concurring in part and dissenting in part.
CONCURRING IN PART, DISSENTING IN PART
My colleagues’ resolution of the question presented by the plaintiffs’ collateral es-toppel claim is eminently correct and so I join part IV of the court’s opinion. However, because I believe Ohio’s voucher program to be constitutional under the First Amendment and the Supreme Court’s Establishment Clause cases interpreting the amendment, I must respectfully dissent from the majority’s treatment of the voucher program’s constitutionality.
My brothers have struck down as unconstitutional Ohio’s effort to establish a school-choice voucher program whose primary purpose is to enable mostly minority poverty-level school children, in Cleveland, Ohio, to escape the devastating consequences of attending Cleveland’s demonstrably failed public schools. My colleagues have done so not on the basis of any independent constitutional analysis of the Ohio Pilot Project Scholarship Program, as the voucher program is formally known, but because they claim the invalidity of the statute is a conclusion foreordained by the United States Supreme Court’s decision in Committee for Public Education and Religious Liberty v. Nyquist, 413 U.S. 756 (1973). I disagree. The New York statute interpreted in Ny-quist and the Ohio statute before us are totally different in all of their essential respects, both in their purposes and their provisions for carrying out their respective purposes. It is impossible to take seriously the majority’s claim that Nyquist governs our result and, for that reason, requires that the Ohio voucher program must be struck down.
Moreover, the majority’s refusal to conduct any meaningful analysis of the Supreme Court’s several Establishment Clause decisions handed down in the 27 years since Nyquist was decided, its insistence that the plainly distinguishable Ny-quist case is directly on point, and the factually unsupported antireligious-schools arguments in the opinion strongly suggest that the majority has simply signed onto the familiar anti-voucher mantra that voucher programs are no more than a scheme to funnel public funds into religious schools.
I.
It is implicit in the majority’s reasoning that there is no need for any independent analysis whether Ohio’s voucher program violates the Establishment Clause because Nyquist is “on point with the matter at hand.” Maj. op. at 953. In my judgment, the majority is mistaken as a matter of fact (the two statutes are totally different) and as a matter of laxo (the relevant Establishment Clause jurisprudence has changed since Nyquist). As to the latter, a reading of the Supreme Court’s Establishment Clause cases decided since 1973 makes it unmistakably clear that the voucher program passes constitutional muster. I do not claim that the Nyquist decision has been overruled, although some of the reasoning in the Nyquist opinion has been “undermined,” as the Ohio Supreme Court put it in Simmons-Harris v. Goff, 86 Ohio St.3d 1, 711 N.E.2d 203, *964208 (Ohio 1999); Nyquist is simply inappo-site to the appeal before us.
The New York statutory provisions struck down in Nyquist and the Ohio voucher program are essentially different laws; they are plainly distinguishable both in their declared purposes and in the manner of their application. For that reason alone, the reasoning and the holding of the Nyquist decision cannot govern our result.
A.
I begin with a comparison of the New York statutory provisions construed in Ny-quist and the Ohio statute before us; a comparison that shows very clearly that the two laws are essentially different. I then examine the Supreme Court Establishment Clause cases decided since Ny-quist, which clearly indicate that the Ohio voucher program is not unconstitutional.
1.
In Nyquist, the Supreme Court was required to decide whether a New York statute containing provisions for both direct and indirect financial assistance to New York’s private schools violated the Establishment Clause. The statute provided for three forms of assistance: (1) direct grants for building maintenance and repairs for private school buildings; (2) tuition reimbursement grants for some low-income parents of children already attending the private schools; and (3) a form of tax relief for parents who failed to qualify for tuition reimbursement under the statute. See Nyquist, 413 U.S. at 762-65, 93 S.Ct. 2955.
The New York legislature enacted the statute for the sole purpose of directly benefitting New York state’s 2,038 financially pressed private schools, wherein some 700,000-800,000 students — almost 20% of the state’s entire elementary and secondary school population — were being educated. See id. at 768, 93 S.Ct. 2955. The legislative “findings” in the New York statute declared: (1) it was in the state’s .interest to provide funding for “maintenance and repair” of the state’s private schools in order to protect the health and safety of those attending the schools; (2) the state had an interest in promoting “alternative educational systems”; and (3) a “precipitous decline” in the number of children attending private schools would perpetuate an already existing fiscal crisis in public schools. Id. at 763-65, 93 S.Ct. 2955.
The Nyquist Court held that the New York law offended the Establishment Clause because “the effect of the aid [was] unmistakably to provide desired financial support for nonpublic, sectarian institutions.” Id. at 783, 93 S.Ct. 2955. Furthermore, the Nyquist Court concluded that “[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.” Id. at 780, 93 S.Ct. 2955.
The Ohio voucher program, which is adequately described in the majority opinion, could not be more unlike the New York statute both in its purpose and in the manner of its application. The essential differences between the New York and Ohio statutes may be summarized as follows:
First, the purpose of the New York statute was to provide financial help to New York’s financially troubled private schools because their closing would force New York’s public schools to absorb the private school students, resulting in massive increased costs and the related burdens of absorbing as many as three quarters of a million new students.
The purpose of the Ohio statute, on the other hand, is to provide financial help to poverty-level students attending the public schools in Cleveland in order to enable them, if they wish, to attend nonreligious private schools, religious private schools, public schools in neighboring districts that wish to participate in the voucher pro*965gram, or to obtain special tutoring while remaining in the Cleveland public schools.
Second, the New York program involved direct financial grants to New York’s private schools, religious and nonreligious, primarily for maintenance and repair. Although the tuition reimbursement and tax relief sections of the statute appeared to benefit the parents of private school children, the Nyquist Court stated that the “tuition reimbursement program also fails the ‘effect’ test, for much the same reasons that govern its maintenance and repair grants.” Id.
Under the Ohio voucher program, on the other hand, there is no provision for any financial grants in any form to any private schools. A voucher recipient receives a scholarship cheek, and the funds therefrom reach a private religious school only after a child’s parents have considered a variety of options available to them and have chosen the religious private school as the best option for their child.
Third, the New York statute permitted government aid to schools that discriminated against children on the basis of religion and, in fact, several qualifying schools imposed religious restrictions on admissions. See id. at 767-68, 93 S.Ct. 2955.
The Ohio voucher program, on the other hand, contains a provision explicitly forbidding participating schools from discriminating against prospective students on the basis of religion. See Ohio Rev.Code § 3313.976(A)(4).
It is clear that the New York statute struck down in Nyquist and the Ohio statute before us are dissimilar laws both in their purposes and the methodologies for carrying out their purposes. As the majority acknowledges, “[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.” Maj. op. at 955 (internal quotation marks and citation omitted). A case construing a statute so manifestly different than the one before us could hardly, as a factual matter, be a binding precedent on this court.
2.
The substantial differences in the purpose and application of the two statutes is not the only reason Nyquist does not govern our result. The additional reason is that the rule of law upon which Nyquist was decided has changed. First, the Ny-quist era categorical prohibition against direct grants to aid religious schools is no longer the law; and second, the criteria for determining when a statute has the forbidden “primary effect” of advancing religion have been modified.
In Lemon v. Kurtzman, 403 U.S. 602, 91 S.Ct. 2105 (1971), the Supreme Court fashioned the following test for assessing whether a statute violates the Establishment Clause:
First, the statute must have a secular legislative purpose; second, its principal or primary effect must be one that neither advances nor inhibits religion ...; finally, the statute must not foster an excessive government entanglement with religion.
Id. at 612-13, 91 S.Ct. 2105 (emphasis added) (internal quotation marks and citation omitted).
The Nyquist Court ruled that the New York statute violated the Lemon test because it had the “impermissible effect of advancing religion.” Nyquist, 413 U.S. at 794. It did so, the Court said, by providing direct financial assistance to religious schools without any restrictions as to the schools’ use of the funds, therefore “advancing] the religious mission of sectarian schools.” Id. at 779-80, 93 S.Ct. 2955. But three years ago in Agostini v. Felton, 521 U.S. 203, 117 S.Ct. 1997, 138 L.Ed.2d 391 (1997), the Supreme Court declared unmistakably that “we have departed from the rule ... that all government aid that directly assists the educational function of religious schools is invalid.” Id. at 225, *966117 S.Ct. 1997. The Agostini Court then proceeded to redefine and narrow the criteria for determining when government aid that finds its way to a religious school has the primary effect of advancing religion.
Again, I do not question for a moment the correctness of the Supreme Court’s decision in Nyquist. I accept it both analytically and precedentially as a faithful 1973 application of the “primary effect” test of Lemon. However, Nyquist was not analyzed and decided under what the Agostini Court called its “changed ... understanding of the criteria used to assess whether aid to religion has an impermissible effect.” Id. at 223, 117 S.Ct. 1997. Since this appeal is also an “impermissible effect” case, our decision cannot be controlled by Nyquist.
B.
What then is the Supreme Court’s “changed ... understanding” of the proper test for determining whether a law has the primary effect of advancing religion?
In Mueller v. Allen, 463 U.S. 388, 103 S.Ct. 3062, 77 L.Ed.2d 721 (1983), the Court held that a Minnesota statute authorizing a tax deduction for certain educational expenses for parents of students attending either public or private schools, religious or nonreligious, did not violate the “impermissible effect” prong of the Lemon test. The Court focused on the fact that the deduction was given directly to the parents, without regard to the type of school, religious or nonreligious, to which the parents might choose to send their children, as a strong indicator of the statute’s “neutrality.” See id. at 397-400, 103 S.Ct. 3062. Any money received at a religious school, the Court said, was “ultimately controlled by the private choices of individual parents.” Id. at 400, 103 S.Ct. 3062.
This principle — that whether public funds find their way to a religious school is of no constitutional consequence if they get t there as a result of genuinely private choice — was reasserted in Witters v. Washington Department of Services for the Blind, 474 U.S. 481, 106 S.Ct. 748, 88 L.Ed.2d 846 (1986). There, a Washington state program survived an Establishment Clause challenge even though it provided vocational rehabilitation assistance for a blind individual to attend a Christian college in order to study to be a Christian pastor. Funds under the program were dispersed directly to the eligible applicants who made the choice of where to expend the educational funds; therefore, “[a]ny aid ... that ultimately flow[ed] to religious institutions [did] so only as a result of the genuinely independent and private choices of aid recipients.” Id. at 488, 106 S.Ct. 748.
In Zobrest v. Catalina Foothills School District, 509 U.S. 1, 113 S.Ct. 2462, 125 L.Ed.2d 1 (1993), the Court upheld the constitutionality of a program providing a sign-language interpreter for a deaf student in a Catholic high school. Relying upon Witters and Mueller, the Zobrest Court concluded that the statute gave parents the choice of where to send their eligible children to school and “distributes benefits neutrally ... without regard to the ‘sectarian-nonsectarian, or public-nonpublic nature’ of the school.” Id. at 10, 113 S.Ct. 2462 (quoting Witters, 474 U.S. at 487, 106 S.Ct. 748).
This line of cases culminated in the Agostini decision in 1997, in which the Supreme Court declared that its understanding of the criteria for determining whether, in any specific program, government aid has the primary effect of advancing religion had “changed.” Indeed, in Agostini the Supreme Court went so far as to modify the Lemon test it had relied upon in Nyquist. The Agostini Court began by recasting Lemon’s “entanglement” inquiry as a factor under the “impermissible effect” prong rather than as a separate and independent criterion. See Agostini, 521 U.S. at 232-34, 117 S.Ct. 1997. It then identified three new sub-criteria to consider when evaluating whether a government-*967aid program violates Lemon's “impermissible effect” prong. These are:
(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.
See id. at 234.
Using this modified Lemon test, the Agostini Court found constitutional a federally mandated New York program that sent public school teachers into private parochial schools to provide remedial education to eligible children. Under the program, children meeting the eligibility requirements received the services, whether they attended a private or public school. See id. at 232, 117 S.Ct. 1997. The Agostini Court concluded that programs in which money ultimately flows to a private, religious school based on the “ ‘genuinely independent and private choices of individuals” do not violate the Establishment Clause. Id. at 226, 117 S.Ct. 1997 (quoting Witters, 474 U.S. at 488, 106 S.Ct. 748).
Finally, in Mitchell v. Helms, 530 U.S. 793, 120 S.Ct. 2530, 147 L.Ed.2d 660 (2000), a plurality opinion written by Justice Thomas, the Court upheld the constitutionality of a federally mandated Louisiana program where educational materials were loaned to public and private schools, both religious and nonreligious. Justice Thomas emphasized that the statute did not have an “impermissible effect” because the “principles of neutrality and private choice, and their relationship to each other [that] were [also] prominent” in the Court’s decisions in Agostini, Zobrest, Witters, and Mueller were present. Id. at 2542.
II.
It is against this background of changed Supreme Court Establishment Clause jurisprudence that we must test the constitutionality of the Ohio voucher program.
The Ohio statute is the product of a 1994 order issued by the United States District Court in Cleveland, directing the Ohio Superintendent of Education to address the educational crisis in Cleveland’s public schools. See Reed v. Rhodes, 869 F.Supp. 1274 (N.D.Ohio 1994). The Ohio legislature and the state’s governor responded with the voucher program that is before us today. See Ohio Rev.Code §§ 3313.974-3313.979. We may safely assume that in fashioning the new law, the Ohio legislators and the governor knew that the challenge they faced was to design a law that would survive a federal constitutional challenge on Establishment Clause grounds. That is not to say that the statute the legislators wrote and the governor signed into law is insulated from federal judicial constitutional scrutiny. Rather, it is to say what the majority does not even acknowledge: this statute is presumed to be constitutional. See McDonald v. Board of Election Comm’rs, 394 U.S. 802, 809, 89 S.Ct. 1404, 22 L.Ed.2d 739 (1969); Hartford Fire Ins. Co. v. Lawrence, Dykes, Goodenberger, Bower & Clancy, 740 F.2d 1362, 1366 (6th Cir.1984). This presumption is not a mere literary figure for rote recitation in all appellate opinions addressing the constitutionality of legislative enactments; it is a bedrock rule of statutory construction, one we are bound assiduously to honor as we begin our assessment of the validity of the Ohio statute.
The first of the Lemon criteria that must be met if a statute is to survive an Establishment Clause challenge is that it have a “secular purpose.” The Ohio voucher program meets this criterion and the plaintiffs agree that it does. The sole purpose of the voucher program is to save Cleveland’s mostly poor, mostly minority, public school children from the devastating consequences of requiring them to remain in the failed Cleveland schools, if they wish to escape. There is also no serious claim that the statute is constitutionally invalid solely because it fosters an “excessive entanglement” between government *968and religion. Rather, the only issue in the case is whether the voucher program has the forbidden “primary effect” of advancing religion. This court’s first duty, therefore, after recognizing that Nyquist’s factually and legally outdated decision is of no help, is to proceed to examine the first two criteria from Agostini’s “impermissible effect” test to determine whether the effect of Ohio’s voucher program is to advance religion, either because (1) the aid it provides results in governmental indoctrination, or (2) the program defines its recipients by reference to religion. See Agostini, 521 U.S. at 234, 117 S.Ct. 1997. These are the only two issues properly before us.
A.
In addressing Agostini’s, first criterion for testing a statute’s claimed impermissible effect, we must ask whether the government aid in the form of the tuition voucher results in “governmental indoctrination.” It is obvious that the Ohio statute does not have the remotest effect of providing governmental indoctrination in any religion, to say nothing of having such a primary effect.
The Supreme Court decisions since Lemon and Nyquist have emphasized that the critical question in determining whether government aid ultimately flowing to religious schools results in governmental indoctrination is if the recipient beneficiaries make a “genuinely independent and private choice[ ]” to “spend” the funds in a religious school. Id. at 226, 117 S.Ct. 1997 (internal quotation marks and citation omitted). If the recipients have such an independent and private choice, then the government’s decision to provide the money to fund that choice does not have the effect of advancing religion. The government is, of necessity, neutral in the matter. Implicit in that constitutional rule of law, as it applies in this case, is that there must be a genuine choice from among a range of alternatives that indicate complete neutrality on the part of the government as to where the recipient parents may choose to spend the government-aid funds. The voucher program does not offend the Establishment Clause because the statute allows parents to make a genuine choice for their children who are currently in Cleveland public schools.
What are the choices Ohio has given these Cleveland parents?
(1) To permit their children to remain in the Cleveland public schools as before;
(2) To accept a tuition voucher for them to attend a Cleveland area nonreligious private school;
(3) To accept a tuition voucher for them to attend a Cleveland area religious private school;
(4) To accept a voucher for them to obtain special tutorial help in the Cleveland schools; or
(5) To accept a voucher for them to attend a public school in a district adjacent to Cleveland, although for the present these districts have declined to participate in the program. See Ohio Rev.Code §§ 3313.976-3313.978.
It is difficult to imagine a statute that could afford its voucher recipients a broader spectrum of educational choice. It is true, of course, that the public school districts adjacent to Cleveland have declined to participate in the voucher program, but there is not the slightest hint in the record that when the Ohio statute was enacted either the legislators or the governor had any idea that the public school districts adjacent to Cleveland would not participate. What we measure today is not whether the children in Cleveland have the fullest conceivable range of options available to them that a panel of federal judges might think to be ideal, but rather, whether the statute, as enacted, has the primary effect of advancing religion by involving the state in governmental indoctrination under Agostini’s first criterion. See Mitchell, 530 U.S. 793, 120 S.Ct. at 2541-*96944. To my knowledge, no federal court has ever held that a school-choice voucher program is unconstitutional because the range of choices does not include a public school option; certainly the majority does not cite such a case.
B.
Neither does the Ohio program “define its recipients by reference to religion,” the second Agostini factor for testing for “impermissible effect.” Agostini, 521 U.S. at 234, 117 S.Ct. 1997. The program defines the first-priority voucher recipients by reference to (1) their attendance in one of Cleveland’s public schools; and (2) a family income that is not more than 200 percent of the federally established poverty level. See Ohio Rev.Code § 3313.978(A). And the statute explicitly forbids a religious test for admission to a participating school, including religious schools. See id. at § 3313.976(A)(4). A parent has the choice of using the voucher in a private religious school, a private nonreligious school, for tutoring in the public school, or in a public school in a neighboring district if any wish to participate. The statute expresses no preference, explicitly or implicitly, either as to the religion of the voucher recipients, or if the recipient chooses a private school, whether the voucher is applied to a religious or nonreligious school.
The Agostini Court recognized, of course, that the eligibility requirements of a government-aid program could “have the effect of advancing religion by creating a financial incentive to undertake religious indoctrination.” Agostini, 521 U.S. at 231, 117 S.Ct. 1997. The Court noted that a financial incentive to choose a religious school over a nonreligious school is not present “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.” Id.
Despite the plain evidence that the aid to the parents of the Cleveland school children is indeed “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,” id., the majority continues to insist that the voucher program is not neutral because it creates a forbidden “incentive” for parents in Cleveland to choose a religious school. As best I can understand it, they rest this conclusion — unsupported though it is by any evidence in the record — on two further conclusions. The first is that because the vast majority — 82 percent — of the private schools participating in the Ohio program are religious, the people of Cleveland are denied a “genuine” choice. This argument is made despite the indisputable fact that of all the private nonreligious private schools participating in the program, not one has ever turned away a voucher applicant for any reason. This not very thinly veiled antipathy the majority has shown toward religious schools — its argument that there are too many religious schools in the program — is meritless for another reason: the Supreme Court has flatly rejected the argument that a high percentage of religious schools participating in a government-aid program is an indicator that the government is engaging in governmental indoctrination of religion. See Mitchell, 530 U.S. 793, 120 S.Ct. at 2542; id. at 2562 (O’Connor, J., concurring); Agostini, 521 U.S. at 229, 117 S.Ct. 1997; Mueller, 463 U.S. at 401, 103 S.Ct. 3062.
The majority then attempts to support its view that the Ohio statute creates a forbidden incentive for parents to choose a religious school by utilizing the transparent argument that this statute should be struck down because the religious schools in the program are too religious. In support of this proposition, the majority devotes considerable attention to the mission statements of several religious schools, which indicate the pervasively religious character of their programs. My brothers *970conclude therefrom that these schools “believe in interweaving religious beliefs with secular subjects” and “incorporate [in their curriculum] religious concepts, motives, and themes.” Maj. op. at 949, 961. Imagine, religious schools that are truly religious!
This plainly hostile attack on the religious schools in the Ohio voucher program is one 1 would of thought unworthy of mention in an opinion from this great court. Is the point being made here that religious schools may participate in a voucher program providing they are not too religious? Or, is it that these poverty-level parents in Cleveland cannot be trusted to understand what they will be exposing their children to if they choose one of these religious schools? One would have thought that the nail was long ago driven into the coffin bearing the discredited arguments that if a voucher program involved too many religious schools, or if those involved are honestly, genuinely, and essentially religious, the statute is therefore invalid as “advancing religion.” This most unattractive argument was utterly rejected in Witters, 474 U.S. at 486-88, 106 S.Ct. 748, and Mueller, 463 U.S. at 397-400, 103 S.Ct. 3062, and also was rejected in Justice Powell’s concurring opinion in Witters. See Witters, 474 U.S. at 492, 106 S.Ct. 748 (Powell, J., concurring). Moreover, Justice Thomas, writing for a clear majority on this point in Mitchell, stated:
In short, nothing in the Establishment Clause requires the exclusion of pervasively sectarian schools from otherwise permissible aid programs, and other doctrines of this Court bar it. This doctrine, born of bigotry, should be buried now.
Mitchell, 530 U.S. 793, 120 S.Ct. at 2552.
The majority, in this case, straining mightily to strike down this law, then conjures still another anti-voucher argument (the reader will recall that the majority’s decisional premise is that this case is controlled by Nyquist and, implicitly, that all else is irrelevant).
My colleagues’ next non-Nyquist argument is that the “school voucher program is not neutral in that it discourages the participation by schools not funded by religious institutions.” Maj. op. at 959. This statement in the majority opinion, which, like so many others in the opinion, is totally without any basis in the evidence, is then fortified by my brothers’ ipse dixit that “religious schools often have lower overhead costs, supplemental income from private donations, and consequently lower tuition needs.” Id. at 959. The only authority my colleagues offer for this speculation is a Fordham University Law Review article. I can only surmise that the point my colleagues wish to make here is that nonreligious schools will not participate in Ohio’s voucher program because the voucher will not cover the cost of educating a student in a nonreligious school. This, my brothers reason, creates an “incentive” for the parents to send their children to the religious schools where they can be educated more efficiently and for fewer dollars.
There is absolutely no evidence in the record to support the majority’s argument that the Ohio statute creates a financial “disincentive” for Cleveland’s neighboring, suburban public school districts to participate in the program. There is no evidence to support what the majority seems to imply — that some parents wishing to use a voucher choose not to do so because other Cleveland area public schools are not participating. And there is no evidence that if a public school chooses to participate in the voucher program, it will lose its state funding.
These arguments are built on a “factual” predicate that has absolutely no basis in the record. There is not a scintilla of evidence in this case that any school, public or private, has been discouraged from participating in the school voucher program because it cannot “afford” to do so. The import of this argument, as best I can *971understand it, is that the parents of the Cleveland school children have an “incentive” to choose Cleveland’s religious schools because there are not enough nonreligious schools participating in the program. Of course, there is no evidence of that either. And there is no evidence that any of the several nonreligious, private schools participating in the program have ever rejected a single voucher applicant for any reason, including a supposed inability to afford the differential between the value of a $2,500 voucher and the actual cost of a nonreligious, private school education.
While I hesitate to dignify the majority’s speculation with speculation of my own, what is at stake in this case is too important to let any of my colleagues’ meritless arguments go unanswered.
It is indisputable that no nonreligious, private school, or any other school for that matter, has ever been discouraged from participating in the Cleveland voucher program and there is no evidence that any private school, religious or nonreligious, has ever turned away a voucher applicant for any reason. Therefore, what my colleagues must be getting at is even more insidious and offensive. The point apparently is that Cleveland parents would never choose to send their children to a religious school in Cleveland if they could afford to send them to a nonreligious, private school, or another public school, but that they cannot do so because the cost differential between the value of a $2,500 voucher and the actual tuition of Cleveland’s nonreligious, private schools is prohibitive. Again, it is of no small importance that there is absolutely no evidence in the record that any Cleveland public school parent has ever declined to enroll his or her child in a nonreligious, private school in Cleveland because there was a differential cost that was prohibitive. It is probably true that no private school, religious or nonreligious, can educate a child for the voucher value of $2,500. But, in all probability, the participating private schools are willing to accept the voucher as meeting a portion of the actual educational costs for these children and are willing to absorb the differential cost as part of their pro bono service in Cleveland to help save as many of these children as possible from the disastrous consequences of continuing in the city’s failed public schools.
But more important than all of this speculation is the reality that the majority’s “neutral only if affordable to all” test is utterly meritless as a matter of law because the now settled Establishment Clause jurisprudence is that whether aid is allocated on the basis of neutral, secular criteria is the key determinant of whether, having made the aid available, the state has engaged in governmental indoctrination in religion. See Agostini, 521 U.S. at 231, 117 S.Ct. 1997.
If the simplicity and clarity of the Supreme Court’s language in Agostini is not sufficient to demonstrate that the Ohio statute does not in any respect operate to advance religion, confirmatory language of crystal clarity appears in the Supreme Court’s recent decision in Mitchell. In a passage in his opinion which enjoys the support of a majority of the Justices, and arguably even the support of the dissenters, Justice Thomas states:
[T]he question whether governmental aid to religious schools results in governmental indoctrination is ultimately a question whether any religious indoctrination that occurs in these schools could reasonably be attributed to governmental action.
Mitchell, 530 U.S. 793, 120 S.Ct. at 2541.
The line of cases decided in the Supreme Court beginning with Mueller in 1983 and ending with Mitchell in 2000 make it unmistakably clear that the majority’s “impermissible incentive” argument has no basis in our Establishment Clause jurisprudence. The rule is now settled that a government program that permits financial aid ultimately to reach religious schools does not offend the Establishment Clause if the government’s role in the pro*972gram is neutral. Neutrality exists if the “governmental aid that goes to a religious institution does so ‘only as a result of the genuinely independent and private choices of individuals.’ ” Id. (quoting Agostini, 521 U.S. at 226, 117 S.Ct. 1997). Justice Thomas wrote that “simply because an aid program offers private schools, and thus religious schools, a benefit that they did not previously receive does not mean that the program, by reducing the cost of securing a religious education, creates ... an ‘incentive’ for parents to choose such an education for their children.” Id. at 2543-44. Finally, Justice Thomas concluded' that the possibility that government aid might be diverted by a sectarian school towards some religious end is irrelevant, for Establishment Clause purposes, if the government aid program provides the aid in a neutral manner. See id. at 2547.
In her concurring opinion in Mitchell, Justice O’Connor, joined by Justice Breyer, agreed that “neutrality is an important reason for upholding government-aid programs” against Establishment Clause challenges, but she reiterated that it was just one factor to consider in challenges to government school-aid programs and not a “factor [of] singular importance.” Id. at 2556-57 (O’Connor, J., concurring). Even Justice Souter’s dissenting opinion, which Justices Stevens and Ginsburg joined, conceded that the Establishment Clause presents no obstacle to government aid if it reaches sectarian schools as a result of the private choices of aid recipients. See id. at 2584 (Souter, J., dissenting).
Justice O’Connor emphasized the distinction between “true private-choice programs” and “per-capita school-aid programs.” Id. at 2559 (O’Connor, J., concurring). The Ohio voucher program, like the programs in Zobrest and Witters, is a true private-choice program because the aid is given directly to eligible individuals, who in turn decide where to spend it. The programs considered in Mitchell and Agostini were examples of per-capita school-aid programs because aid was distributed based on the number of students attending each school, regardless of whether the school was religiously based or not. Justice O’Connor concluded that true private-choice programs were more likely to survive Establishment Clause challenges, even though government aid was diverted to the religious schools, because “ ‘[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.’” Id. at 2558 (quoting Witters, 474 U.S. at 488, 106 S.Ct. 748) (O’Connor, J., concurring).
True private-choice programs, by their very nature, cannot have the forbidden “primary effect” of the government “advancing religion” because the aid is given directly to the beneficiary and that student or parent retains control over where the aid will be applied. “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.” Id. at 2559. Furthermore, when government aid flows to a religious school as a result of “independent decisions made by numerous individuals ..., [n]o reasonable observer is likely to draw from the facts ... an inference that the State itself is endorsing a religious practice or belief.” Id. (internal quotation marks and citation omitted).
The majority opinion in this case claims the voucher program “involves the grant of state aid directly and predominantly to the coffers of the private, religious schools.” Maj. op. at 960. Furthermore, according to the majority, “[t]here is no neutral aid when that aid principally flows to religious institutions....” Id. at 961. The majority ignores that this view has been flat-out rejected by the Supreme Court in the decisions I have discussed which make it very clear that the number of religious schools participating in the voucher program, the thoroughness of the religious training that occurs there, and the use to which such *973schools might put the funds are all totally irrelevant to the question of the government’s neutrality, when the government aid reaches a religious school only as a result of the recipient’s “ ‘genuinely independent and private choice[ ].’ ” See Mitchell, 530 U.S. 793, 120 S.Ct. at 2541 (citation omitted); Zobrest, 509 U.S. at 8, 113 S.Ct. 2462; Witters, 474 U.S. at 486, 106 S.Ct. 748. By ignoring the recent Supreme Court cases emphasizing the importance of genuinely independent and private choices and the distinction between true private-choice programs and per-capita school-aid programs, the majority has failed to conduct a “meaningful” independent analysis of the voucher program’s constitutionality.
III.
In summary, and to repeat, according to the Supreme Court, a true private-choice program does not result in “governmental indoctrination” so long as the path of the government aid is determined by the “ ‘genuinely independent and private ehoice[ ]’ ” of the aid recipients. See Mitchell, 530 U.S. 793, 120 S.Ct. at 2541 (citation omitted); id. at 2557-60 (O’Connor, J., concurring); Agostini, 521 U.S. at 226, 117 S.Ct. 1997; Zobrest, 509 U.S. at 10, 12, 113 S.Ct. 2462; Witters, 474 U.S. at 488, 106 S.Ct. 748; Mueller, 463 U.S. at 397-98, 400, 103 S.Ct. 3062.
Ohio’s voucher program easily meets this test. Before a voucher is “spent” at a religious school, Cleveland parents must independently make two important choices:
First, they must decide whether their child will take advantage of the voucher alternatives at all, or select another option, such as remaining in the Cleveland schools, undertaking home schooling, or attending one of Cleveland’s well regarded community schools. Second, if a child’s parents choose the voucher option, they must make the further “genuinely independent and private choice[ ]” whether to use the voucher at a private school, nonreligious or religious, or for special tutoring in the Cleveland public schools. The voucher-use choice of attending a public school in a neighboring district is not presently available to Cleveland parents because no neighboring district has opted into the voucher program.
It is difficult to imagine how a voucher statute could be crafted that more clearly and decisively forecloses the government from having any role in the religious indoctrination of Cleveland school children, or forecloses it from defining the recipients of the vouchers by reference to religion, than through the range of free and independent choices the statute gives to the parents whose children attend the Cleveland public schools.
IV.
In striking down this statute today, the majority perpetuates the long history of lower federal court hostility to educational choice. It does so by reaching back to a 1973 Supreme Court decision, Nyquist, that construes a statute that is light years away from the voucher program before us and that rests upon law that has been altered in an important respect by subsequent Supreme Court decisions. My colleagues refuse to acknowledge that the program in Nyquist is factually distinguishable in essential ways from the Ohio voucher program and that the Supreme Court has explicitly declared that the criteria for determining whether a statute authorizing government aid to schools violates the Establishment Clause have changed. And then, almost as if recognizing that its Ni/gmsi-is-directly-on-point argument cannot withstand close scrutiny, the majority resorts to the lamentable tactic of attempting to arouse support for its view by making the familiar but unworthy arguments that the voucher program has too many religious schools and that they are too religious. This argument should fail, first, because it is rooted in nativist hostility toward religious schools and, second, because it has been explicitly rejected *974by the Supreme Court as a legitimate determinant of whether a government is engaging in religious indoctrination.
Despite the majority’s disclaimer that “courts do not make educational policy; we do not sit in omnipotent judgment as to the efficacy of one scheme or program versus another,” maj. op. at 951, the majority opinion is nothing more than an attack upon the philosophical and cultural desirability of publicly funded educational choice for the poor. This case and its result-sentencing nearly 4,000 poverty-level, mostly minority, children in Cleveland to return to the indisputably failed Cleveland public schools from which, in many cases, they escaped as long as three years ago-is an exercise in raw judicial power having no basis in the First Amendment or in the Supreme Court’s Establishment Clause jurisprudence.
In all events, a matter of this gravity and of such immense importance to the Cleveland children who are directly affected, and indeed to the nation, should not be determined by just two judges of this court. Therefore, I respectfully urge my colleagues to take this case for en banc review, if they are asked to do so, and decide the vitally important Establishment Clause issue it presents, after giving careful consideration to the full panoply of Supreme Court Establishment Clause jurisprudence, and not just one, inapposite 1973 case.
As to what is written in part IV of the majority opinion, I have no disagreement.