. It is simply not the proper role of the courts to change statutory construction by judicial fiat. See Brogan v. United States, 522 U.S. 398, 408, 118 S.Ct. 805, 139 L.Ed.2d 830 (1998); Nat'l Life and Accident Ins. Co. v. United States, 524 F.2d 559, 560 (6th Cir.1975) (“The Courts ... do not have the power to repeal or amend the enactments of the legislature even though they may disagree with the result; rather it is their function to give the natural and plain meaning effect to statutes .... ”). Should we consider the Community Schools program in our analysis of the constitutionality of the school voucher program, we would open the door to a wide-reaching analysis which would permit us to consider any and all scholarship programs available to children who qualify for the school voucher program: we would be considering and comparing every available option for Cleveland children. Such an analysis would expand our jurisdiction far beyond the case at hand; we are presented only with the question of whether the school voucher program violates the Establishment Clause, and we must limit ourselves to that issue, regardless of the temptations Defendants’ arguments present.
We find that Nyquist governs our result. Factually, the program at hand is a tuition grant program for low-income parents whose children attend private school parallel to the tuition reimbursement program found impermissible in Nyquist. Under both the New York statute in Nyquist, as well as the Ohio Statute at issue, parents receive government funds, either in direct payment for private school tuition or as a reimbursement for the same, and in both cases, the great majority of schools benefitted by these tuition dollars are sectarian. The Nyquist Court itself found there to be no distinction between “a reimbursement, a reward, or a subsidy, [as in all three,] the substantive impact is still the same.” 413 U.S. at 786, 93 S.Ct. 2955. As in Nyquist, the Ohio program contains no “effective means of guaranteeing that the state aid derived from public funds will be used exclusively for secular, neutral, and nonideological purposes.” Id. at 780, 93 S.Ct. 2955. *959Here, there is clearly “no endeavor to guarantee the separation between secular and religious functions and to ensure that State financial aid supports only the former.” Id. at 783, 93 S.Ct. 2955. In both Nyquist and this case, there are no restrictions on the religious schools as to their use of the tuition funds — the funds may be used for religious instruction or materials as easily as for erasers and playground equipment.
Despite the language of the statute, there is no evidence that the tuition vouchers serve as a neutral form of state assistance which would excuse the direct funding of religious institutions by the state, despite the statute’s language. Admittedly, the voucher program does not restrict entry into the program to religious or sectarian schools, but facial neutrality alone does not bring state action into compliance with the First Amendment. See Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 534, 113 S.Ct. 2217, 124 L.Ed.2d 472 (1993). The school voucher program is not neutral in that it discourages the participation by schools not funded by religious institutions, and the Cleveland program limits the schools to which a parent can apply the voucher funds to those within the program. Practically speaking, the tuition restrictions mandated by the statute limit the ability of nonsectarian schools to participate in the program, as religious schools often have lower overhead costs, supplemental income from private donations, and consequently lower tuition needs. See Martha Minow, Reforming School Reform, 68 Fordham L.Rev. 257, 262 (1999)(finding that voucher funding levels typically “approximate[ ] the tuition level set by parochial schools [which] reflects subsidies from other sources”). In fact, Defendants admit that there is incentive for private nonsectarian schools to participate in the community schools program rather than in the school voucher program. See Brief of State at 10. The evidence illustrates this point in that 82% of participating schools are sectarian, just as in Nyquist where 85% of the participating schools were sectarian. Beyond that, we note that the number of available places for students in sectarian schools is higher than 82%, as many of the sectarian schools are larger and provide a greater number of places for children in the voucher program. Moreover, close to 96% of the students enrolled in the program for the 1999-2000 school year attended sectarian institutions.
The alleged choice afforded both public and private school participants in this program is illusory in that the program’s design does not result in the participation of the adjacent public schools from outside the Cleveland school district. Per pupil expenditures in the public schools are backed by $7,097 in public funding. See Brief of Senel Taylor Intervenors at 17. At a maximum of $2,250, there is a financial disincentive for public schools outside the district to take on students via the school voucher program. Since its inception, no public schools from outside Cleveland have registered in the school voucher program, and there are no spaces available for children who wish to attend a suburban public school in place of a private school under the program. Therefore, the program clearly has the impermissible effect of promoting sectarian schools.
This is not the type of case which would fall into the exception to the Nyquist rules. Here, state assistance is only available to those students who attend private schools-the aid is clearly dependent on whether parents choose public or private schools. That the majority of places available in the program are for students attending sectarian schools is not unpersuasive. This program provides incentives for parents to choose schools other than mainstream public ones, but that choice does not extend to schools outside of the program. Students may not choose to attend community or magnet schools using a voucher, they may not apply a voucher to tuition at a private school outside the Cleveland School District, and they may not receive a voucher *960for a private school within the Cleveland School District which has not registered as part of the program. Rather, the program provides financial assistance for those parents who wish to place their children in the particular private schools, mostly religious, which take part in the program. We find such a scheme directly akin to Nyquist’s offensive aid to only private school students, and not an instance where “the class of beneficiaries included all schoolchildren, those in public as well as those in private schools.” 413 U.S. at 782 n. 38. The effect of this program, like Nyquist and unlike Mueller and subsequent cases, is one where “public assistance amounting to tuition grants was provided only to parents of children in nonpublic schools.” Mueller, 463 U.S. at 398, 103 S.Ct. 3062.
Contrary to the tax deduction generally available in Mueller, the Ohio voucher program is available to curtail only those expenses which students attending certain private schools accrue. See 463 U.S. at 398, 103 S.Ct. 3062 (distinguishing the Mueller program from that in Nyquist because in Nyquist, “tuition grants [were] provided only to parents of children in nonpublic schools”). The idea of parental choice as a determining factor which breaks a government-church nexus is inappropriate in the context of government limitation of the available choices to overwhelmingly sectarian private schools which can afford the tuition restrictions placed upon them and which have registered with the program. The absence of any meaningful public school choice from the decision matrix yields a limited and restricted palette for parents which is solely caused by state legislative structuring.
In contradistinction to Witters, a student under the Ohio statute cannot apply state aid to any school he or she chooses, including public schools, since under the Ohio program, no public schools have enrolled, nor are likely to enroll. Similarly, Agosti-ni is inapposite because the services made available to students at parochial schools through the placement of public school teachers to teach secular subjects were available to all qualifying students without regard to the nature of the institution they attended. 521 U.S. at 232, 117 S.Ct. 1997 (finding that under the New York program at issue in Agostini, and unlike the Ny-quist program, services “are available to all children who meet the Act’s eligibility requirements no matter ... where they go to school). While the program upheld in Mitchell provided for the loan of instructional equipment by state agencies to both public and private schools, in this case aid predominantly flows directly to the coffers of religious institutions. Unlike Mitchell, under the Ohio statute, there are not “numerous private citizens who are free to direct the aid elsewhere” as the majority of the choices available to parents and students are religious institutions. 530 U.S. 793, 120 S.Ct. at 2533 (Thomas, J. plurality). The voucher program at issue constitutes the type of “direct monetary subsidies to religious institutions,” that Justice O’Connor found impermissible in Mitchell. Id. at 2559-60 (O’Connor, J. concurring). This pi'ogram is dissimilar to that upheld in both Agostini and Mitchell, as here aid goes only to students enrolled in private schools, thereby fostering the type of government entanglement prohibited under the Establishment Clause.
To approve this program would approve the actual diversion of government aid to religious institutions in endorsement of religious education, something “in tension” with the precedents of the Supreme Court. Mitchell, 530 U.S. 793, 120 S.Ct. at 2556. We find that when, as here, the government has established a program which does not permit private citizens to direct government aid freely as is their private choice, but which restricts their choice to a panoply of religious institutions and spaces with only a few alternative possibilities, then the Establishment Clause is violated. This scheme involves the grant of state aid directly and predominantly to the coffers of the private, religious schools, and it is unquestioned that these institutions incor*961porate religious concepts, motives, and themes into all facets of their educational planning. There is no neutral aid when that aid principally flows to religious institutions; nor is there truly “private choice” when the available choices resulting from the program design are predominantly religious.
We conclude that unlike Mitchell, Agos-tini, Witters and Mueller, the Ohio scholarship program is designed in a manner calculated to attract religious institutions and chooses the beneficiaries of aid by non-neutral criteria. The effect of the voucher program is in direct contravention to these Supreme Court cases which mandate that the state aid be neutrally available to all students who qualify, that the parents receiving the state aid have the option of applying the funds to secular organizations or causes as well as to religious institutions, and that the state aid does not provide an incentive to choose a religious institution over a secular institution. Accordingly, we hold that no genuine issue of material fact remains for trial that the voucher program has the primary effect of advancing religion, and that it constitutes an endorsement of religion and sectarian education in violation of the Establishment Clause. We therefore affirm the district court’s order granting summary judgment to Plaintiffs.
IV.
Intervenor Taylor asserts that the district court erred by holding that the Ohio Supreme Court’s opinion in Simmons-Harris v. Goff, 86 Ohio St.3d 1, 711 N.E.2d 203 (1999), did not estop Plaintiffs’ claim, and by refusing to certify the question of estoppel to the Ohio Supreme Court. This Court reviews the issue of collateral estoppel as part of the summary judgment determination de novo. This Court reviews the district court’s denial of certification for an abuse of discretion. See Transamerica Ins. Co. v. Duro Bag Mfg. Co., 50 F.3d 370, 372 (6th Cir.1995).
Under both federal and state law, an “issue must have been necessary to support the judgment ... in the prior proceeding” in order to find collateral estoppel. Knox County Educ. Ass’n v. Knox County Bd. of Educ., 158 F.3d 361, 376-77 (6th Cir.1998); accord MetroHealth Med. Ctr. v. Hoffmann-LaRoche, Inc., 80 Ohio St.3d 212, 685 N.E.2d 529, 533 (1997) (“Issue preclusion precludes the relitigation of an issue that has been actually and necessarily litigated and determined in a prior action.”); cf. Ameigh v. Baycliffs Corp., 81 Ohio St.3d 247, 690 N.E.2d 872, 875 (1998)(“Where the judgment of a court is not dispositive on issues which a party later seeks to litigate, res judicata is not applicable ... even if the prior court decision has discussed the issues that are the subject of the current litigation.”). A determination is not essential to the judgment if the judgment could be supported by an adequate and independent state ground. See Lambrix v. Singletary, 520 U.S. 518 522-24, 117 S.Ct. 1517, 137 L.Ed.2d 771 (1997). In Goff, the Ohio Supreme Court held that the 1995 school voucher program was enacted in violation of the one-subject rule of the Ohio Constitution, and on that basis, ordered the program in its entirety “stricken” from the Ohio statute books. 711 N.E.2d at 203. That ruling entitled the plaintiffs to the relief they requested, and therefore, any discussion of other grounds for striking or upholding the statute could be neither necessary nor essential to the holding. The Ohio Court determined that the entire program could not stand; therefore, any analysis as to the constitutionality of particular portions of the program was by definition advisory or dicta, and cannot be relied upon to bar further litigation. Because Plaintiffs would not be able to obtain Supreme Court review of the Ohio Supreme Court’s determination as to the Establishment Clause, such determinations cannot constitute collateral estoppel. See Cal. v. Rooney, 483 U.S. 307, 311, 107 S.Ct. 2852, 97 L.Ed.2d 258 (1987) (declining to review a Fourth Amendment ruling adverse to *962the state of California where that ruling was unnecessary to a judgment in favor of the state).
Similarly, the district court did not err in refusing to certify the question to the Ohio Supreme Court. The governing law as to whether a party is estopped from relitigating an issue not essential to the court’s determination is clear and uncontroverted; we therefore find no abuse of discretion in the district court’s determination.
V.
Before concluding, we must pause to briefly address the dissent, not for the purpose of dignifying its hyperbole, but to quash any putatively substantive argument which may have found its way through the gratuitous insults. The dissent first makes the bald-faced assertion that the majority has struck down the voucher program as unconstitutional without any “meaningful” independent analysis, and that the majority simply concludes that the program is “foreordained” to be found unconstitutional under Nyquist According to the dissent, the New York statute in Nyquist is “totally different” from the Ohio statute before us today, thus making it impossible to “take seriously” the majority’s conclusion that Nyquist is controlling. However, even a cursory reading of the majority opinion clearly indicates that it is the dissent and its rhetoric which should not be taken seriously. As carefully set forth in Part III of this opinion, the Ohio statute at issue has the same effect as that of the New York statute held unconstitutional under the Establishment Clause in Nyquist. Both statutes have the impermissible effect of benefitting only students in particular private, and mostly religious, schools, irrespective of the illusory choice provided on the face of the Ohio statute. The fact that the dissent may not agree with the analysis set forth in the opinion to illustrate this point does not ipso facto render the analysis “meaningless.”
Second, in a similar vein, the dissent claims that the majority reaches its conclusion that the voucher program is unconstitutional under the Establishment Clause without conducting any “meaningful” analysis into the Supreme Court’s several cases on this issue since Nyquist was handed down. However, in Part II of this opinion, the majority painstakingly sets forth First Amendment Establishment Clause jurisprudence and its evolution since Nyquist, while carefully applying that law to the statute at hand in the following section. Again, it is obvious that the dissent’s bald-faced assertion that this analysis is not “meaningful” is apparently born out of nothing more than its disagreement with the outcome of this case, rather than with an objective observation. It is the majority which employs the evolving jurisprudential standards in reaching its outcome, while the dissent empicas a rigid antiquated standard to reach its result driven outcome in contravention of the Supreme Court’s latest pronouncements.1 See, e.g., Agostini, 521 U.S. at 222-26, 117 S.Ct. 1997.
VI.
We recognize the importance of this case and the precedential value it espous*963es. Equally as important, we are aware of the critical nature of questions of educational policy, and the need to establish successful schools and academic programs for children. We find, however, that even more important is the need to uphold the Constitution of the United States and, in this case, to override the State of Ohio’s statutory scheme where it constitutes an impermissible infringement under the Establishment Clause of the First Amendment. We therefore AFFIRM the district court’s order finding the school voucher program unconstitutional, as well as the court’s determination that Plaintiffs are not collaterally estopped.
. Judge Ryan inappropriately calls for an en banc review of the matter in his dissent. There are rules and procedures governing a call for en banc review once a case has been decided, whether the call is made by a party or sua sponte by an active judge of this Court or member of the original panel, and it is the process contemplated by these rules and procedures which should be used to invoke en banc review after the Court's opinion has been issued. See Craft v. CIR, 233 F.3d 358, (6th Cir.2000) (criticizing the concurring opinion's exhortation for en banc review of the matter, while noting that Fed.R.App.P. 35(b) and 6 Cir. I.O.P. 35(c) set forth the appropriate procedures to follow when calling for such review). The unfortunate practice of arguing for en banc review of a case in a panel member’s separate opinion — instead of simply permitting the Court's regular operating procedures for seeking such review to be followed subsequent to the issuance of the majority opinion — is one which should not be perpetuated.