THE WORST CHOICE FOR SCHOOL CHOICE: TUITION TAX CREDITS ARE A BAD IDEA AND DIRECT FUNDING IS WISER Michael J. Broyde* and Anna G. Gabianelli** Abstract School choice is on the rise, and states use various mechanisms to implement it. One prevalent mechanism is also a uniquely problematic one: the tax credit. Tax credits are deficient at equitably distributing a benefit like school choice; they are costly, and they invite fraud. Instead of using tax credits, states opting for school choice programs should use direct funding. Direct funding will more efficiently achieve the goals of school choice because it can be regulated like any other government benefit, even if it ends up subsidizing religious private schools. Tax credits’ prevalence is not inexplicable, of course. It is based on a prior legal understanding that states were constitutionally restricted from directly funding religious schools. Historically, states that wanted to include religious private schools in their school choice programs therefore felt pushed to use tax credits as their only constitutionally viable option. However, the landscape has changed. The Supreme Court held in 2022 that direct funding of religious private schools is not only constitutionally permissible, but it is required if a state funds non-religious private schools and provides no neutral basis for excluding religious ones. The initial reason for tax credits’ popularity therefore no longer exists; both tax credits and direct funding alike are constitutionally acceptable. It is time, therefore, to revisit the merits of tax credits and ask whether, knowing what we know now, it is worth disposing of them in favor of direct funding. This Article answers that question with a resounding yes. Tax credits carry significant disadvantages—specifically, inequitable distribution and difficulties in regulation—that direct funding does not. Now that the law is clear, states choosing to sponsor school choice should discontinue their use of tax credits in favor of direct funding. * Michael J. Broyde is a Professor of Law at Emory University, the Berman Projects Director in its Center for the Study of Law and Religion, and the Director of the SJD Program at Emory University School of Law. ** Anna G. Gabianelli graduated with High Honors from Emory Law in 2023, currently works in New York City, and will serve as a judicial law clerk in the coming years. The authors would like to thank Samuel Brunson, Alex Zhang, Adam Chodorow, Hillel Levin, Allen Calhoun, Nicole Garnett, and all of the participants in the joint faculty workshop between the law schools of Emory University and the University of Georgia, for their insightful comments on drafts. We also thank Ashley Audisho and Cory Conley from Emory University School of Law for their excellent research assistance and Ari Liberman for his editorial suggestions. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 78 I. INTRODUCTION .............................................................................................. 78 II. THE LAW THEN AND NOW ............................................................................. 81 A. Under the U.S. Constitution ...................................................................... 81 1. 1947–1983: Public Aid Restricted to Secular Purposes ....................... 82 2. 1983–2011: Neutral Availability and Private Choice .......................... 84 3. 2011: Standing ...................................................................................... 86 4. 2017: Free Exercise .............................................................................. 87 B. Meanwhile, in the States ........................................................................... 90 1. No-Aid Provisions and State Interpretations ........................................ 91 2. No-Aid Provisions after 2017 ............................................................... 93 III. WHY TAX CREDITS ARE THE WRONG POLICY CHOICE .................................. 94 A. Tax Credits Have the Effect of Direct Payments while Circumventing Rules Governing Direct Payments .................................................................... 95 B. Bad Policy ................................................................................................. 97 1. Why Do States Subsidize School Choice? ............................................. 97 a. Equity ................................................................................................ 98 b. Pure Economics ................................................................................ 98 c. Freedom of Choice ............................................................................ 99 2. Why Do Tax Credits Fail to Achieve the Stated Goals of School Choice? ............................................................................................................. 100 a. Inequitable Distribution .................................................................. 100 b. Fraud Concerns ............................................................................... 109 3. These Concerns Show that Tax Credits are a Poor Instrument, Regardless of a State’s Reason for Implementing School Choice. ............. 112 a. Equity .............................................................................................. 112 b. Pure Economic Growth and Freedom of Choice ............................ 112 c. Inefficiency ..................................................................................... 115 IV. SOLUTIONS .................................................................................................. 115 V. ANTICIPATED OBJECTIONS AND OUR RESPONSES........................................ 119 A. Legal Arguments ..................................................................................... 119 1. Establishment Clause .......................................................................... 119 2. No-aid Clauses .................................................................................... 122 3. Other State Constitutional Issues........................................................ 124 4. Reforming, Instead of Replacing, Tax Credits .................................... 125 B. Policy Arguments.................................................................................... 127 VI. CONCLUSION ................................................................................................ 128 I. INTRODUCTION Across the nation, state-sponsored school choice programs are becoming increasingly prevalent. Such programs make state money available to families for educational costs, often including private school tuition.1 States sponsor these 1 States offering tuition tax credits include Georgia (Qualified Education Expense Tax Credit, Ga. Code Ann. § 48-7-29.16), Arizona (Ariz. Rev. Stat. Ann. § 43-1089), Florida (Florida Tax Credit Scholarship Program, Fla. Stat. Ann. § 212.099), Pennsylvania (24 Pa. Stat. Ann. § 20-2005-B), Montana (Mont. Code Ann. § 15-30-3111), and Missouri (Missouri Empowerment Scholarship 2023] THE WORST CHOICE FOR SCHOOL CHOICE 79 programs for various reasons, including to foster competition among private schools and to increase opportunities for students at struggling public schools. They likewise use various methods to administer them, including distributing vouchers, specially designated savings accounts, or tax credits to incentivize donations to organizations that provide scholarships. The method a state chooses to administer school choice matters, and this Article explains why tax credits are the wrong choice. At the outset, it is important to note that the following piece is not meant to argue for or against school choice. The merits of such programs as a general matter are contested, yes, but we do not enter the debate. Instead, we argue strictly that once a state has chosen to administer school choice, tax credits are the worst option for its execution, and direct funding is far better. It is our perspective that if a state operates school choice, its goal should be to administer the program in a way that ensures that everyone has an equal opportunity to benefit from it. Moreover, our policy objectives are for states opting into these programs to ensure a truly progressive system of school choice. States could (and should) go further, in line with the goals of a progressive tax code, and provide more aid to needier families, and avoid altogether regressive systems–for example, concentrating aid on the wealthy–which would not only be unpopular but inequitable. In any event, proper administration is key to ensuring these progressive objectives. Tax credits are ineffective at achieving any goal of school choice. They disproportionately benefit wealthy students at private schools who would have had school choice without tax credits, cost governments more than direct funding would, and are difficult to monitor. Given these pronounced flaws, it might seem confounding as to why they are such a prevalent mechanism. But, a closer historical consideration reveals exactly why this is the case. Simply, states flocked to tax credit options because they thought they had to. Tax credits became a popular method of school choice because they were believed to be the only constitutional option for funding religious schools. For most of the twentieth century, the Establishment Clause was understood to prohibit direct funding, and, furthermore, several states also enacted “no-aid” provisions in their state constitutions, which explicitly prohibited the government from financially supporting religious schools. Tax credits were a workaround because of their indirect structure. We will consider the structure of this program at some length later, but, in short, these tax credit programs involve an exchange of money, first between individual donors and private “scholarship” organizations (hereinafter “scholarship organizations”) and then between these scholarship organizations and private schools, which award funds to families as scholarships. The tax credit part comes in the first exchange. The donor might choose the (religious or secular) private school (or even the student) to which the scholarship organization directs their money, and then, at tax time, the donors receive a tax credit based on the amount of their donation to the scholarship organization. In states where the tax credit Accounts Program, Mo. Rev. Stat. § 135.712). These programs continue to grow; see e.g., Mike DeForest, Record Number of Students to Attend Private Schools Using State Vouchers, CLICKORLANDO.COM (Aug. 11, 2023), https://www.clickorlando.com/news/local/2023/08/11/recor d-number-of-students-to-attend-private-schools-using-state-vouchers/ [perma.cc/UWA2-HMJ6]. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 80 matches the value of the donation, the donor breaks even, effectively having directed some of their tax payment to a private school, while technically, no “state” funds are directly given to private schools. In other words, instead of the state directly paying a private school, an individual pays a private school, and the state reduces the individual’s tax bill by the amount of his donation. One can plainly see how this process is intentionally cumbersome and difficult to regulate—though it was seen as the only way to provide aid to religious schools in accordance with prior constitutional understandings.2 Recently this understanding has changed. Carson v. Makin3 clarified that there is no constitutional issue with governments’ subsidizing religious schools. Moreover, the Court held that the Free Exercise Clause forbids states that provide funding to private schools from disqualifying religious schools from that support only because they are religious; indeed, sometimes a state must support religious institutions. This change negates one of the main rationales for why tuition tax credits became so popular; their indirect structure is no longer necessary to comply with the Establishment Clause. Yet, still, many tuition tax credit systems, and the decisions upholding them as constitutional under state no-aid provisions, remain in force. Tax credits are no longer necessary for constitutionality, so the question becomes whether they are worth keeping around. We argue that they are not. States should do away with tuition tax credits and opt for direct funding4 in the form of vouchers.5 On policy grounds, tax credits are a poor mechanism for school choice. By design, they separate the government from aid going to religious schools. As collateral results, they are ineffective and prone to fraud. Starting from scratch with the law as it is today, we maintain that no state wishing to achieve educational improvement or equity would use tax credits. Instead, as argued below, a decision to fund private schools through tax credits only signals that the state prioritizes funding wealthy private schools and neglects private schools serving poor families. 2 See infra, text accompanying notes 110–170. 3 See Carson v. Makin, 142 S. Ct. 1987, 2002 (2022). 4 We recognize that both vouchers and tuition tax credits can be characterized as indirect; in both cases, funding is directed by parents and does not transfer immediately from the state to the school. We describe vouchers as direct because the government directly funds the private school in question, regardless of whether it is religious. With tax credits, there is no transaction between the government and the parents, similar to how the government provides tax deductions for charitable contributions. In that sense, tax credits are indirect. 5 We use the term “vouchers” to include all programs in which the government provides direct aid to the parents which can be spent on the education of their children, including programs like Arizona’s Empowerment Scholarship Account, a version of the more common “education savings account.” In each of these voucher-type programs, parents are given public funds to use for tuition or other educational expenses. See generally Nicole Stelle Garnett, Unlocking the Potential of Private-School Choice: Avoiding and Overcoming Obstacles to Successful Implementation, MANHATTAN INST. REP. (Mar. 16, 2023), https://manhattan.institute/article/unlocking-the-potential- of-private-school-choice-avoiding-and-overcoming-obstacles-to-successful-implementation [perm a.cc/2BMY-UXPT]. We do not see these various diverse funding programs as substantially different from each other in contrast to tuition tax credits, as they work independent of the tax code. The virtues and vices of each of these types of voucher programs is discussed in the above report. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 81 Direct funding, on the other hand, can be better regulated in ways commensurate with other public benefits. Reducing the misuse and fraud will enable states to achieve equitable school choice goals much more effectively. A choice not to use direct funding is a choice to stick with the inequity and inefficiency of tax credits—a choice to continue publicly subsidizing private education for the rich with little government oversight. Of course, proponents of tax credits would avoid articulating that their policy goal is to subsidize exclusively the wealthy’s private education and to remain programmatically unregulated, but those are—undoubtedly—the only reasons to continue using tax credits. States now clearly have the option to choose direct funding. States offering school choice, then, should use that mechanism, rejecting tax credits as an unnecessary and inefficient vestige of previous law, or else abstain from funding school choice at all. This Article’s four major parts support this conclusion. First, we describe the history of the federal and state laws that explain why tax credits have come to dominate as a tool for school choice but bolster the position that more direct forms are equally constitutional. The second section then discusses why tax credits are a poor mechanism for instituting school choice. The third section then describes why direct funding in the form of vouchers is a much more effective and desirable policy choice for states with school choice programs. The final sections respond to some possible criticisms of this position, followed by a summary and conclusion. II. THE LAW THEN AND NOW To understand why tax credits are so prevalent today—and to see why they were once appealing—it is necessary to understand how the First Amendment and state law have historically been applied. While states may now fund religious schools, and in some circumstances they must, the law took a long and winding road to get here. A. Under the U.S. Constitution Since the mid-twentieth century, the Supreme Court has faced a parade of cases challenging public aid in religious education. In the Court’s own words, it “can only dimly perceive the lines of demarcation in this extraordinarily sensitive area of constitutional law.”6 Accordingly, the doctrine has changed substantially over time. The First Amendment’s Religion Clauses state that legislators “shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof . . . .”7 The cases described below illustrate how federal rules governing public aid to private schools have changed. Cases initially required that aid be restricted to secular purposes, then became less restrictive with the doctrine of private choice, and, finally, were mostly disposed of by standing Free Exercise 6 Mueller v. Allen, 463 U.S. 388, 393 (1983) (quoting Lemon v. Kurtzman, 403 U.S. 602, 612 (1971)). 7 U.S. CONST. amend. I. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 82 challenges. Now, of course, states may directly fund private schools.8 If they do, however, they may not exclude religious schools on the basis that they are religious.9 But these are only recent developments, and the earlier cases show much more restrictive interpretations of the Religion Clauses, and, more poignantly, the higher importance of indirectness in school choice funding schemes which tax credits provided. 1. 1947–1983: Public Aid Restricted to Secular Purposes Beginning in the mid-twentieth century, public aid to private schools was permissible only if it was restricted to secular purposes. The Supreme Court first applied the Religion Clauses to the states in 1947.10 It upheld a New Jersey program that reimbursed parents, including parents of children at parochial schools, for the costs of transporting their children to school.11 The First Amendment allowed this program because it was available to all families, regardless of religion, and the services were “separate and so indisputably marked off from the religious function” of parochial schools.12 State aid to private religious schools remained permissible for the next forty years, as long as it was restricted to secular purposes. The Court upheld state statutes under which states loaned secular textbooks to students in public and nonpublic schools for free13 and funded facilities at religious schools that would never be used for religious purposes.14 It was similarly permissible for states to “include church-related schools in programs providing bus transportation, school lunches, and public health facilities . . . .”15 It was not a constitutional problem that state support subsidized some costs of attending religious schools; lightening the 8 See Trinity Lutheran Church of Columbia, Inc. v. Comer, 582 U.S. 449, 467 (2017) (“[T]he exclusion of Trinity Lutheran from a public benefit for which it is otherwise qualified, solely because it is a church, is odious to our Constitution all the same, and cannot stand.”); Espinoza v. Montana Dep’t of Revenue, 140 S. Ct. 2246, 2261 (2020) (“A State need not subsidize private education. But once a State decides to do so, it cannot disqualify some private schools solely because they are religious.”); Carson v. Makin, 142 S. Ct. 1987, 2002 (2022) (holding that Maine’s tuition assistance payments program could not exclude otherwise eligible schools on the basis of their religious exercise). 9 Carson, 142 S. Ct. at 1997. 10 Everson v. Bd. of Educ., 330 U.S. 1, 8 (1947); Sch. Dist. v. Ball, 473 U.S. 373, 381 (1985), overruled by Agostini v. Felton, 521 U.S. 203, 235 (1997) (“Everson made clear that the guarantees of the Establishment Clause apply to the State. . . .”). Even before the Religion Clauses were incorporated against the states, Louisiana taxpayers challenged a state program that loaned secular textbooks to all schoolchildren, including those in religious schools. Cochran v. La. State Bd. of Educ., 281 U.S. 370, 374 (1930); They argued that the program violated the Fourteenth Amendment by taking their private property (their taxes paid) for a private purpose (aiding nonpublic schools). Id. The Supreme Court upheld the program because it served the public purpose of education. Id. at 375. The program did not single out religious schools or their students; furthermore, the program benefited the students, not the schools. Id. at 374-75. 11 Everson, 330 U.S. at 17. 12 Id. at 18. 13 Bd. of Educ. v. Allen, 392 U.S. 236, 238 (1968); Meek v. Pittenger, 421 U.S. 349, 362 (1975), overruled by Mitchell v. Helms, 530 U.S. 793, 808 (2000); Wolman v. Walter, 433 U.S. 229, 232 (1977), overruled by Mitchell, 530 U.S. at 801. 14 Tilton v. Richardson, 403 U.S. 672, 689 (1971). 15 Meek, 421 U.S. at 364, overruled by Mitchell, 530 U.S. at 808. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 83 financial burden did “not alone demonstrate an unconstitutional degree of support for a religious institution.”16 While clearly secular resources like transportation were permissible during this time, the Court invalidated other provisions not as easily separable from religious instruction. These included “auxiliary services” like counseling and specialized instruction to be provided on the grounds of nonpublic schools.17 These services were impermissible because the monitoring required to ensure they remained secular would necessarily require excessive entanglement between church and state.18 The Court decided along these lines in 1973 in Committee for Public Education and Religious Liberty v. Nyquist.19 New York taxpayers challenged state law that provided direct grants from the state to nonpublic schools for facilities maintenance, reimbursed certain parents of students at nonpublic schools for some tuition expenses, and allowed other parents tax benefits corresponding to the amount they paid in nonpublic school tuition. Each of these provisions was invalidated. The New York law had the impermissible primary effect of advancing religion because the funds were not restricted to secular uses:20 “In 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.”21 Regarding the tuition reimbursements and tax benefits, the Court rejected the argument that aid was permissible because it went to parents or students, instead of to the school. Earlier cases considered this point; for example, the Court noted that when states loan textbooks directly to students instead of to schools, “the financial benefit is to parents and children, not to schools.”22 In Nyquist, the Court suggested that such statements were mostly dicta: “far from providing a per se immunity from examination of the substance of the State’s program, the fact that aid is disbursed to parents rather than to the schools is only one among many factors to be considered.”23 The Court did not decide the disputed issue of whether the tax benefit in Nyquist was a tax credit or a deduction.24 It had qualities of both, and its constitutionality did not depend on its label.25 It was a “form of encouragement and reward for sending [one’s] children to nonpublic schools,” just like the tuition 16 See Meek, 421 U.S. at 360, overruled by Mitchell, 530 U.S. at 808 (quoting Allen, 392 U.S. at 244). 17 Meek, 421 U.S. 349, overruled by Mitchell, 530 U.S. at 808; Sch. Dist. v. Ball, 473 U.S. 373, 397–98 (1985), overruled by Agostini v. Felton, 521 U.S. 203, 235 (1997). 18 Meek, 421 U.S. at 370, overruled by Mitchell, 530 U.S. at 808. 19 Comm. for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 774 (1973). 20 Id. at 756, 774, 780, 793. 21 Id. at 780. 22 Bd. of Educ. v. Allen, 392 U.S. 236, 244 (1968). 23 Nyquist, 413 U.S. at 781. 24 Id. at 789–90. 25 Id. (“As Mr. Chief Justice Burger’s opinion for the Court in Lemon v. Kurtzman, 403 U.S., at 614, notes, constitutional analysis is not a ‘legalistic minuet in which precise rules and forms must govern.’ Instead, we must ‘examine the form of the relationship for the light that it casts on the substance.’”) (citing Lemon v. Kurtzman, 403 U.S. 602, 614 (1971). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 84 reimbursement.26 “The only difference[,] that one parent receives an actual cash payment while the other is allowed to reduce by an arbitrary amount the sum he would otherwise be obliged to pay over to the State,” was of no moment.27 Whether the aid took the form of a reimbursement or tax credit, “neither [was] sufficiently restricted to assure that it [would] not have the impermissible effect of advancing the sectarian activities of religious schools.”28 These cases illustrate the rule that governed public aid in religious education cases beginning in 1947 and ending in 1983: the Establishment Clause permitted public aid to private schools as long as it was restricted to secular purposes. 2. 1983–2011: Neutral Availability and Private Choice The rule governing public aid to private schools evolved in 1983. Such aid, even if not restricted to secular purposes, was permitted as long as it was neutrally available and available to private schools only as a result of non-governmental choice, regardless of whether it served nonsecular purposes. Mueller v. Allen29 ushered in this expansion. There, the Supreme Court, divided five to four, upheld a Minnesota statute allowing parents to deduct expenditures on their children’s tuition, textbooks, and transportation from their taxable income, regardless of where the children attended school.30 The Court found that the law did not have the primary effect of advancing religion for several reasons: 1) the deduction was a permissible exercise of the Minnesota legislature’s discretion in distributing its tax burden,31 2) it was available to all families with schoolchildren,32 and 3) private schools received funds “only as a result of numerous, private choices of individual parents.”33 These reasons shaped subsequent cases and so merit further examination. The first reason that the Establishment Clause allowed Minnesota’s tax deduction was that it was a valid exercise of the state legislature’s power.34 This reason obviously would not justify, the Court wrote, a direct grant or reimbursement to parents like the one invalidated in Nyquist.35 Even Nyquist’s tax benefits did not resemble “genuine tax deduction[s]” because the amount that families saved did not correspond to the amount spent on tuition.36 Mueller’s educational expenses deduction, on the other hand, was one of many available under Minnesota law and perceived as an evenhanded attempt to equitably 26 Id. at 791. 27 Id. 28 Id. at 794. 29 Mueller v. Allen, 463 U.S. 388 (1983). 30 Id. 31 Id. at 396. 32 Id. at 397. 33 Id. at 399. 34 Id. at 396 (“[T]he Minnesota legislature's judgment that a deduction for educational expenses fairly equalizes the tax burden of its citizens and encourages desirable expenditures for educational purposes is entitled to substantial deference.”). 35 Mueller, 463 U.S. at 396 n.6 (“Plainly, [Nyquist’s] outright grants [in Comm. for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756 (1973)] to low-income parents did not take the form of ordinary tax benefits.”). 36 Id. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 85 distribute the tax burden.37 The Court thus approached the Mueller deduction with more deference than it did the tax benefits in Nyquist. The four dissenting Justices found no difference between the tax benefits in those two cases. In their view, both were impermissibly unrestricted to secular purposes and violated the Establishment Clause, regardless of their indirectness.38 They relied on Nyquist to assert that the aid’s form did not decide its constitutionality: [F]or purposes of determining whether such aid has the effect of advancing religion,’ it makes no difference whether the qualifying ‘parent receives an actual cash payment [or] is allowed to reduce . . . the sum he would otherwise be obliged to pay over to the State.’ It is equally irrelevant whether a reduction in taxes takes the form of a tax ‘credit,’ a tax ‘modification,’ or a tax ‘deduction.’39 Nonetheless, the majority’s view that the Mueller tax deduction was genuine differentiated it from the benefits struck down in Nyquist. The second reason listed by the Court—that the Minnesota aid was neutrally available to all parents—was also a critical distinction between Mueller and Nyquist. In Nyquist, aid was available exclusively to families with children in nonpublic schools.40 Mueller’s aid was available regardless of where children attended school and regardless of religion. The Court held that “a program, like [Minnesota’s], that neutrally provides state assistance to a broad spectrum of citizens is not readily subject to challenge under the Establishment Clause.” The law’s challengers argued the tax deduction was not neutrally available in practice. As the bulk of the deductible expenses were tuition payments, and given that public school students do not pay tuition, most of the “neutrally available” deductions were not actually available to their families. The Supreme Court found, however, that statistical reports of who happened to claim tax benefits had no place in constitutional interpretation.41 The Mueller Court’s third reason for upholding the Minnesota program was that private schools received aid only as a result of private choice. This reason is closely related to the notion of indirectness that Nyquist held was “only one among many factors to be considered.”42 Mueller reinvigorated the importance of indirectness, pointing out that where the Court invalidated public aid to private schools, it was usually in cases where funds were paid directly from the state to the 37 Id. at 396. 38 Id. at 406–07 (Marshall, J., dissenting). 39 Id. at 407–08 (Marshall, J., dissenting) (citations omitted) (quoting Nyquist, 413 U.S. at 790–91). 40 Id. at 398. (both quotes). 41 Mueller, 463 U.S. at 401 (“We would be loath to adopt a rule grounding the constitutionality of a facially neutral law on annual reports reciting the extent to which various classes of private citizens claimed benefits under the law. . . . [T]hat private persons fail in a particular year to claim the tax relief to which they are entitled—under a facially neutral statute—should be of little importance in determining the constitutionality of the statute permitting such relief.”). 42 Nyquist, 413 U.S.at 781 (describing “the fact that aid is disbursed to parents rather than to the schools”). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 86 school without passing through the students’ or parents’ hands.43 An appealing function of tax deductions was that “aid to parochial schools is available only as a result of decisions of individual parents[, so] no ‘imprimatur of State approval,’ can be deemed to have been conferred on any particular religion, or on religion generally.”44 Mueller marked a major change in cases assessing aid to public schools. In subsequent years, the Court upheld government provisions of a language interpreter at a Catholic school,45 tuition assistance for a religious school’s vocational training,46 and even made clear that government vouchers for tuition expenses are legal as well.47 States were also permitted to provide auxiliary teachers and equipment to private schools when such programs were neutrally available and dispensed to all eligible students, no matter where they attended school.48 This phase of Establishment Clause decisions also helped clarify historical cases in two ways. First, the Court replaced the amorphous notion of aid’s “directness” with the private choice requirement. The private choice requirement was met if private citizens had the opportunity to direct the aid elsewhere;49 “private choice is easier to see when aid literally passes through the hands of individuals,” but “literal” passage was not what the Establishment Clause required.50 Accordingly, even aid that was “allocated . . . based on the per capita number of students at each school” reflected private choice.51 Second, the public aid in Nyquist remained distinguishable from the expanding field of permissible aid under Mueller. Because Nyquist’s program was available only to private school students, it would have failed to meet the neutrality requirement.52 3. 2011: Standing In the 2011 Arizona Christian Sch. Tuition Org. v. Winn five-to-four decision, the Supreme Court cast many of these earlier Establishment Clause rulings into doubt.53 There, Arizona taxpayers challenged a state statute allowing tax credits for donors to funds that provided scholarships to private school students, including those in religious schools. The Winn Court did not assess neutrality and private choice because it decided the taxpayers lacked standing. It acknowledged that while taxpayers generally have standing to challenge government spending 43 Mueller, 463 U.S. at 399 (“[A]ll but one of our recent cases invalidating state aid to parochial schools have involved the direct transmission of assistance from the state to the schools themselves.”). 44 Id. (citation omitted). 45 Zobrest v. Catalina Foothills Sch. Dist., 509 U.S. 1, 13–14(1993). 46 Witters v. Washington Dep't of Servs. for the Blind, 474 U.S. 481, 489 (1986). 47 Zelman v. Simmons-Harris, 536 U.S. 639, 662–63 (2002). 48 See generally Agostini v. Felton, 521 U.S. 203 (1997) (overruling in part Sch. Dist. of City of Grand Rapids v. Ball, 473 U.S. 373, 395 (1985) (invalidating auxiliary services “unmediated by the tax code and the ‘numerous, private choices of individual parents of school-age children’”). 49 See Mitchell v. Helms, 530 U.S. 793, 816 (2000). 50 Id. at 816. 51 Id. at 829–30. 52 Zelman v. Simmons-Harris, 536 U.S. 639, 662 (2002) (“[W]e now hold that Nyquist does not govern neutral educational assistance programs that, like the program here, offer aid directly to a broad class of individual recipients defined without regard to religion.”). 53 See generally Arizona Christian Sch. Tuition Org. v. Winn, 563 U.S. 125 (2011). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 87 under the Establishment Clause, Arizona’s tax credits did not constitute government expenditures. Contributions to scholarship funds were never “drawn from general tax revenues,” so no money had “been extracted from [the challengers] and handed to a religious institution in violation of [their] conscience.”54 The Court acknowledged that it had decided earlier cases that may have had the same standing deficiency as Winn but suggested that those cases might have had another source of standing.55 A four-Justice dissent emphasized that no court before had distinguished between tax benefits and appropriations when applying the rule that taxpayers had standing for Establishment Clause challenges.56 That distinction had always been irrelevant because “cash grants and targeted tax breaks are means of accomplishing the same government objective.”57 Given that “targeted tax breaks are often ‘economically and functionally indistinguishable from a direct monetary subsidy,’” allowing standing to turn on such a formalistic difference nullifies the rule allowing for taxpayer standing in Establishment Clause cases.58 After Winn, neutrality and private choice presumptively remained the standard, but lack of taxpayer standing foreclosed further Establishment Clause challenges of public aid for private schools. Importantly, one consequence of Winn was a further increase in states’ use of tuition tax credits, now that it was clear that taxpayers lacked standing to challenge them.59 4. 2017: Free Exercise The standing obstacle impeded Establishment Clause challenges, but it did not end all litigation about public aid to private schools. Government aid was instead challenged on Free Exercise grounds in states that funded only secular private schools. Such was the case in Trinity Lutheran. There, a Missouri program provided grants for public and private schools to cover the costs of recycled rubber playground surfaces.60 The state’s policy was to disqualify religious entities from eligibility to receive a grant—a policy entirely based on the Missouri Constitution’s no-aid provision.61 Thus, when Trinity Lutheran Church and Preschool applied for a grant and was disqualified because of religious affiliation, a Free Exercise challenge developed. The Supreme Court held that the policy of disqualifying religious applicants violated Free Exercise. The policy conditioned a benefit on 54 Id. at 144. 55 See id. at 145. 56 Id. at 151 (Kagan, J., dissenting) (“In the decades since Flast, no court—not one—has differentiated between appropriations and tax expenditures in deciding whether litigants have standing.”). 57 Id. at 148 (Kagan, J., dissenting). 58 Id. at 157 (Kagan, J., dissenting) (quoting Rosenberger v. Rector & Visitors of Univ. of Virginia, 515 U.S. 819, 859 (1995)). 59 Thanks to Professor Adam Chodorow for illuminating this point. 60 See Trinity Lutheran Church of Columbia, Inc. v. Comer, 582 U.S. 449, 453 (2017). 61 Mont. Const. art. I, § 7 (requiring “[t]hat no money shall ever be taken from the public treasury, directly or indirectly, in aid of any church, sect or denomination of religion, or in aid of any priest, preacher, minister or teacher thereof, as such; and that no preference shall be given to nor any discrimination made against any church, sect or creed of religion, or any form of religious faith or worship.”). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 88 abstention from religious belief, impermissibly deterring the exercise of First Amendment rights. Yet, the grant program did not violate the Establishment Clause. It was faith-neutral, and playground resurfacing money was a secular resource.62 Accordingly, the Court distinguished Trinity Lutheran from a state’s denial of a public scholarship to a student seeking religious training that had been upheld under the Establishment Clause. Using taxpayer money to fund the training of clergy implicated antiestablishment interests in a way that using it to resurface preschool playgrounds did not.63 Justice Breyer agreed, writing separately that Missouri’s grants were “a general program designed to secure or to improve the health and safety of children,” no different from “such ‘general government services as ordinary police and fire protection.’”64 Under his view, the Establishment Clause allows grants for playground surfaces for the same reason it allows providing fire safety equipment to religious schools.65 The two-Justice dissent disagreed, stating that this was the exact kind of direct government funding to a religious institution that the Establishment Clause prohibits.66 That violation could not be rectified because the facilities benefitting from the grant could not be restricted to secular purposes,67 and funding did not depend at all on private choice.68 In 2020, the Supreme Court applied the same reasoning to a Free Exercise challenge of a Montana tax credit program.69 Montana granted tax credits of up to $150 for contributions to organizations that funded private school scholarships. The Montana Supreme Court held that those scholarships could not be used at religious schools because of the Montana Constitution’s no-aid provision.70 Its order 62 Trinity Lutheran argued that a state-funded grant for playground resurfacing complied with the Establishment Clause partly because playgrounds were used only for secular purposes. Brief for Petitioner at 30, Trinity Lutheran Church of Columbia, Inc. v. Pauley, 563 U.S. 125 (2017) (No. 1 5-577), 2016 WL 1496879 (U.S. 2016) (“[R]ubber playground surfacing material . . . is devoid of any religious content and . . . cannot possibly be diverted to a religious use. . .”). 63 See Trinity Lutheran Church of Columbia, 563 U.S. at 465. 64 See id. at 471 (Breyer, J., concurring). 65 Justice Gorsuch predicted that this distinction between discrimination based on religious status as opposed to religious use would fade into immateriality. See id. at 469 (Gorsuch, J., concurring in part) (“Is it a religious group that built the playground? Or did a group build the playground so it might be used to advance a religious mission? The distinction blurs in much the same way the line between acts and omissions can blur when stared at too long.”). After Espinoza and Carson, it is clear that he was right. 66 See id. at 473 (Sotomayor, J., dissenting) (“This Court has repeatedly warned that funding of exactly this kind—payments from the government to a house of worship—would cross the line drawn by the Establishment Clause.”). 67 See id. at 476 (Sotomayor, J., dissenting) (“The playground surface cannot be confined to secular use any more than lumber used to frame the Church’s walls, glass stained and used to form its windows, or nails used to build its altar.”). 68 See id. at 474 n.2 (Sotomayor, J., dissenting) (“Because Missouri decides which Scrap Tire Program applicants receive state funding, this case does not implicate a line of decisions about indirect aid programs in which aid reaches religious institutions ‘only as a result of the genuine and independent choices of private individuals.’”) (quoting Zelman v. Simmons-Harris, 536 U.S. 639, 649 (2002)). 69 See generally Espinoza v. Montana Dep’t of Revenue, 140 S. Ct. 2246 (2020). 70 Mont. Const. art. X, § 6 (“(1) The legislature, counties, cities, towns, school districts, and public corporations shall not make any direct or indirect appropriation or payment from any public fund or 2023] THE WORST CHOICE FOR SCHOOL CHOICE 89 invalidated the entire tax credit program. Families hoping to use the scholarships at religious schools appealed to the U.S. Supreme Court, which decided that the tax credit program must continue. The Court based this decision on the Religion Clauses. First, it held that the tax credits did not violate the Establishment Clause because they were neutrally available and reflected scholarship recipients’ private choice to attend religious schools. Second, the Court held that Montana’s no-aid provision, interpreted as prohibiting this scholarship program by the Montana Supreme Court, violated Free Exercise. The provision “bar[red] religious schools from public benefits solely because of the religious character of the schools.”71 Under Trinity Lutheran, that distinction was impermissible discrimination. The Montana Supreme Court had discontinued the tax credit program entirely before the U.S. Supreme Court decided this case. According to the dissenting Justices in Espinoza, that meant that the discriminatory granting of scholarships had ended; “[t]here simply [were] no scholarship funds to be had.”72 The majority disagreed and reversed the decision to end the program, because it had been made “by a court, and not based on some innocuous principle of state law.”73 The majority reasoned that the Montana Supreme Court invalidated the program because it violated the state’s no-aid provision, which itself violated Free Exercise. If the Montana Supreme Court had realized its state provision violated federal law, it would have had no basis on which to end the program. State courts “must not give effect to state laws that conflict with federal law,”74 so “the Montana Supreme Court should have ‘disregard[ed]’ the no-aid provision and decided this case ‘conformably to the [C]onstitution’ of the United States.”75 In short, when a state court is faced with a program that violates the federal Constitution to comply with the state constitution, it must invalidate the state provision (here, the no-aid clause), not the program itself.76 This rule exists only in the context of judicial decisions; state legislatures remain free to end or alter school choice programs.77 More importantly, shortly after Espinoza, Carson v. Makin similarly held that Free Exercise required a state to financially support religious schools.78 A monies, or any grant of lands or other property for any sectarian purpose or to aid any church, school, academy, seminary, college, university, or other literary or scientific institution, controlled in whole or in part by any church, sect, or denomination. (2) This section shall not apply to funds from federal sources provided to the state for the express purpose of distribution to non-public education.”). 71 Espinoza, 140 S. Ct. at 2255. 72 Id. at 2279. 73 Id. at 2262 (emphasis added) (“The Montana Legislature created the scholarship program; the Legislature never chose to end it, for policy or other reasons.”). 74 Espinoza, 140 S. Ct. at 2262 (citing Armstrong v. Exceptional Child Ctr., Inc., 575 U.S. 320, 324 (2015)). 75 Id. (citing Marbury v. Madison, 5 U.S. 137, 178 (1803)). 76 No-aid clauses are not facially unconstitutional. States can constitutionally comply with no-aid clauses by never sponsoring school choice (for either religious or non-religious private schools). When a state is already sponsoring school choice for non-religious schools, however, it cannot constitutionally refuse to extend that program to religious schools because of its no-aid clause (that application of the no-aid clause would be unconstitutional, and under Espinoza, requires courts to invalidate the no-aid clause instead of ending the program). 77 Espinoza v. Montana Dep’t of Revenue, 140 S. Ct. 2246, 2262 (2020) (emphasis added) (“The Montana Legislature created the scholarship program; the Legislature never chose to end it, for policy or other reasons.”). 78 Carson v. Makin, 142 S. Ct. 1987, 1989 (2022). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 90 Maine program assisted students living in remote districts without secondary public schools. The program would pay tuition at a private secondary school for those students, as long as the school was nonsectarian. This distinction was not based on a state constitutional provision; it came instead from “an opinion by the Maine attorney general taking the position that public funding of private religious schools violated the Establishment Clause of the First Amendment.”79 The Court stated that such funding did not violate the Establishment Clause because it was neutrally available and reflected private choice, but excluding religious schools from this public benefit violated Free Exercise.80 It was a straightforward application of Espinoza’s holding that “[a] State need not subsidize private education. But once a State decides to do so, it cannot disqualify schools solely because they are religious.”81 Nor can a state supreme court address this problem by judicially ending the aid program altogether. State legislatures can choose to aid private schools or not, but when they do, the U.S. Supreme Court has now held that courts deciding on programs that exclude religious schools must order inclusion of religious schools. This last phase of Supreme Court case law dramatically shifted the meaning of the Religion Clauses.82 States used to have to prove that they met an exception to the general rule that they could not fund religious entities. Espinoza and Carson obviated the lingering possibility that subsidizing religious entities was permissible as long as it focused on arguably secular resources, like fire protection or playground surfaces. If that were still the rule, Espinoza and Carson could not have upheld aid for tuition costs, because they are inseparable from the entities’ sectarianism. Taxpayers have not had standing to challenge tax credits on Establishment Clause grounds since 2011. Now it is clear that even if they did, they would lose. Free Exercise requires states to fund religious entities in some circumstances—situations where the state is funding private secular institutions compel the state to fund private religious institutions as well.83 B. Meanwhile, in the States As federal law on aid to religious schools evolved, so did state law. The sections below outline the rise and fall of state no-aid provisions and explain why they are mostly unenforceable today. The evolution of no-aid provisions and their interpretations is important for understanding why tax credit systems exist in their current form today. 79 Id. at 1994. 80 Id. at 1998. 81 Espinoza, 140 S. Ct. at 2262 (2020) (quoted by Carson v. Makin, 142 S. Ct. at 1997). 82 Trinity Lutheran Church of Columbia, Inc. v. Pauley, 788 F.3d 779, 783 (8th Cir. 2015), rev’d and remanded sub nom. Trinity Lutheran Church of Columbia, Inc. v. Comer, 137 S. Ct. 2012, 2016 (2017) (describing a ruling that a no-aid provision violates the First Amendment, the relief that Trinity Lutheran later won, as “unprecedented”). 83 Steven Green, Symposium: RIP state “Blaine Amendments” — Espinoza and the “no-aid” principle, SCOTUSBLOG (Jun. 30, 2020), https://www.scotusblog.com/2020/06/symposium-rip-sta te-blaine-amendments-espinoza-and-the-no-aid-principle/ [perma.cc/H3SR-AK6R] (describing Espinoza as “inflat[ing] the free exercise component but ignor[ing] the freedom-enhancing aspect of non-establishment.”). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 91 1. No-Aid Provisions and State Interpretations Many state constitutions contain no-aid provisions explicitly stating that no public funds should be paid to religious institutions.84 Traditionally, they had been held as a permissible way for states “to enforce a more strict policy of church and state separation than that required by the First Amendment.”85 They are often referred to as “Blaine Amendments,” referring to an 1876 proposed amendment to the U.S. Constitution that would have explicitly prohibited state funding of religious schools.86 The term carries a negative connotation because the Blaine Amendment itself was based on widespread anti-Catholic sentiment.87 Some states’ no-aid provisions directly followed the failure of the Blaine Amendment and indeed reflect anti-Catholic views. Others, however, were adopted before the Blaine Amendment conflict and were not born of that bias.88 In any event, state no-aid provisions persist. Some state high courts, like Montana’s in Espinoza, have held that no-aid provisions prohibit tuition tax credits.89 Others have gone the route in Winn, holding that their state provisions do not apply to tax credits because tax credits are not actually expenditures.90 As state high courts are the ultimate decision makers on the meanings of their own constitutions, states are entitled to interpret their no-aid provisions differently. States interpreting along the lines of Winn, including Florida, Georgia, and Arizona, serve to exemplify why the mechanism of tuition tax credits became 84 Id. (“38 state constitutions contain provisions that prohibit public monies being spent in aid of religious institutions or religious education.”). 85 Luetkemeyer v. Kaufmann, 364 F. Supp. 376, 386 (W.D. Mo. 1973), aff’d. 419 U.S. 888 (1974) (stating such a practice “does not present any substantial federal constitutional question”). Compare Locke v. Davey, 540 U.S. 712 (2004) (upholding Washington prohibition on state funding of training of ministers based on Washington constitutional provision that “draws a more stringent line than that drawn by the United States Constitution”) with Carson v. Makin, 142 S. Ct. 1987, 1989 (2022) (invalidating “Maine's decision to continue excluding religious schools from its tuition assistance program after Zelman [which] promotes stricter separation of church and state than the Federal Constitution requires”). 86 See, e.g., Mitchell v. Helms, 530 U.S. 793, 828 (2000); Green, supra note 83. 87 Green, supra note 83. 88 Id. (“15 states adopted no-funding provisions prior to the Blaine Amendment, with several arising in states with little or no discernible religious conflict. They also arose at a time that states were establishing their public schools and were seeking to guarantee the financial security of those fledgling schools.”). 89 Espinoza v. Montana Dep’t of Revenue, 35 P.3d 603, rev’d and remanded, 140 S. Ct. 2246, 2249 (2020) (“When the Legislature indirectly pays general tuition payments at sectarian schools, the Legislature effectively subsidizes the sectarian school's educational program. That type of government subsidy in aid of sectarian schools is precisely what the Delegates intended [the Montana Constitution] to prohibit.”). 90 Kotterman v. Killian, 972 P.2d 606, 621 (1999) (“[T]his tax credit is not an appropriation of public money.”); Gaddy v. Georgia Dep't of Revenue, 802 S.E.2d 225, 230 (2017) (“We also reject the assertion that plaintiffs have standing because these tax credits actually amount to unconstitutional expenditures of tax revenues or public funds. The statutes that govern the Program demonstrate that only private funds, and not public revenue, are used.”); McCall v. Scott, 199 So. 3d 359, 370 (Fla. Dist. Ct. App. 2016) (“[T]he legislative actions challenged in this case, the authorization of tax credits under the FTCSP and the payment of private funds to private schools via scholarships authorized under the FTCSP, involve no appropriation from the public treasury. The program is funded through voluntary, private donations by individual and corporate taxpayers.”). about:blank COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 92 so prevalent:91 they circumvent no-aid provisions. No-aid provisions forbade government funding of religious schools. In these states, however, tax credits were not interpreted as government funding, so they allowed the government to support religious schools and comply with no-aid provisions. Consider Florida, for instance. Its no-aid provision states that “[n]o revenue of the state or any political subdivision or agency thereof shall ever be taken from the public treasury directly or indirectly in aid of any church, sect, or religious denomination or in aid of any sectarian institution.”92 The state provides tax credits to donors to organizations that provide scholarships, including to religious schools.93 When taxpayers challenged the tax credits for violating the no-aid provision back in 2016, Florida courts found that the plaintiffs lacked standing.94 Florida taxpayers have standing where they can show that government spending or taxing violates a specific constitutional limit on the government’s powers to tax and spend. Here, “[b]ecause there was no diversion of any state revenues from public schools to private schools through the operation of” the tax credits, the plaintiffs could not make that showing.95 Accordingly, the tax credit system and the no-aid provision were left to coexist. The court then distinguished another Florida case, where taxpayers challenged “statutes that authorized the state to direct appropriations to sectarian institutions.”96 Without the structure of the tax credit, those taxpayers had standing because they alleged that state statutes allowed appropriations to aid sectarian institutions.97 This example illustrates why states aid religious schools through tax credits rather than by direct means of funding—many states thought they were required to do so by their no-aid provisions. The tax credit system arose not because states thought it was the best way to support private schools, but because it was the only way they thought they could do so constitutionally. This led to the tax credit scheme used by states today: taxpayers make contributions to non-profit organizations,98 in some states with the option to indicate which school or child their contribution should be directed to.99 Those organizations direct most of that money to private 91 In 2008, when Georgia passed its Qualified Education Expense Credit, five other states already had similar programs. Delon Pinto & Wade Walker, Elementary and Secondary Education: Amend Titles 20 and 48 of the Official Code of Georgia Annotated Relating, Respectively to Education and Revenue Taxation, So As to Provide for A Program of Educational Improvement; 25 GA. ST. U. L. REV. 73, 82 (2008) (Arizona, Florida, Pennsylvania, Iowa, and Rhode Island). 92 Fla. Const. art. I, § 3. 93 McCall, 199 So. 3d at 363. 94 Id. at 359. 95 Id. at 366. 96 Id. at 359 (emphasis added) (describing Council for Secular Humanism, Inc. v. McNeil, 44 So. 3d 112 (Fla. Dist. Ct. App. 2010)). 97 See Council for Secular Humanism, 44 So. 3d 112, 122 (Fla. Dist. Ct. App. 2010). 98 In Georgia, a student scholarship organization must use “at least 90 percent of its annual revenue received from donations for scholarships or tuition grants to allow students to attend any qualified school of their parents’ choice,” and may not “[limit] availability to only students of one school.” Ga. Code Ann. § 20-2A-1 (West). 99 Georgia GOAL Scholarship Program, Inc., Designation Options, Georgia GOAL Scholarship Program (2023), https://www.goalscholarship.org/for_donors/page/designation-options [perma.cc/ 7JPJ-FR7M] (GA allows donors to designate school); King Foundation dba Arizona Tax Credit, 2023] THE WORST CHOICE FOR SCHOOL CHOICE 93 schools for scholarships, and the taxpayers can claim a dollar-for-dollar tax credit up to a certain amount.100 In some states, including Georgia, the only eligibility requirement for receiving a scholarship is that a student is transferring into a private school from a public one. In others, students’ household income must fall below a certain level, or the student must qualify for benefits like food assistance.101 2. No-Aid Provisions after 2017 In 2017, the Supreme Court in Trinity Lutheran held that Missouri could not, on the basis of its no-aid provision, exclude religious schools from otherwise available public benefits. Its decision technically only invalidated Missouri’s policy of automatically disqualifying religious schools from competition for state grants and did not directly affect Missouri’s no-aid provision.102 Speculation about what that decision meant for no-aid provisions103 was mostly resolved three years later in Espinoza. There, the Court held that Montana’s application of its no-aid provision violated Free Exercise.104 Finally, in Carson, the Court repudiated Maine’s understanding that the Establishment Clause prohibits states from funding religious schools. The Court held not only that no such prohibition exists, but also that Free Exercise requires including religious schools in otherwise-available aid. Taking the trilogy of cases together, it now seems that no-aid provisions comply with Free Exercise only in two situations. First, if the no-aid provision is interpreted like Arizona’s, meaning it allows for indirect aid like tax credits, then it is constitutional as long as the only type of aid the state provides to private schools F.A.Q., Arizona Tax Credit, https://aztxcr.org/faqs/ [perma.cc/GS8X-KTMS] (Arizona allows donors to designate schools or recommend individual students). Florida does not allow designation of a contribution to a school or student. Scholarship-funding organizations “may not restrict or reserve scholarships for use at a particular private school or provide scholarships to a child of an owner or operator.” Fla. Stat. Ann. § 1002.395 (West). They “[m]ust allow an eligible student to attend any eligible private school and must allow a parent to transfer a scholarship during a school year to any other eligible private school of the parent’s choice.” Id. In 2017, over half of the tax credit programs prohibited donors from designating their contribution to a specific individual, and about a third of programs prohibited donors from designating a specific school. U.S. Gov’t Accountability Off., GAO-19-664, Private School Choice: Accountability in State Tax Credit Scholarship Programs (2019), https://www.gao.gov/assets/gao-19-664.pdf [perma.cc/7JLE-7DE M]. 100 See, e.g., Ariz. Dep’t of Revenue, Credits for Contributions to Certified School Tuition Organizations, Ariz. Dep’t of Revenue, https://azdor.gov/tax-credits/credits-contributions-certified- school-tuition-organizations [perma.cc/8AKK-B5KT]; Fla. Dep’t of Educ, FTC for Parents, Fla. Dep’t of Educ (2023), https://www.fldoe.org/schools/school-choice/k-12-scholarship-programs/ftc /ftc-parent-info.stml [perma.cc/GHW6-T9ER]. 101 In Florida, to qualify for a scholarship, students must be on a list of students already receiving public benefits, have a household income level below 375% of the federal poverty level, or be in foster care. Fla. Stat. Ann. § 1002.395 (West). 102 Trinity Lutheran Church of Columbia, Inc. v. Comer, 582 U.S. 449, 467 (2017). 103 James Hirsen, Symposium: A takedown of the Blaine Amendments, SCOTUSBLOG (Jul. 2, 2020), https://www.scotusblog.com/2020/07/symposium-a-takedown-of-the-blaine-amendments/ [perma. cc/3MWM-F8ZK] (describing Trinity Lutheran as “a funeral dirge . . . for Blaine funding restrictions”). 104 Espinoza v. Mont. Dep't of Revenue, 140 S. Ct. 2246, 2262 (2020). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 94 is indirect.105 That way, religious schools remain eligible for all the aid available to private schools. The second way that a no-aid provision could now comply with Free Exercise is if the state provides no aid to private schools whatsoever.106 The takeaway for the purposes of tax credits is this: many states chose tax credits as the means to fund private schools because they thought the Establishment Clause or their state no-aid provisions required them to. The Supreme Court has now made clear that the Establishment Clause does not forbid direct payments.107 Moreover, no-aid provisions, if interpreted to have any force, violate Free Exercise.108 Therefore, the only thing “requiring” states to use tuition tax credits could be an impotent no-aid provision interpreted to prohibit direct funding but allow tax credits. Such a provision is useless because, as the next section describes, tax credits are an inefficient, ineffective way to achieve the goals of direct funding. III. WHY TAX CREDITS ARE THE WRONG POLICY CHOICE Because direct aid to parochial schools is permitted—even sometimes mandatory—tax credits are no longer necessary for compliance with the Establishment Clause. The constitutionality of the other reason for the tax credit structure, state no-aid provisions, is also now severely suspect. Nevertheless, tax credits remain ubiquitous. Now that the reasons for their development have been obviated, the time has come to ask whether tax credits are actually worthwhile. In other words, just because tax credits are no longer necessary, is it time to change the status quo? This section argues in the affirmative. Tax credits are not only unnecessary, but they are also bad law and bad policy. The decisions upholding them as bypasses around the Establishment Clause or no-aid state laws rest on debatable legal ground; after all, tax credits may 105 Oklahoma has seen some interesting developments with respect to its no-aid provision since Carson. The Oklahoma Constitution requires public schools to be “nonsectarian” and “free from sectarian control.” When considering a Catholic charter school, the state’s then-Attorney General John M. O’Connor wrote that the Supreme Court would probably permit a religious charter school under Carson. The subsequent state Attorney General, Gentner Drummond, withdrew that opinion and advised that a religious charter school would violate the Oklahoma Constitution. Mark Walsh, A Major Reversal on Religious Charter Schools in Oklahoma, EDUCATION WEEK (Feb. 24, 2023), https://www.edweek.org/policy-politics/a-major-reversal-on-religious-charter-schools-in-oklahom a/2023/02 [perma.cc/KV3N-GB9E]. 106 Espinoza, 140 S. Ct. at 2261; (quoted by Carson v. Makin, 142 S. Ct. 1987, 1997 (2022)) (“A State need not subsidize private education. But once a State decides to do so, it cannot disqualify some private schools solely because they are religious.”). We are inclined to think that in a universe in which vouchers are normative and paid in large enough quantities that all students use them, the distinction between “public,” “charter,” and “private” schools is merely formulaic, since all are funded by the government. 107 Trinity Lutheran implies that private choice is not necessary to circumvent the Establishment Clause, see 582 U.S. 449, and Carson required direct funding to religious schools, see 142 S. Ct. at 2014 (2022). 108 Steven K. Green, Requiem for State “Blaine Amendments”, 64 JOURNAL OF CHURCH AND STATE, 437, 440 (2022) (“Even though the Supreme Court stopped short of declaring Montana’s no-funding provision to be unconstitutional, the Espinoza decision possibly made it—and similar provisions in thirty-seven other states—dead-letter laws.”); Steven Green, Symposium: RIP state “Blaine Amendments” — Espinoza and the “no-aid” principle, SCOTUSBLOG (Jun. 30, 2020), https://ww w.scotusblog.com/2020/06/symposium-rip-state-blaine-amendments-espinoza-and-the-no-aid-prin ciple/ [perma.cc/3MCD-5JBE] (“[Espinoza] ha[s] now tainted all no-aid provisions, and the more general principle against government funding of religion, with the aura of discrimination.”). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 95 formally circumvent constitutional issues, but they have the same effect as direct payments. But far from being harmless vestigial features of past legal landscapes, tax credits fail to properly achieve their goals and are an inefficient waste of public resources. The weaknesses in their legal basis and their inefficiencies were not an accident; instead, those qualities were features thought necessary to the validity of tax credits. The weaknesses discussed in detail below all exist for the same reason, which is to distance the government—and its oversight—from the aid. The further removed the government was from the tax credit system, the more defensible tax credits were as compliant with the Establishment Clause. Inherent in that distance was also a lack of regulation—the government needed to avoid entangling itself with tax credits so that they could not be challenged as public aid. With nonexistent regulation, the government avoided connection with malfeasance in the tax credit system, but it also created a system that is not only ineffective but also rife with abuse, as noted in the following sections.109 A. Tax Credits Have the Effect of Direct Payments while Circumventing Rules Governing Direct Payments The history of Establishment Clause jurisprudence and no-aid provisions demonstrates that tax credits were used as a way around rules that would otherwise prohibit government aid to religious schools.110 They allowed the government to do indirectly what it could not do directly. The result is an absurdly contrived system of funding private schools with both educational and policy failings. The purpose of this section is to demonstrate that tax credits have the same effect as direct payments, but are a much more cumbersome mechanism for aiding private schools. Tax credits are a formalistic bypass amounting to a dollar-for-dollar reduction in someone’s tax bill in exchange for a donation. Georgia private schools, for example, promote the program on their websites under the heading “Redirect Your Tax Dollars.”111 Married taxpayers filing jointly under Georgia’s tax credit 109 Recall that tuition tax credits work by allowing individuals or corporations to receive a tax credit, which is the right not to pay that amount to the state in taxes, in exchange for giving money to an intermediary organization (a “scholarship organization”). Then, the scholarship organization makes a donation to a private school, often the private school of the donor’s choice, as a scholarship. In contrast to vouchers (where the government collects taxes and then allows individuals to direct a portion of that to a private school), a government with a tax credit program forgoes receiving dollars as taxes if they were paid to a scholarship organization. For more on this, see Carl Davis, State Tax Subsidies for Private K-12 Education, INSTITUTE ON TAXATION & ECONOMIC POLICY (Oct. 2016), https://itep.sfo2.digitaloceanspaces.com/k12taxsubsidies.pdf [perma.cc/55WZ-37RT]. 110 See, e.g., Brief of Americans United for Separation of Church and State, Gaddy v. Ga. Dep’t of Revenue, 802 S.E.2d 225 (2017) (No. S17A0177), 2017 WL 373535 at 3 (“Programs like the Tax Credit Program, which divert tax dollars from the public schools to private religious schools, are precisely what the No-Aid Clause, with its careful inclusion of a prohibition on indirect as well as direct aid, was meant to forbid.”); Kotterman v. Killian, 972 P.2d 606 (1999) (Feldman, J., dissenting) (“It is a dangerous doctrine that permits the state to divert money otherwise due the state treasury and apply it to uses forbidden by the state's constitution.”). 111 COVENANT CHRISTIAN SCHOOL, Georgia Private School Tax Credit, https://www.ccssmyrna.org /give/georgia-private-school-tax-credit.cfm [perma.cc/RP2E-7C26] (“Redirect Your Tax Dollars to [Covenant Christian School]”); Porter Academy, Redirect Your Tax Dollars, Porter Academy, https://www.porteracademy.org/support-porter-academy/redirect-your-tax-dollars/ [perma.cc/A69- L-AVS4]. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 96 program who donate $5,000 to be used for scholarships at private schools will have their state tax bill reduced by $5,000.112 As long as their state tax bill before the credit would be at least $5,000, they lose no money on the donation. In other words, instead of paying $5,000 to Georgia in taxes, they donated $5,000 to be used for private school scholarships. The taxpayers lose zero dollars, because their tax bill was reduced by the amount they donated, and the state loses out on $5,000 in tax revenue that it would have otherwise gained. Nevertheless, Georgia characterizes funds flowing through scholarship organizations as “only private funds, and not public revenue.” Because private entities control the funds, they are not subject to constitutional restrictions on direct payments. The Georgia Supreme Court characterized tuition tax credits as follows: Individuals and corporations choose the [scholarship organizations] to which they wish to direct contributions; these private [scholarship organizations] select the student recipients of the scholarships they award; and the students and their parents decide whether to use their scholarships at religious or other private schools. The State controls none of these decisions. Nor does it control the contributed funds or the educational entities that ultimately receive the funds.113 The money granted in tax credits resembles government spending— Georgia budgets the amount that can be claimed in credits every year114—but technically, it is not government spending. Nor is it regulated and enforced like government spending. Whether the government pays someone $5,000 or cuts his tax bill by $5,000, the effect is the same. The tax credit system, however, lacks the regulations customarily attached to government benefits.115 In view of the standing restrictions and avoidance of constitutional snags, we understand why tax credits became a deeply appealing method to support parochial schools. Though based on extremely formalistic legal reasoning, they succeeded as a work-around. And while insulating the government was a primary reason for tax credits’ success, tax credits also came with political advantages and 112 Ga. Code Ann. § 48-7-29.16 (West); Georgia GOAL Scholarship Program, Inc., Frequently Asked Questions, Georgia GOAL Scholarship Program, https://www.goalscholarship.org/about_g oal/page/frequently-asked-questions [perma.cc/WBY3-LVEF]. 113 Gaddy v. Ga. Dep’t of Revenue, 802 S.E.2d 225, 230 (2017). 114 In 2008, when Georgia’s Qualified Education Expenses credit was first passed, the cap was $50 million, and state legislators considered how the cap would affect the state’s budget. Pinto & Walker, supra note 91, at 74–75. For the tax year 2023, Georgia allocated $120 million to be used to reimburse taxpayers for their donations to private school tuition funds. Ga. Code Ann. § 48-7-29 .16 (West); Ga. Dep’t of Revenue, Tax Credit Summaries, Dep’t of Revenue, https://dor.georgia.g ov/tax-credit-summaries [perma.cc/76BB-PCC7]. That cap was met on January 3, 2023, the first day that the Georgia Department of Revenue accepted tax credit applications for that year. After the allocated $120 million had been applied for, the Department of Revenue ceased accepting applications for the credit. The fact that Georgia allocates $120 million from its budget to be used for tax credits demonstrates the similarity between tax credits and payments from the government— if tax credits did not have the same effect as withdrawing money from the state treasury, Georgia would not need to limit them like this. Unlike tax deductions, which reduce the taxpayer’s bill indirectly by reducing her taxable income, tax credits are treated like state revenue by being accounted for in the budget. 115 See Section II.B.2.ii, Fraud Concerns, infra p. 111. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 97 other appealing draws.116 Still, tax credits were, and remain, a convoluted, albeit legal, schema. Now that the Court has clarified that direct payments are permissible and sometimes mandatory, the extra complication inherent to tax credits is unnecessary, administratively expensive, and prevents aid from going where it is most needed. If someone with a goal of funding private school scholarships started from a blank slate today, it would be absurd to invent a complicated system like this. Thus, we contend, these systems survive only because of their established popularity. The extra complication is not just a neutral inconvenience. The next section explains why the convoluted system is more than unnecessary; it is both educationally and administratively counterproductive. As a result of the tax credit system, educational quality, and ultimately the state and its people, suffer. B. Bad Policy Tax credits were initially popular because states believed they needed to starkly separate themselves from public aid directed at private schools. For the same reason, tax credits are an unmonitored, minimally effective method of aiding poorer schools and schools serving the poor. Beliefs about the Establishment Clause and state no-aid clauses led states to enact tax credit systems and then look the other way. That failure to monitor was not accidental; state governments intentionally chose not to play a role, lest the re-directed tax payments be considered “state money” and entangle the government with religion. Because of the lack of regulation, tax credits are an unsurprisingly inefficient and ineffective way for the government to support private schools. There are plenty of stated reasons for supporting school choice, but, as detailed below, tax credits serve none of these goals well. Tax credits have ended up selectively benefitting—and being abused by—wealthy schools at the expense of poor schools. Direct aid better achieves the goal of promoting quality education, and now that it is clearly constitutional, states should replace tax credit systems with direct aid. The following subsections break down why tax credits are a poor mechanism for school choice. Section one catalogs some of the main reasons that states pursue school choice. Those reasons include plenty of worthwhile objectives, decreasing inequity and fostering competition among them. Notably absent from those purported goals, however, is subsidizing private school tuition for the rich. Section two explains why this unspoken result—selectively helping the rich pay for private school—is actually what tax credits achieve. It describes the two main flaws of tax credits: their inequitable distribution and evasion of regulation. Section three demonstrates that because of those flaws, tax credits are a counterproductive instrument for achieving any of the stated goals of school choice. 1. Why Do States Subsidize School Choice? To understand why tax credits are not only a poor, but also counterproductive, tool for school choice, it is first necessary to understand why states pursue school choice at all. There are plenty of reasons why state governments want to fund private schools, “from raising the quality of local schools 116 These political reasons, as discussed below, are still insufficient to justify using tax credits when direct funding—unencumbered by the tax code, available regardless of people’s tax filing status, and more readily regulated—is possible. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 98 and boosting religious freedom to empowering inner-city parents.”117 Broadly, they seek to improve the quality of education. The stated goal of Georgia’s private school tax credit, for example, is “educational improvement.”118 Some states articulate more specific goals. Those goals roughly break down into three groups, including educational equity, economic competition, and protecting individual freedom of choice. Below we will explain these different goals and how school choice theoretically achieves them. Keep in mind our reason for doing so: to demonstrate that tax credits do not actually achieve any of these goals. We by no means suggest that these goals are unworthy, or that school choice itself is good or bad. Our purpose here is only to provide necessary background to the larger point that tax credits are a poor tool for achieving any of the stated purposes of school choice. a. Equity With that, we will discuss the first stated goal of school choice: educational equity. Under this view, public aid enables private school access for students who could not otherwise afford it.119 Historically, this has been a leading justification for school choice.120 Florida, for example, seeks to “expand[] educational opportunities for children of families that have limited financial resources.”121 This can be seen as an extension of school choice’s origins as a response to desegregation in the South; it was advocated as empowering against paternalistic government control to allow Black families choices of schools.122 Some programs, for example, make funds available specifically to students at low-performing public schools or with family incomes below a certain level. Prioritizing those students for school choice is seen as a way to increase their opportunity to achieve regardless of the district they live in.123 b. Pure Economics Not all rationales are about equity, however. A purely economic theory championed by Milton Friedman has supported school choice since at least the mid- 117 Bruce Fuller, Richard F. Elmore, & Gary Orfield, Policy-Making in the Dark: Illuminating the School Choice Debate, in WHO CHOOSES? WHO LOSES? CULTURE, INSTITUTIONS, AND THE UNEQUAL EFFECTS OF SCHOOL CHOICE 2 (1996). 118 Pinto & Walker, supra note 91, at 73. 119 Hillel Y. Levin, Tax Credit Scholarship Programs and the Changing Ecology of Public Education, 45 ARIZ. ST. L.J. 1033, 1073 (2013) (describing one distinct group of school choice proponents as seeking redistribution of opportunity); Garnett, supra note 5, at 8. 120 Garnett, supra note 5, at 8 (“[A]dvocates have historically argued that parental choice is needed for children who are not well served by district public schools, including economically disadvantaged students and students with disabilities.”). 121 Fla. Stat. Ann. § 1002.395 (1)(b), 4–5 (West); see also Fla. Dep’t of Educ., Florida Tax Credit FAQs, Fla. Dep’t of Educ., https://www.fldoe.org/schools/school-choice/k-12-scholarship-program s/ftc/ftc-faqs.stml [perma.cc/8DTS-EUAD]. 122 Fuller, Elmore, & Orfield, supra note 117, at 3. (“Originally advocated in the South as a way to avoid the desegregation of public schools, school choice came to be seen by the Left as a way to empower poor and working-class families to challenge paternalistic bureaucracies.”). 123 Pinto & Walker, supra note 91, at 84; Hillel Y. Levin, Tax Credit Scholarship Programs and the Changing Ecology of Public Education, 45 ARIZ. ST. L. J. 1033, 1073 (2013). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 99 twentieth century, gaining popularity in the 1980s.124 Friedman proposed separating educational funding from administration; “[g]overnments could require a minimum level of education which they could finance by giving parents vouchers redeemable for a specified maximum sum per child per year if spent on ‘approved’ educational services.”125 This would foster competition among schools, “meet the just complaints of parents that if they send their children to private nonsubsidized schools they are required to pay twice for education,” and reduce the government’s overall role in administering education.126 Friedman acknowledged that an obstacle to school choice was “the general disrepute of cash grants to individuals (‘handouts’) with the absence of an efficient administrative machinery to handle the distribution of vouchers and check their use.”127 Writing in 1955, he asserted that such machinery was developed through “the enormous extension of personal taxation and of social security programs.”128 He was not wrong to say that there is a way to distribute government money to families to enable school choice, even if it is not as straightforward as funding a public school, as explained below in section two. c. Freedom of Choice Another theory focuses on the importance of freedom of choice in American society.129 School choice proponents emphasize individuality; school choice allows parents to benefit from state money allocated for educational costs to personalize their children’s education “rather than accepting an education system designed to teach to the average.”130 Economic effects aside, these supporters of school choice “view[] education primarily as a private good, the benefits of which are reaped by the child herself and her family,” and the most important reason for school choice is to protect parents’ freedom of choice.131 124 See generally Milton Friedman, The Role of Government in Education, in ECONOMICS AND THE PUBLIC INTEREST (Robert A. Solo ed.) (1955); Jeffrey R. Henig, The Local Dynamics of Choice, in Bruce Fuller et. al, WHO CHOOSES? WHO LOSES? CULTURE, INSTITUTIONS, AND THE UNEQUAL EFFECTS OF SCHOOL CHOICE, supra note 117, at 95 (discussing “the market metaphor”). 125 Friedman, supra note 124, at 127; see William McGurn, Milton Friedman’s School Choice Revolution, WALL ST. J. (April 3, 2023). 126 Friedman, supra note 124, at 130; Fuller, Elmore, & Orfield, supra note 117 at 12 (“Choice proponents also argue that, by unleashing market dynamics and incentives, schools will be held more accountable, . . . and school principals, presently entangled in bureaucratic rules and hog-tied by teacher unions, will be able to reward strong teachers and prune their schools of weak teachers.”). 127 Friedman, supra note 124, at 132. 128 Id. 129 Rob Reich, Common Schooling and Educational Choice as a Response to Pluralism, in School Choice Policies and Outcomes 21–22 (Walter Feinberg and Christopher Lubienski ed., 2008) (“[T]he equality and efficiency arguments have a virtual monopoly on the attention of policymakers and pundits. . . . [I]t is my contention that the legitimacy of school choice is founded in liberty.”). 130 Jeb Bush, School Choice is Sweeping the Nation from Florida to Utah, WALL ST. J. (Feb. 3, 2023), https://www.wsj.com/articles/school-choice-is-sweeping-the-nation-from-florida-to-utah-je b-bush-education-learning-students-children-parents-11675435984 [perma.cc/9VDJ-68Y8]. 131 See Levin, supra note 123; Reich, supra note 129. (“Permitting parents to select a school for their children is crucial to respecting the liberty interests of parents. . . . [L]iberal societies must protect some version of school choice because the normative significance of pluralism requires the state to respect the liberty interests of parents. . . .”). https://www.wsj.com/articles/school-choice-is-sweeping-the-nation-from-florida-to-utah-jeb-bush-education-learning-students-children-parents-11675435984 https://www.wsj.com/articles/school-choice-is-sweeping-the-nation-from-florida-to-utah-jeb-bush-education-learning-students-children-parents-11675435984 COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 100 All of these goals are worthwhile; states should certainly pursue educational equity, competitive schools, and robust freedom of choice for their residents. Importantly, states should pursue those goals for all of their residents, not just a select few. School choice may well be a great way to do that (again, this Article is neutral with regard to whether school choice itself is good or bad). School choice by way of tax credits, however, will not achieve these goals. 2. Why Do Tax Credits Fail to Achieve the Stated Goals of School Choice? This section describes some phenomena endemic to tax credits today: they inequitably distribute resources and are functionally unmonitored by states. Neither is accidental; both are collateral effects of tax credits’ being used to circumvent the Establishment Clause and no-aid clauses. As explained below, these phenomena make tax credits a horrible tool for achieving the goals of school choice. a. Inequitable Distribution By their nature, tax credits end up funding private schools attended by wealthy students in wealthy areas. Instead of serving their stated purposes, tax credits actually exacerbate educational inequity. They purport to throw a lifeline to students in struggling schools by either a) directly enabling them to leave a struggling public school, or b) improving that struggling public school by creating public-private competition. However, the real effects are quite different. Tax credits end up being a tool for the wealthy to direct more funds to their schools of choice, leaving the public schools struggling. The following subsections provide further detail, but the main point is that tax credits are an unmonitored tool selectively assisting the wealthy with directing state money toward their preferred schools. It is a stark reality that tax credits do not equitably benefit all students who would like to attend private schools. Their benefits are selectively concentrated on wealthier schools, not schools that have actual need for public aid. Schools that receive money from tax credit donations are both attended by families wealthy enough to pay significant income tax and organized enough to encourage donations. On the student level, the fact that scholarship eligibility is not always based on income means that tax credit scholarships do not necessarily benefit the neediest students. Assisting the upper-middle class with private school tuition is simply not what state tax law is for, and it does not even serve the goal of school choice— families who can afford private school already have a choice, and tax credits do not widen the choices of poorer families that they purport to benefit. At the school level, tax credits only benefit private schools both 1) attended by students from families wealthy enough to be motivated by tax credits and 2) organized enough to publicize, encourage, and obtain donations. Tax credits are celebrated as benefitting individual students and families,132 but, in reality, they 132 Charles J. Russo, M.Div., J.D., Ed.D. & William E. Thro, M.A., J.D., The Demise of the Blaine Amendment and A Triumph for Religious Freedom and School Choice: Espinoza v. Montana Department of Revenue, 46 U. DAYTON L. REV. 131, 163 (2021) (“It is, then, essential to afford parents greater, potentially life-changing opportunities to have their children educated in the schools of their choice, presumably because their values are consistent with those modeled and taught in the 2023] THE WORST CHOICE FOR SCHOOL CHOICE 101 benefit the private school attended by donors’ children by serving as an extra stream of revenue enabling decreased tuition costs. In many states, donations to tuition organizations do not go to a general fund to be used for all private school scholarships. Instead, taxpayers donating to a scholarship organization designate which particular private school they want their donation to support. Schools encourage current students’ families to donate, knowing those contributions will be earmarked for that school.133 In some states, donors can even designate the specific student to receive their donation.134 Schools must have wealthy students to benefit from tax credits because, by definition, only people who make enough money to pay income taxes are incentivized by tax credits. To a family who pays income tax, a tax credit is a form of redirecting tax dollars–donating one dollar means a dollar comes off the tax bill, and the family breaks even. But a family that does not pay income taxes loses that donated dollar; since no income tax would have been paid either way, the donated dollar is of little benefit. Thus, only families who pay income taxes have the economic incentive to make a donation that will be refunded at tax time.135 Because a family’s income must exceed a certain threshold to owe income taxes,136 and the amount owed increases with earnings, poorer families have no reason to donate to scholarship organizations. Even refundable tax credits—which reimburse filers with no state tax liability for the amount of their donation—put poor families in a worse position to participate. Refundable tax credits still require parting from a sum of money for much of the year, because donors do not receive their refund until tax time the following year.137 Not having access to an amount of money for some time may not be a problem for families with wealth beyond their needs, but, for others, it is likely prohibitive. When donors can designate a school, schools already attended by wealthier families will be the ones benefiting from tax credits. Schools frame tax credits as additional giving, which is natural to encourage at a private school attended by the schools they select.”); Isabel Chou, "Opportunity" for All?: How Tax Credit Scholarships Will Fare in New Jersey, 64 RUTGERS L. REV. 295, 300 (2011) (“For these parents, a voucher or scholarship would provide their children with an immediate opportunity to receive a better education at a private school.”). 133 COVENANT CHRISTIAN SCHOOL, supra note 111. 134 This problem leads to some tax credits’ working like vouchers in practice. When donors can designate the student to benefit from their donation, for example, there is no difference between that student receiving a voucher and receiving an allocation of money that was donated in exchange for a tax credit. 135 See William G. Gale, Tax Credits: Social Policy in Bad Disguise, BROOKINGS (February 16, 1999), https://www.brookings.edu/opinions/tax-credits-social-policy-in-bad-disguise/ [perma.cc/62 DH-EG92] (“Unless the credits are refundable—that is, unless they actually give people cash back instead of just reducing tax liabilities—they will not help low-income households.”). 136 For example, to owe $5,000 (the maximum tax refund under Georgia’s private school tax credit law) in Georgia income tax, a family must make $91,000 a year. Tax Tables & Georgia Tax Rate Schedule, https://dor.georgia.gov/document/document/2022-tax-tablepdf/download [perma.cc/M42 A-YK3D]. 137 It is true that refundable tax credits can be designed to help low-income households. See Andrew T. Hayashi & Justin J. Hopkins, Charitable Tax Deduction and Civic Engagement, 2023 U. ILL. L. REV. 1179, 1179 (2023). As explained below, even though the tax credit system can be manipulated to avoid the problems of excluding low-income households, direct funding is a simpler and easier way to distribute school choice. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 102 wealthy. But at a private school serving a low-income population, which may charge only nominal tuition, fundraising is less likely to come from parents of attending students.138 In Georgia, the leading scholarship organizations distribute scholarships on recommendation from their partner schools; scholarship seekers generally apply through the admissions department of the school they wish to attend, not through the scholarship organization.139 Even though the funds are given to the school to be used as scholarships, the school retains a lot of power in recommending scholarship recipients, and the money received from scholarship organizations can replace parts of the budget that otherwise would have gone to financial aid. The money coming from the scholarship organization is meant to be used for scholarships, but because the money comes from a private organization, and because the school has so much control over how it is used, the money may go to fund tuition for students who could have paid on their own, which over time keeps tuition low and serves as an extra stream of revenue for the school. In many states, including Georgia, there is no wealth-based eligibility requirement for a scholarship, so if donations to a school exceed student need, the school can allocate the excess funds among other students or reduce tuition generally. In that case, there is no reason to redistribute excess funds to students with need at other schools. In fact, it would be unwise to do so. The school is better off reducing its own tuition or otherwise taking advantage of the new revenue stream internally. Since money is fungible, schools will naturally reduce tuition or otherwise improve the school when provided with another revenue stream.140 Tax credits thus have the effect of driving down tuition costs at private schools already attended by the wealthy. The available data in Georgia bears this out. The 138 According to the U.S. Department of Education, the percentage of parents who report participating in school fundraising is over twenty points higher for parents at or above the poverty threshold than it is for parents below it. The percentage of parents participating in school fundraising also increases with level of parent education. Rachel Hanson and Chris Pugliese, Parent and Family Involvement in Education: 2019 (NCES 2020-076). U.S. Department of Education: National Center for Education Statistics, Parent and Family Involvement in Education: 2019, A-5 (2020) https://nc es.ed.gov/pubs2020/2020076full.pdf [perma.cc/U332-LYSW]. 139 GEORGIA GOAL SCHOLARSHIP PROGRAM INC [SCHOLARSHIP PROCESS] https://www.goalscholars hip.org/for_parents/page/scholarship-process [perma.cc/72D2-3M63]; APOGEE SCHOLARSHIP FUND: HOW TO APPLY FOR SCHOLARSHIPS https://www.apogee123.org/info/scholarships [perma.c c/5X4J-QTVR]. This also means that if a school does not partner with an SSO, it cannot receive scholarship money from the tax credit program. Georgia’s seven largest SSOs, which accounted for 89% of donations and 90% of scholarships from 2017–19, list over 500 partner schools on their websites. Greg S. Griffin & Leslie McGuire, SPECIAL REPORT ON QUALIFIED EDUCATION EXPENSE CREDIT AND STUDENT SCHOLARSHIP PROGRAM, GA. DEP’T OF AUDITS AND ACCOUNTS PERFORMANCE AUDIT DIV., 6 (2021), https://www.audits.ga.gov/ReportSearch/download/25619 [p erma.cc/P584-9Z25] (State auditors could not determine what “percentage of private schools in the state this represents, however, because GaDOE does not maintain an accurate and complete list of all private schools operating in the state.”). 140 We recognize that the question of how much of a subsidy for the wealthy is present varies from state program to state program. Some states, like Georgia and Arizona, have no means testing at all, and other have very diverse factors, some of which functionally limit the program to the middle class or lower. But it seems to us that all tuition tax credit programs exclude the poor who pay no taxes; these programs are not subsidies for the wealthy only, but the middle class as well. Our point, however, is that it is the desperately poor, located frequently in public school districts that are very low quality, who need these programs the most and are least likely to gain from tuition tax credit. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 103 percentage of scholarships in Georgia awarded to students in the lowest income quartile decreased, and the percentage going to students in the highest income quartile increased, from 2015 to 2018.141 Fifty-eight percent of scholarships assisted students in families earning above the state’s median income.142 Relatedly, tax credits incentivize private schools to admit and retain wealthy students. As discussed above, the wealthy are uniquely positioned to make donations in exchange for tax credits. Private schools know that the wealthier their student population is, the more funding they can expect to receive from scholarship organizations. When private schools recruit students accordingly, economic stratification between schools will only worsen. Wealthy families can leverage their status when applying, and private schools know admitting wealthy students is likely to lead to extra revenue from scholarship organizations. As the contribution limit increases, so will the incentive for private schools to admit rich applicants. Private schools will segregate into those with wealthy students, who can fund a lot of tax credits, and those without, who will not receive any funding from scholarship organizations. Effectively, the wealthy get reimbursed by the state for payments made to their private school of choice. Private schools receiving extra revenue from scholarship organizations can keep tuition low, but students at poorer and middle class private schools must continue to pay rising tuition without reimbursements from the state. In short, tax credits make subsidies of private education available exclusively to the wealthy. Of course, not every private school’s attendees are always in the same tax bracket—it could be possible that the ability to designate a school to receive a donation enables the richer families at a private school to benefit poorer families attending that same school. However, due to the “increasingly polarized pattern of school enrollment[,]” private schools are “increasingly segregated by income.”143 The reality is that private schools attended by affluent students are not typically the same private schools attended by middle- or low-income students. Furthermore, few schools that serve the impoverished also serve the wealthy. There are some private schools that focus on serving low-income students. The Cristo Rey Network of high schools, for example, “delivers . . . education in the Catholic tradition for students with limited economic resources.”144 It charges nominal tuition which “generally provides less than [ten] percent of the cost of 141 Samantha Sunne & Donnell Suggs, Scholarship Tax Credit Program: $600M With Little Oversight, THE CURRENT (Sep. 23, 2020, updated Feb. 3, 2022), https://thecurrentga.org/2020/09/2 3/scholarship-tax-credit-program-600m-with-no-oversight/ [perma.cc/DA43-JPCE]. 142 Id. 143 Richard J. Murnane, Sean F. Reardon, Preeya P. Mbekeani & Anne Lamb, Who Goes to Private School? Long-term enrollment trends by family income, EDUCATION NEXT, 18(4), 58 –66 (2018); see also Richard J. Murnane, Sean F. Reardon, Long-Term Trends in Private School Enrollments by Family Income, AERA OPEN (2018) https://files.eric.ed.gov/fulltext/EJ1194132.pdf [perma.cc/E R4K-SWFT] (comparing “high-priced private schools that children from affluent families attend and the lower-priced . . . private schools that students from lower-income families attend”). 144 Mission and History, CRISTO REY NETWORK, https://www.cristoreynetwork.org/about/mission- history [perma.cc/LDJ3-L9XS]. See, e.g., Why All Girls, MOTHER CAROLINE ACADEMY AND INFORMATION CENTER, https://www.mcaec.org/why-all-girls/ [perma.cc/HL7Y-JE7N]; Admissions Criteria MHS, https://www.mhskids.org/admissions/admissions-criteria/ [perma.cc/4S9U-9VV2] (Milton Hershey School applicants must come from low-income family). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 104 educating a student.”145 Such a school, which does not receive significant tuition payments and is designed for students who would not otherwise have the choice to attend private school, seems an ideal candidate for public aid. But the average income of a family of four for a Cristo Rey student is $38,000,146 which in Georgia, means owing no more than $2,012.50 in state income taxes,147 and less in fact. Contributing anything more than that to a scholarship organization would be a loss, so Cristo Rey can expect, at best, SSO contributions of around $2,000. A school with an average family income of $95,000, however, can expect more—a family with that income would owe $5,000 in state income taxes, so that school could expect a $5,000 donation—the law’s maximum—at no cost to the parents, and even a gain.148 In any event, as described below, many scholarships funded by tax credit donations are not awarded to students who actually need a scholarship to attend private school. Just as this mechanism benefits wealthy families and schools, it also benefits schools with the resources and organization to encourage families to donate and designate the school. Convincing parents to donate thousands of dollars to a student scholarship organization, even though they will receive those dollars back at tax time, takes a certain level of organization that typically only private schools 145 Sarah Stewart, Work Study Program Provides Revenue to School Experience to Students, NATIONAL ASSOCIATION OF INDEPENDENT SCHOOLS https://www.nais.org/magazine/independent-s chool/summer-2016/work-study-program-provides-revenue-to-school-and/ [perma.cc/UP5G-M7G B]. 146 Frequently Asked Questions, CRISTO REY NETWORK, https://www.cristoreynetwork.org/about/ faqs [perma.cc/7F79-VGM3]. 147 Tax Rate Schedule, GA. DEP’T OF JUST., https://dor.georgia.gov/tax-tables-georgia-tax-rate-sche dule [perma.cc/YPU4-HXH5]. 148 Gains are possible if contributions for tax credits are also deductible as charitable contributions under § 501(c)(3) of the Internal Revenue Code. In that case, taxpayers reduce their federal taxable income by the amount of the donation, despite losing zero dollars on the donation because they received the same amount back from the state. As of August 2019, per U.S. Treasury Regulation, individual contributions to charitable organizations for which the taxpayer receives a state tax credit are not deductible from federal taxable income. 26 C.F.R. § 1.170A-1(h)(3)(i); see also 47A C.J.S. Internal Revenue § 204 (“State tax credit as a ‘return benefit.’”). In some cases, individuals may still count their contribution as payment of state or local taxes. 26 C.F.R. § 1.170A-1(h)(3)(ix) (“Safe harbor for individuals”). Contributions by C Corporations and some passthrough entities may also still be claimed as business expenses. 26 C.F.R. § 1.162-15(a)(3) (“Safe harbors for C Corporations and specified passthrough entities making payments in exchange for State or local tax credits.”). Tuition organizations convey that contributions are not federally deductible as charitable contributions with varying clarity. See, e.g., Donor Frequently Asked Questions, ARIZ. PRIVATE SCH. ORG., https://apsto.org/individualdonorfaq [perma.cc/2JYT-EA7H] (advising donors to consult with a tax advisor for deductibility of their contributions); Frequently Asked Questions, ARIZ. TAX CREDIT https://aztxcr.org/faqs/ [perma.cc/C8X9-WVT7] (“Since we are a 501(c)(3) non- profit, your donation may be eligible for a federal tax deduction. . . .”). Gains would also be possible if states allowed tax credits to be claimed in amounts that exceeded a taxpayer’s state tax liability, which does not appear to be common. Georgia, for example, does not allow the claiming of tax credits in excess of a taxpayer’s state income tax liability, but it does allow unused tax credits to be used against tax liability for five years following the donation. GA Code § 48-7-29.16(e) (2022). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 105 with wealthy populations have.149 To receive funds from tax credit donations, a private school must partner with an SSO and have donors designate the school with their contributions. Both of those require resources. On private schools’ webpages encouraging donations to scholarship organizations, for example, there is usually a school administrator for donors to contact.150 At least one school provides information on an interest-free loan to cover the donation until the tax credit is paid.151 These schools clearly invest resources in communicating the availability of tax credits and encouraging donations. Schools serving wealthier families have also been found to have more involved parents, which makes requesting donations easier.152 So not only are poorer private schools less likely to enroll students from families willing and able to donate, they are less likely to have the resources to communicate to parents that the donation opportunity even exists. Because of the hoops that a school must jump through to garner funding from donations, the donations end up going to private schools with money to spend on fundraising outreach and the organizational base to manage contributions. Often, schools with the organizational infrastructure necessary to benefit from tax credits are religious schools. Parents at religious schools report more participation in school fundraising than parents at non-religious schools do, whether because of a shared purpose, strong community, or peer pressure.153 Tax credits may have been intended to benefit all schools, but because the aid is based on the tax code, it is used disproportionately by schools with more centralized communities and more resources, which tend to be the schools that are not in need of public aid. In hindsight, these hoops look like an unfortunate complication that (perhaps accidentally) put wealthier schools in a better position to benefit from tax credits. The inequitable result may not have been intended, but the hoops definitely were. Outsourcing the fundraising to schools and allowing private organizations to manage donations and award scholarships were fundamental to states’ acceptance of tax credit systems because they distanced the government from the aid. These complications were features of the tax credit system because they gave the impression that the tax credit scholarships were not state money—the funds were raised and managed by other parties, so they could not have been used to connect the government to religious schools. 149 According to the U.S. Dep’t of Educ., the percentage of parents who reported receiving communications from the school addressed to all parents increased with the level of parent education, and the percentage of parents who reported receiving such communications was over ten points higher for parents above the poverty threshold than for parents below it. Rachel Hanson & Chris Pugliese, Parent and Family Involvement in Education: 2019 (2020). 150 See, e.g., Westminster, Ga. Private Sch. Tax Credit Program, https://www.westminster.net/supp ort-westminster/georgia-private-school-tax-credit-program [perma.cc/Z7FK-2WDS] (“Director for Annual Giving”); Learn More About ALEF Fund, https://drive.google.com/file/d/1EtOKo39_uN7r WdFbsunCaoed1MuUyGW4/view [perma.cc/HZG5-YEVY]. 151 Alumni Donation Info., THE WEBER SCH., https://www.weberschool.org/alumni/alumni-giving [perma.cc/J2DR-THUN]. 152 Patricia A. Bauch & Ellen B. Goldring, Parent Involvement and School Responsiveness: Facilitating the Home-School Connection in Schools of Choice, 17 EDUC. EVALUATION AND POL’Y ANALYSIS 1, 1-21 (1995). 153 Parents at religious schools report more participation in school fundraising. INSTITUTE OF EDUCATION SCIENCES, PARENT AND FAMILY INVOLVEMENT IN EDUCATION: 2019 (2020). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 106 In a system like Georgia’s, where parents can designate the school their contribution should benefit, whether a school receives funds through the tax credit program is not based on need—it is based on whether the school has wealthy enough parents to benefit from a tax credit and whether the school has spent enough resources on communicating and encouraging donations to the program. The Georgia statute enacting the tax credit program does not mention anything about a donor designating a school to receive her donation (although it does provide that no school or scholarship organization may represent that the contribution will be directed toward any individual).154 This is another feature of tax credits that, far from being an accident, made them appeal to state governments. Because private organizations manage the funds, the state can disclaim responsibility for donations being concentrated at wealthy schools. While the statute does not prohibit designating schools to receive donations, the practice is much more naturally attributable to the private organizations and schools themselves than the government. This is just another way that outsourcing the management of tax credit donations allows state governments to distance themselves from the enterprise and convey compliance with the Establishment Clause and state no-aid provisions. On the student level, donations for tax credits are not directed to students who need them to have school choice. Eligibility for these scholarships often, and increasingly, does not depend at all on a student’s wealth, so many donation-funded scholarships benefit students whose families could have financed a private education without assistance.155 As a result, public funds end up paying for private school for students who might not actually need them to have school choice, and do not contribute to school choice for the impoverished and schools that serve the impoverished. In Georgia, for example, students are eligible for a Qualified Education Expense Tax Credit (“QEETC”) scholarship if they attended a public school for at least six weeks immediately before receiving the scholarship.156 Once donations are made to scholarship organizations, “state law does not dictate how scholarships are distributed.”157 In 2017, 56.7% of scholarship recipients’ families were in the third or fourth percentile for income.158 One of Georgia’s leading scholarship organizations awarded $10,932,789 in scholarships in 2019. 47% of that amount was awarded to families with incomes over 400% of the federal poverty line.159 154 Ga. Code Ann. § 48-7-29.16 (West). 155 Garnett, supra note 5, at 17. (“All, most, or many students are now eligible to participate in private-school-choice programs in several states, and several more states seem poised to join this expansive parental-choice roster.”); The Editorial Board, The Rising Demand for Sch. Choice, WALL ST. J. (July 30, 2023), https://www.wsj.com/articles/the-rising-demand-for-school-choice-ari zona-florida-implementation-esa-1c47ce5f [perma.cc/E39G-PPCD] (detailing rise in participation in Indiana’s school choice program after vouchers became “nearly universal” in part because of a raised income cap). 156 Qualified Education Expense Tax Credit, ED CHOICE https://www.edchoice.org/school-choice/p rograms/georgia-qualified-education-expense-tax-credit/#student_eligibility [perma.cc/9MCA-VX YW]. 157 Griffin & McGuire, supra note 139. 158 Clair Suggs, Shifting Public Funds to Private Schools: High Costs, Poor Track Record, GA. BUDGET & POL’Y INST. (April 26, 2018), https://gbpi.org/shifting-public-funds-to-private-schools- high-costs-poor-track-record/ [perma.cc/BMX7-SDGF]. 159 Griffin & McGuire, supra note 139. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 107 16% of it went to families with incomes lower than 125% of the federal poverty line.160 One Georgia scholarship organization’s website seems to indicate that participation in the tax credit program means both donating and applying for a scholarship.161 These facts show that scholarships funded by donations for tax credits are not directed toward families who actually need them to have a choice in school attendance. Rather, all these statistics show that donating to a scholarship organization is understood to come with financial benefit to the donor. Some states provide more direction to scholarship organizations regarding eligibility for scholarships. Eligibility for Pennsylvania’s Opportunity Scholarship, for example, is based on whether a student lives within the attendance boundary of a failing public school.162 The Florida Tax Credit Scholarship program used to be available only to students receiving other state benefits, who have a household income level that does not exceed 375% of the federal poverty level, or who are in foster care.163 However, beginning in July 2023, any Florida resident “eligible to enroll in kindergarten through grade 12 in a public school in [Florida]” is eligible, though “priority must be given” to students at certain low income levels or who are in foster care.164 Where tax credit programs do not have income-based eligibility requirements, the impact is easily seen. Arizona, for instance, has two such programs. One leading scholarship organization there awarded 75.05% of its total scholarships for one of these programs to students whose household income exceeded 342.25% of the poverty level in fiscal year 2020–21.165 To be sure, the fact that scholarships go to wealthy students can be easily attributable to schools and scholarship organizations instead of to the state. In Arizona, scholarship organizations encourage families seeking scholarships to create a website encouraging donations to their specific children.166 Yet, importantly, schools with the organizational wherewithal to advise students of this opportunity, as well as students with familial help and technological access, are in an advantaged position to benefit. Because any parents who pay state taxes gladly take advantage of this program to benefit the school their own children attend, 160 Id. 161 GaSSO Scholarship Process Details, GA. STUDENT SCHOLARSHIP ORG., https://www.georgiass o.com/donors [perma.cc/FB4M-MSVN] (“Step 3: . . . Donate on the [SSO’s] website Step 4: . . . Complete Scholarship application on [SSO’s] website”). 162 24 Pa. Stat. Ann. § 20-2009-B (West); Overview of the Opportunity Scholarship Tax Credit Program, PA. DEP’T OF EDUC., https://www.education.pa.gov/K-12/Opportunity%20Scholarship% 20Tax%20Credit%20Program/Pages/Overview-of-the-Opportunity-Scholarship-Tax-Credit-Progr am.aspx [perma.cc/2VDN-LEF9] (“A low-achieving school is defined as a public elementary or secondary school in Pennsylvania that ranks in the bottom 15 percent of its designation as an elementary or secondary school based on combined . . . scores on the most recent annual assessment . . . .”). 163 Fla. Stat. Ann. § 1002.395 (amended 2023) (West). 164 Id. at § 1002.395(3)(2). 165 Financial Reports, ARIZ. PRIVATE SCH. TUITION ORG., https://apsto.org/FinancialReport [perma .cc/4V9L-Z8AU]. 166 How it Works, ARIZ. TUITION ORG., https://www.azto.org/applicants/how-it-works [perma.cc/6 N6T-BXHK]. Arizona also implemented universal vouchers in 2023, separate from its tax credit programs. See Sarah Mervosh, $7,200 for Every Student: Arizona’s Ultimate Experiment in School Choice, N.Y. TIMES (July 24, 2023), https://www.nytimes.com/2023/07/24/us/arizona-private-scho ol-vouchers.html [perma.cc/39DH-85TF]. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 108 schools with wealthy parents gain from this program the most, increasing governmental revenue to their schools and thus reducing tuition in the long run.167 Again, this is a feature, not a bug; it is a result of the thought-necessary separation between religious schools and state governments that tax credits provide. The more the onus is on the family and the school to reach out for donations, the less it appears connected to the state. Many states have provisions that aim to prevent donors from directing their contribution toward their own children.168 These provisions are meant to preclude “quid pro quos” under which donors could fund their own child’s education with donations that will be refunded in tax credits.169 Donations to student scholarship organizations cannot substitute for a family’s tuition payments, for instance.170 However, while on paper a contribution does not fund the donor’s child’s tuition, such donations are still an extra revenue source for the school that will abate tuition costs. Parents who pay income taxes can contribute to that revenue source for free (or better). The tax credit donations may not be formal quid pro quos, but they give 167 We are aware of the fact that scholarship granting organizations—which collect the donations that generate the tuition tax credits—typically cannot be captured by only one school, but we are not persuaded that this is at all important. No matter how many schools are under one umbrella, we think it is exceedingly likely that the scholarship granting organizations all honor the requests of parents as to which school should receive their donation, thus ensuring that tuition tax credits go to specific schools that parents direct. Solving this issue (if it is a problem) would be easy to do, but we suspect that tuition tax credits would lose their attractiveness in fact if there was no quid-pro- quo to the parents’ children’s school at all. Every parent knows that all money that goes to their children’s school reduces their burden. See generally Garnett, supra note 5. 168 In Georgia, Donors may not designate their contribution “for the direct benefit of any particular individual,” and a student scholarship organization may not represent or direct a school to represent that the taxpayer will receive a scholarship for the direct benefit of any particular individual. Ga. Code Ann. § 48-7-29.16 (West). “This prohibition is designed to prevent a guardian from using the tax credit as a tax avoidance mechanism while ensuring a scholarship for a dependent.” Griffin & McGuire, supra note 139, at 25. Scholarship organizations are not required to report to the state whether donors designate funds to particular students. Garnett, supra note 5, at 3. In Arizona, donors to school tuition organizations may not designate the donation for their own dependent, designate a student beneficiary as a condition of their donation, or “agree with another person to designate each other’s contribution to the school tuition organization for the direct benefit of each other’s dependent, a practice commonly known as swapping.” ARIZ. DEP’T OF REVENUE, School Tax Credits, 4 (2022) https://azdor.gov/sites/default/files/2023-03/PUBLICATIO N_707.pdf [perma.cc/P4VT-2FW2]. In Florida, “[t]he taxpayer making the contribution may not designate a specific child as the beneficiary of the contribution.” Fla. Stat. Ann. § 1002.395 (2)(f) (West). 169 If parents could fund their own child’s tuition through their contribution to a scholarship organization, which would be reimbursed to those parents by the state, it would function just as a voucher does. Instead of the government giving money to the family to spend on the child’s education, the parent gives money to the scholarship organization that will fund the child’s tuition, then the parent is reimbursed for that money by the state in the form of a tax credit. Arizona’s system, by making quid pro quo or donations for one’s own child illegal, seeks to prevent this from happening but does a poor job at enforcement. A much simpler solution, that we propose below, is to simply allow vouchers, making the private middleman unnecessary. We see no good reason why Arizona would want to support a tax credit system for school choice while not allowing vouchers. Arizona is moving in this direction; as of July 2023, the state has introduced universal vouchers through a program separate from its tax credits. Mervosh, supra note 166. 170 See Ga. Code Ann. § 48-7-29.16 (West). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 109 the impression to parents that they are an avenue to support the school with public money.171 In short, critics were not wrong that there is a way to distribute government money to families to enable school choice, even if it is not as straightforward as funding a public school. But when states felt a need to separate themselves from funding religious education, they inevitably separated themselves from regulating how state aid was directed to private schools. That separation exacerbated the lack of “administrative machinery” that Milton Friedman recognized.172 And it just so happens that the avenue chosen by states to fund private schools—while keeping their distance—was tax credits. That choice ends up disserving the original purpose. Friedman’s theory relied on everyone having choice, but this is facilitated most easily by vouchers, not by tax credits. Tax credits are easily used by the rich and much more difficult for the poor to take advantage of. The economic reasons for choice favor vouchers, accessible equally by all, over tax credits. Even when tax credit scholarships are awarded to needy students, they do not cover the full cost of private school tuition.173 When the scholarship still leaves some cost for attending private school, families with money to cover that cost may eagerly use it to switch out of a struggling public school. That cost may be prohibitive to others.174 This is another way that tax credits focus their benefits on the middle class and the rich but are of no use to the truly poor. Tax credit scholarships are limited in value both because they depend on private fundraising and because the amount that can be donated is usually capped. By contrast, vouchers have neither of those limitations and are usually awarded in larger amounts than tax credit scholarships.175 b. Fraud Concerns Suffice to say, tax credit systems do not always work the way they are intended to on paper. Another major issue concerns their unique vulnerability to fraud. Despite attempts to monitor scholarship organizations and private schools, 171 See, e.g., Georgia GOAL, MOUNT PISGAH CHRISTIAN SCHOOL, https://www.mountpisgahschool .org/giving/georgia-goal [perma.cc/6CR3-XK4T] (encouraging tax credit contributions that “make it possible for [the school] to . . . provide additional programs and participation in the arts and athletics by increased enrollment [and] have greater financial flexibility in funding and supporting other [school] programs and opportunities”); Alumni Giving, THE WEBER SCHOOL, https://www.we berschool.org/alumni/alumni-giving [perma.cc/FLP7-XCNA] (“Supporting [the school] through [tax credits] is just like donating money . . . and it won’t cost you anything.”); COVENANT CHRISTIAN SCHOOL, supra note 111; Georgia Private School Tax Credit Program, THE WESTMINSTER SCHOOLS, https://www.westminster.net/support-westminster/georgia-private-school-tax-credit-pro gram [perma.cc/Q5P3-FS8H] (“[T]he Georgia Private School Tax Credit Program, combined with support from [our school’s] endowed funds and annual [school] Fund, plays a critical role in the School’s overall financial aid program. Participation in this program makes a real impact on the School’s ability to enroll bright, motivated, and curious students regardless of their family’s ability to pay full tuition.”). 172 Friedman, supra note 124, at 132. 173 Garnett, supra note 5, at 6. 174 Id. at 7 (“[A]bsent financial assistance, the scholarships provided through private-school-choice programs place many private schools well out of the reach of many scholarship recipients.”). 175 Id. at 6. about:blank COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 110 the features of the system interfere with any real ability to ensure funds are being spent as intended. For example, Georgia lacks adequate controls to prevent taxpayers from claiming credits in excess of their tax liability, which could lead to some taxpayers actually turning a profit through the tax credit system.176 Georgia law also does not require scholarship organizations to report some important information, including whether only eligible students receive funds and assurances that donors do not designate their funds to a particular student.177 Even when state law does require disclosures, the information is minimally helpful. Georgia requires independent compliance audits from scholarship organizations, but the reports vary greatly and lack detail.178 With weak oversight comes weak enforcement. Georgia, for instance, does not always verify that reports submitted by scholarship organizations meet legal requirements. When the state does verify the reports, punishment for non- compliance is simply being removed from the list of active scholarship organizations.179 Compare that to other types of fraud with government money, like Medicaid fraud. Georgia has an office that “rigorously reviews, investigates, and audits Medicaid providers and recipients to uncover criminal conduct, administrative wrongdoing, poor management practices, and other waste, fraud, and abuse.”180 There is no analogous office for scholarship organizations. As scholarship organizations do not technically handle government money—as they receive donations from private parties who later claim tax credits—the typically robust protections that accompany government benefit programs are absent. Courts have held that tax credit systems like this are not government spending. Some state courts, including Georgia’s, have upheld tax credits while acknowledging that direct payments would be unconstitutional under the same circumstances.181 These decisions made hay of the formal distinction between tax benefits and appropriations.182 Both forms of aid are means to the end of financing 176 See Griffin & McGuire, supra note 139, at 11. State law limits what taxpayers can claim based on their tax liability, but “controls are not present to prevent [satisfaction of claims in excess of tax liability].” 177 See id. at 19 (“SSOs have several requirements directly related to scholarship management practices that are not required to be reported to the state. . . . [L]ittle information is coming to the state regarding whether the SSOs are actually complying with these scholarship management requires under other code sections.”). 178 See id. at 15–16 (“Using data from the compliance audits, we attempted to compile a complete record of administrative expenses and revenues by SSO; however, it was not possible because of a lack of consistency and detail in the reports. . . . While we could have run a calculation, it is not clear that all SSOs were categorizing funds similarly because the statute does not define the terms.”) 179 See id. at 2 (“We also found that DOR did not ensure that compliance audits verified and reported on all relevant requirements. Finally, noncompliant SSOs were not always removed from GaDOE’s list of active participants in a timely manner.”). 180 Office of Inspector General, GEORGIA DEPARTMENT OF COMMUNITY HEALTH, https://dch.georgi a.gov/office-inspector-general [perma.cc/Z2S8-RNLU]. 181 See generally Gaddy v. Ga. Dep’t of Revenue, 802 S.E.2d 225 (2017); Kotterman v. Killian, 972 P.2d 606 (1999); McCall v. Scott, 199 So. 3d 359 (Fla. Dist. Ct. App. 2016). 182 To others, that same distinction is “one in search of a difference.” Ariz. Christian Sch. Tuition Org. v. Winn, 563 U.S. 125, 157 (2011) (Kagan, J., dissenting) (“[T]argeted tax breaks are often ‘economically and functionally indistinguishable from a direct monetary subsidy.’ . . . [H]ere too 2023] THE WORST CHOICE FOR SCHOOL CHOICE 111 an organization. Decisions denying taxpayer standing, however, “enable[] the government to end-run” the rule that taxpayers can challenge government spending in support of religion.183 As another example, Arizona prohibits donors from designating their own children as scholarship recipients, but it also prohibits donors from agreeing with another donor that they will each designate each other’s child, a practice known as “swapping.”184 Nonetheless, there is widespread evidence that the practice persists, which is unsurprising since it is enforced essentially by the honor system.185 Arizona scholarship organizations include language about the impermissibility of swapping on their websites,186 but no strong enforcement mechanism exists because the whole system is privatized. Again, this lack of regulation is a feature of tax credits once thought necessary for compliance with the Establishment Clause or state no-aid provisions. To keep tax credits compliant with those laws, it was necessary that they were not considered government money. The responsibility to ensure compliance with rules, like the one against swapping, falls on private organizations, therefore, and not the government. form prevails over substance, and differences that make no difference determine access to the Judiciary.”); To them, tax credits are constitutionally no different from government funding. Kotterman v. Killian, 193 Ariz. 273, 309 (1999) (Feldman, J., dissenting) (“[A tax credit] is a direct government subsidy limited to supporting the very causes the state’s constitution forbids the government to support.”); Stanley S. Surrey, Tax Incentives As A Device for Implementing Government Policy: A Comparison with Direct Government Expenditures, 83 HARV. L. REV. 705, 717 (1970) (“A dollar is a dollar—both for the person who receives it and the government that pays it, whether the dollar comes with a tax credit label or a direct expenditure label.”); Opinion of the Justs. to the Senate, 401 Mass. 1201, 1203–04 (1987) (“The fact that the expenditure here takes the form of a tax deduction rather than a direct payment out of the Commonwealth's treasury does not alter the result, for it has been recognized that the tax subsidies or tax expenditures of this sort are the practical equivalent of direct government grants.”). Under that view, decisions upholding tax credits as not implicating no-aid provisions and the Establishment Clause would have been wrong all along. The correctness of that view is immaterial to this Article. Regardless of the merits of constitutional decisions on tax credits, they are inefficient and rife with fraud. These citations serve only to illustrate objections to one of the historical legal reasons that tax credits became so popular— objections to the idea that tax credits were somehow legal where direct funding was not. These objections are moot after Espinoza and Carson, which hold that states may, and sometimes must, subsidize religious schools regardless of the form that subsidy takes. Our point is not that these objections were right or wrong, but that tax credits are a poor policy choice, and now that direct funding is clearly permissible, it should replace old tax credit systems. 183 Ariz. Christian Sch. Tuition Org. v. Winn, 563 U.S. at 148 (2011) (Kagan, J., dissenting.). 184 Donation Restriction, ARIZONA TUITION ORGANIZATION, https://www.azto.org/contributors/don ation-restrictions [perma.cc/EV6S-TFW6]. 185 Michelle Reese, Private School Tax Credits Rife with Abuse, EAST VALLEY TRIBUNE (Dec. 12, 2017), https://www.eastvalleytribune.com/news/private-school-tax-credits-rife-with-abuse/article_ 7debd2e5-d000-5aed-b813-a0d252377755.html [perma.cc/CL7M-4PRK]. 186 See Donation Restriction, supra note 184 (“Arizona Tuition Organization (AZTO) does not condone, endorse, or accept the practice of swapping donations.”); Donor FAQ, ARIZONA PRIVATE SCHOOL TUITION ORGANIZATION, https://apsto.org/individualdonorfaq [perma.cc/4XK3-YMKW] (“Unfortunately, a donor cannot claim a tax credit if the donation benefits their own dependent or they ‘swap’ donations with other parents.”). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 112 3. These Concerns Show that Tax Credits are a Poor Instrument, Regardless of a State’s Reason for Implementing School Choice. The flaws outlined above make tuition tax credits a poor mechanism for achieving the goals of school choice: equity, pure economic growth, and individual choice. Below, we explore the reasons for this in greater detail. a. Equity This is most obvious when it comes to the goal of equity. As explained above, schools that have wealthier students and that are more organized are in a better position to take advantage of tuition tax credits, within the letter of the law or otherwise.187 Whether fraud outright violates a statute or exists as an undetectable wink-and-nod that tuition will stay low if parents donate, it is not an equitable way to distribute resources. Either way, schools with wealthy populations and organizational chops can count on a separate stream of revenue coming from the state, whereas schools serving the poor cannot. This outcome is exactly the opposite of the equity sought by school choice proponents. Instead of the public aid being directed to the private schools that need it the most, funds are disproportionately directed to the schools and students with more wealth than most. If tax credits in their current form are the only option, the goal of equity would actually be better served by having no aid to private schools at all. With no aid, resources intended for needy schools would at least not be redirected toward wealthy ones. b. Pure Economic Growth and Freedom of Choice Even if a state funds school choice on a purely economic theory, without any desire to equitably distribute resources, tax credits remain an inferior option to direct funding. For the reasons set out above, tax credits will only be available to some of the population. If a state believes that choice will foster healthy competition among schools and allow the best to rise to the top, that interest should be heightened for low-performing schools. Using tax credits instead of direct funding ensures that desired effects are had only on schools that 1) serve families wealthy enough to pay income tax and 2) are organized enough to take advantage of tax credits. Economic competition already exists in the schools where families pay sizeable tuition; those families had school choice regardless of government funding, and they exercised it by choosing a certain private school. To truly foster competition among schools, school choice must create a choice where there was not one before. When it comes to pressuring public schools to improve, this means giving a meaningful choice to students who previously only 187 Recent reports note that Yeshiva schools have other sources of public funding. See Brian M. Rosenthal, How Hasidic Schools Reaped a Windfall of Special Education Funding, N.Y. TIMES (Dec. 30, 2022), https://www.nytimes.com/2022/12/29/nyregion/hasidic-orthodox-jewish-special-e ducation.html [perma.cc/K2JS-653R]; Eliza Shapiro & Brian M. Rosenthal, In Hasidic Enclaves, Failing Private Schools Flush With Public Money, N.Y. TIMES (Sept. 12, 2022), https://www.nytim es.com/2022/09/11/nyregion/hasidic-yeshivas-schools-new-york.html [perma.cc/GXQ6-J44A]; Brian M. Rosenthal & Eliza Shapiro, Hasidic School to Pay $8 Million After Admitting to Widespread Fraud, N.Y. TIMES (Oct. 12, 2022), https://www.nytimes.com/2022/10/24/nyregion/ha sidic-yeshiva-fraud-central-united-talmudical-academy.html [perma.cc/Q9E6-JBHX]. Legal or otherwise, it takes institutional infrastructure to benefit from public funds. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 113 had the option of public school. Healthy economic competition is also still needed among struggling low-income private schools, which are not well set up to benefit from tax credits. If a state wants those schools to improve by competing for government money, it should concentrate that money on low-performing schools that both need to improve and need the money. Tax credits do the opposite. This is also the reason that tax credits do not serve the goal of individual freedom of choice—they only actually create choice for people within a certain tier of wealth. Consider Georgia’s or Arizona’s system, for instance, in which donors can direct their donations to a school that their children already attend. The donor taking advantage of the tax credit already attends that school and pays tuition. The donation through the tax credit does not necessitate that someone from a low- performing private school will transfer into the school that received the donation. The school that received the donation might direct those funds toward financial aid for existing students. A poor private school may provide a high-quality education, but because of the families it serves, it does not stand to benefit from tax credits. In short, the purely economic theory is not served by tax credits because its effects are focused on one specific area, which already enjoys a healthy level of competition. Another reason that tax credits do not foster economic competition is the very problem that Milton Friedman anticipated: a lack of regulation and states’ inability to ensure that schools actually spend designated funds the way they are allowed to. The administrative issues outlined above are a direct result of states’ beliefs that they were not permitted to fund religious education. They also show how instead of addressing the concern that states would not be able to regulate money allocated to private parties for education, they exacerbate it. c. Inefficiency The tax credit structure also means that some donated money will go to a scholarship organization’s operating costs instead of to actual scholarships. As the number of scholarship organizations increases, so does the amount of money lost to overhead.188 In Georgia, for example, scholarship organizations must use at least 92% of donated money for scholarships.189 Without a tax credit system, there would be no need to sacrifice the remaining eight percent to a scholarship organization; if everyone paid their full tax liability and the state assisted private schools directly, it could use one hundred percent of the amount that would have been donated for that purpose. Taking out the private third party would mean more money would be available for the government to support the private schools that need it. Moreover, a 2021 Georgia audit found that it was impossible to determine whether the rate that scholarship organizations retain was reasonable because of insufficient data on scholarship organizations’ revenue and expenses.190 This finding reveals a much larger inefficiency of tax credit programs: they require extensive monitoring but remain vulnerable to waste and fraud. Administration of direct funding will cost a state something as well. If a state administers the program, however, the required monitoring will be much more 188 Garnett, supra note 5, at 14 (“The proliferation of SGOs competing for taxpayer donations dilutes the impact of scholarship tax-credit programs.”). 189 Ga. Code Ann. § 20-2A-1 (West); Ga. Code Ann. § 20-2A-2 (West). 190 See Griffin & McGuire, supra note 139, at 15. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 114 achievable, and it should be the state’s, not a private entity’s role, to administer social benefits.191 All of these issues are reasons that tax credits for private education are just not worth their cost. For every dollar the state gives up in tax revenue, it should see a dollar of improvement in education, and it should be progressive in its aid— schools that serve the poor should get more, or at least the same, aid. In theory, school choice should reduce the number of students attending failing public schools, allowing those public schools to spend more per student and improve, while also supporting students attending private schools. The results, however, show that that is not the reality. “[S]tudents in public schools that have more school choice do not perform noticeably better than students in public schools without school choice,”192 and school choice does not reliably improve educational outcomes for students who choose private school.193 Tax credit systems also result in public funding of private schools that are not subject to the same regulations as public schools. The government regulates public schools for a reason, but tax credits use public money to fund schools that are not required to publish data on finances, attendance, curriculum, or test scores.194 Naturally, it is difficult to determine whether students who choose private school under school choice programs fare better than they would have at public school.195 Even if choice does improve long-term educational outcomes for students who choose private school,196 the fact remains that the privilege of choice is not equally distributed. Recipients of tax credit scholarships are often students who do not need scholarships to attend private school, and poorer, tuition-funded private 191 Interestingly, one of the most successful tax credit programs thrives in part because it operates though only one scholarship organization. Florida’s “Step Up For Students” is the only scholarship organization involved in one of Florida’s tuition tax credit programs. According to a Manhattan Institute Report, the scholarship organization “operates much like a privately operated statewide voucher program. It provides a single point of contact for schools as well as students interested in taking advantage of school choice in Florida and guarantees that scholarships follow children when they transfer from one school to another.” Garnett, supra note 5, at 14. This supports the move from tax credits to vouchers—this tax credit system is successful not because it is a tax credit, but because it operates like a statewide voucher. 192 Pinto & Walker, supra note 91, at 85. 193 Paul Teske & Mark Schneider, What Research Can Tell Policymakers about School Choice, 20(4) JOURNAL OF POLICY ANALYSIS AND MANAGEMENT 609, 619 (2001) (cataloging studies and concluding that “choice programs demonstrate modest to moderate test score improvements for some, but not all, students who participate”). 194 U.S. Gov’t Accountability Off., supra note 99; see also Stephen J. Owens, Opinion: Vouchers prop up private schools with public money, Atlanta J.-Const. (Dec. 21, 2022), https://www.ajc.com/ education/get-schooled-blog/opinion-vouchers-prop-up-private-schools-with-public-money/JBHM UA4SPRHMDCFCS2OAZTB3K4/ [perma.cc/L2BC-QMDL] (“[P]rivate schools are not held to federal protections for services provided for students with disabilities or those learning English.”). 195 Teske & Schneider, supra note 193. 196 Patrick J. Wolf et al., Do Voucher Students Attain Higher Levels of Education? 20 (Annenburg Inst. at Brown Univ. Ed Working Paper No. 19-115, 2019), http://www.edworkingpapers.com/ai19- 115 [perma.cc/NW77-QRQC] (“Our findings contribute to a growing body of evaluation results indicating that private school voucher programs positively affect student educational attainment.”); Matthew M. Chingos & Daniel Kuehn, THE EFFECTS OF STATEWIDE PRIVATE SCHOOL CHOICE ON COLLEGE ENROLLMENT AND GRADUATION (Urban Inst. 2017), https://www.urban.org/sites/default/ files/publication/93471/2017_12_05_the_effects_of_statewide_private_school_choice_on_college _enrollment_and_graduation_finalized_2.pdf [perma.cc/TP2B-G3X9]. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 115 schools do not have the attendees or the resources that lead to them receiving donations. Inevitably, some students will remain at public schools, which will become more stratified as a result of school choice.197 For programs that provide such little benefit to educational systems for the poor, states sacrifice immense tax revenue and leave open opportunities to abuse the system. Tax credit programs thus frequently exacerbate the problems they were intended to repair, leading to more private funding for already wealthy private schools and leaving low-cost private schools and public schools to founder. It is unsurprising that tax credit programs are not massively successful— their goal was to increase school choice, but when they were initially developed, legislators believed it was unlawful for state money to support religious schools. Naturally, the roundabout system that resulted was inefficient. Had the law been what it is now—not imposing any barrier between public money and religious schools—tax credit programs probably would not have been legislators’ first choice to strengthen education. Regardless of whether the circumvention of understood laws was initially proper, we now have the opportunity to repair the inefficiencies of tax credit programs and effectively improve education for all–public or private, religious or secular, wealthy or poor. Because the Establishment Clause jurisprudence is now clear, instead of clinging to the vestigial approach that has failed to prove itself, states should pursue their goals of school choice through direct funding. IV. SOLUTIONS States that wish to fund private schools should therefore discard their tax credit systems in favor of direct funding that is more likely to improve education for those who need it. The most important change that a state should make with direct funding is to ensure that whatever aid it provides to private schools should benefit students in need, instead of subsidizing tuition payments that otherwise could have been paid privately. Because the benefits of tax breaks are more appealing to higher earners who owe more in taxes, subsidies should avoid the tax code entirely.198 The state should aid all students, not just the children of high earners. Merely disallowing scholarship organization donors from indicating the school or student to receive the donation would not solve this problem, however. Florida’s Tax Credit Scholarship Program does just this already—donors may not indicate a certain student or school to receive their donation, and the private organization managing the donations may not represent that the funds will be directed to a particular student or school.199 This change might be expected to 197 Kristie J.R. Phillips et al., School Choice & Social Stratification: How Intra-district Transfers Shift the Racial/ethnic and Economic Composition of Schools, 51 SOCIAL SCIENCE RESEARCH 30 (2014); Julia A. McWilliams, The Neighborhood School Stigma: School Choice, Stratification, and Shame, 15 POLICY FUTURES IN EDUCATION 221 (2017). 198 Coverdell education savings accounts, for example, allow investments to grow tax-free and be spent tax-free on educational expenses. But only families with money to invest stand to benefit from them. INTERNAL REVENUE SERV., U.S. DEP’T OF THE TREASURY, PUB. NO. 970 TAX BENEFITS FOR EDUCATION (2022), https://www.irs.gov/publications/p970#idm140550552112176 [perma.cc/8J2 M-WFXB]. 199 Fla. Stat. Ann. § 1002.395 (West 2023). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 116 destroy any incentive to donate, but Florida’s system is still robust; its Tax Credit Scholarship Program awarded $568 million in scholarships for the 2021–22 school year.200 However, it still suffers from a lack of regulation. This is because—even despite requiring (at least until July 2023201) that only certain children qualify for tax credit-funded scholarships202—it was not the government in charge of ensuring that only qualified students receive scholarships, but the private organizations that receive donations.203 The private organizations are also in charge of ensuring that the private schools submit proper financial records regarding their use of the funds.204 Like in other states, outsourcing the regulation of the tax credit system to private groups creates a gap in oversight that, at worst, invites abuse. If a state truly wants to ensure that public funds are directed to students who need them to attend private school, it should dispose of the intermediary layer of the private organization receiving donations.205 Even putting aside the possibility of abuse, Florida has no reason to redirect money that would otherwise be state tax revenue toward private schools. Nothing requires this extra complication. Florida can now just allocate some of its state tax revenue to support private schools, making budgeting simpler and more regular, and it would give more predictability to their state aid. Interestingly, Florida operates another scholarship program that does this. Florida’s Family Empowerment Scholarship for Educational Options, enacted in 2019, provides vouchers that can be used at private schools, but its funds come directly from the state’s budget, instead of from donations in exchange for tax credits.206 Florida expanded eligibility for this program in 2023, making these vouchers available to all Florida students regardless of their families’ income.207 While the Family Empowerment Scholarship funds are still managed by private organizations,208 the law directs those organizations to prioritize students with 200 Florida Tax Credit Scholarship Program, FLA. DEPT. OF EDUC., OFF. OF INDEP. EDUC. & PARENTAL CHOICE (2022), https://www.fldoe.org/core/fileparse.php/5606/urlt/FTC-Oct-2022-line. pdf [perma.cc/Z86U-H3UT]. 201 Fla. Stat. Ann. § 1002.395 (West) (amended 2023). Effective July 1, 2023, there are no income- based eligibility requirements for Florida’s Family Empowerment Scholarship Program. The law directs the private organizations distributing scholarships to give priority to students with low- income levels or who are in foster care. 2023 Fla. Sess. Law Serv. Ch. 2023-16 (West). 202 Ariz. Private Sch. Tuition Org., supra note 165; Fla. Stat. Ann. § 1002.395 (West). Only children who qualify for other government benefits, who are in foster care, or whose household income does not exceed 400 percent of the federal poverty level qualify for tax-credit funded scholarships. 203 Fla. Dept. of Educ., Off. of Indep. Educ. & Parental Choice, supra note 200, at 2. 204 Fla. Stat. Ann. § 1002.421 (West 2023). 205 Garnett, supra note 5, at 14 (noting that scholarship organizations are accompanied by a “risk that elite schools with ties to wealthier, more sophisticated, donors may capture a share of the available tax benefits that is disproportionate to the number of participants whom they serve.”). 206 “[Family Empowerment Scholarships] are government-funded and the payments come from the state of Florida to the Scholarship Funding Organization . . . [Florida Tax Credit] scholarships are privately funded. The funding for a student receiving [a Florida Tax Credit] scholarship comes from donations to the Scholarship Organization that serves their household.” Frequently Asked Questions about the Family Empowerment Scholarship for Educational Options (“FES-EO”) Program, AAA SCHOLARSHIP FOUNDATION, https://www.aaascholarships.org/wp-content/uploads/2022/03/AAA- FES-EO-FAQ-2021-22-rev20220302.pdf [perma.cc/5JWK-Z3VD]. 207 2023 Fla. Sess. Law Serv. Ch. 2023-16 (West). 208 Fla. Stat. Ann. § 1002.394 (West 2023). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 117 lower family incomes or who are in foster care.209 Yes, given that it is managed by private organizations, Florida’s voucher program may suffer from lack of oversight just like tax credits do. However, the voucher program is preferable over the tax credit nonetheless—the funding is at least acknowledged as coming from the state budget. Tax credit systems, on the other hand, claim to be privately funded when they are in fact state expenditures. Most importantly, the Florida voucher program is preferable over a tax credit system like Georgia’s, where scholarships disproportionately benefit the rich because donors can direct their donations to specific schools. Florida’s system at least makes funds available to all students; both tax credit scholarships and direct vouchers are to be distributed by the scholarship-funding organizations to any eligible student.210 This portability is very important to avoiding the concentration of public funds on well-resourced schools.211 Still, the Florida system could of course go further by increasing oversight by minimizing the role of private scholarship-funding organizations and actually having the state administer scholarships. It could also require directing funds toward students who actually need them to choose private school as opposed to making them uniformly available.212 But at the very least, it is preferable over a system that concentrates funds on already wealthy private schools. Arizona has also taken a large step away from convoluted tax credits and toward direct funding. It recently enacted a voucher program that makes all students eligible for a voucher in the amount of around $7,200 deposited in an Education Savings Account (“ESA”). That money can be used for private school tuition or homeschooling. The good news about Arizona’s program is that the money can be used at a school chosen by the student and her family, not by a donor looking to shore up a specific private school’s revenue with state money. The room for improvement lies in execution and administration: voucher recipients “tend to be relatively well-off,”213 which may be due to inadequate publicity of the vouchers or administrative hurdles to claiming them. Even when funding has nothing to do with a tax credit, the more administrative hurdles that a family faces in using it, the more likely it is to disproportionately benefit wealthier children with educated parents who have the time and energy to navigate the process.214 The vouchers are also available to students who are already attending private school.215 Since there is no economic eligibility requirement, these vouchers are likely assisting students who could have attended private school without them. The government is effectively covering part of private school tuition for anyone, even those who never 209 2023 Fla. Sess. Law Serv. Ch. 2023-16 (West). 210 Id. 211 See Garnett, supra note 5, at 14. 212 Florida seems to be moving in the opposite direction with regard to need-based eligibility. In 2023, it changed its statutes from mandatory income-based eligibility to directing scholarship- funding organizations to prioritize low-income applicants. 2023 Fla. Sess. Law Serv. Ch. 2023-16 (West). 213 Mervosh, supra note 166. 214 Garnett, supra note 5, at 16 (“While ESA programs provide maximum flexibility, even well- educated parents may find the record-keeping required as a condition of participation frustrating. Parents with less formal education may find it nearly impossible.”). 215 Mervosh, supra note 166. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 118 thought of attending public school. To school choice supporters focused exclusively on equity, this is a weakness of Arizona’s program—the money would be better spent on giving more help to poor students, so the voucher could actually cover the cost of tuition. It is not a problem, however, for the goals of school choice focused purely on economics or individual choice. An ideal system could identify private schools in need of money and fund them more, consistent with the general goals of a progressive tax code: focusing on those serving low-income populations, charging minimal tuition, and relying on donations to operate. Whether aid comes directly from the state or from donations incentivized by tax credits, the responsibility for identifying qualified schools should fall on the state, not private organizations that partner with schools. Deference to private organizations only results in wealthier schools benefiting the most from scholarships. If a private scholarship organization has the choice between partnering with a wealthy private school and a poor one, it will choose the wealthy school that has the resources and organization to generate donations.216 States should also minimize the hoops schools must jump through to receive aid. The more hoops there are, the more aid is disproportionately directed to schools with organizational resources, as opposed to schools actually in need.217 The goal of public aid to private schools is to make private education affordable to all, so it should be directed to struggling private schools, or to all students generally, but not merely to those sophisticated enough to jump through complex administrative hoops. Requiring schools to take administrative steps, like partnering with scholarship organizations and generating donations, has the opposite effect. It directs donations toward schools with resources and infrastructure as opposed to struggling ones. Making public aid “easier” to get by removing the administrative barriers will allow the schools that need it to actually benefit from it. Removing administrative barriers to receiving donations also would not increase the likelihood of abuse because the regulatory responsibility would be placed on states instead of on private organizations. Instead of requiring private schools to communicate with parents, fundraise, and partner with a private scholarship organization, none of which have anything to do with being deserving of public funds, states should require schools to demonstrate that they actually are deserving of public funds. Georgia’s system, for example, should be revised to include qualifications for scholarship recipients that actually reflect financial need or universally fund all. Moreover, the state, not private organizations seeking more donations, should ensure that public funds actually go to qualified schools and students. Under a system like that, there is no need to use tax credits, and it becomes clear that the reasons tax credits were desirable in the first place—that they obscure the state’s role in providing aid and reduce transparency—do not serve the goal of educational equity and may actually be counterproductive. 216 See, e.g., GOAL Program Results, GA. GOAL SCHOLARSHIP PROGRAM, INC. (July 31, 2023), https://www.goalscholarship.org/results/ [perma.cc/7E83-XFAK]; Financial Reports, ARIZONA CHRISTIAN SCHOOL TUITION ORGANIZATION, https://acsto.org/about/financial-reports [perma.cc/F S4J-ADLU] (publicizing amounts they have provided in scholarships). 217 Garnett, supra note 5, at 14 (noting that scholarship organizations are accompanied by a “risk that elite schools with ties to wealthier, more sophisticated, donors may capture a share of the available tax benefits that is disproportionate to the number of participants whom they serve”). about:blank about:blank 2023] THE WORST CHOICE FOR SCHOOL CHOICE 119 The Maine program challenged in Carson for excluding religious schools allocated direct aid based on where a student lives. Living in a district that does not operate its own public school qualifies a student for the program. Qualified students designate the school they plan to attend, which can be public or private, and their home district pays that school for their cost of attendance.218 One good quality of this program is the absence of an income requirement to benefit from it (as opposed to tax credits, which only benefit those who earn a certain amount). The program does not appeal to someone because it can refund what she pays the state in taxes; it is available simply because she lives in a district without a public school. There is also not the same potential for abuse as there is with the tax credit system—a student chooses a school and the state funds his attendance; there is no illusion that the public money should be used for scholarships but ends up serving the student in lieu of public school.219 The public money is supposed to benefit the specific student. One feature of Maine’s system is that it does not necessarily direct money toward the rich or poor, but, rather, is universal. A rich student and a poor student would both get private school paid for by the state as long as they live in a remote district. This is somewhat a feature of Maine’s program, which aims to ensure that every student in a state with many remote areas “shall be provided an opportunity to receive the benefits of a free public education,”220 as opposed to equalizing educational opportunities or increasing choice. V. ANTICIPATED OBJECTIONS AND OUR RESPONSES This section addresses two contemplated responses to the proposal to dispose of tax credits. The first is legal, namely, that states still may not fund religious schools because of federal or state law. The second is about policy: even if states are not required to use tax credits to fund public schools, they may still wish to. A. Legal Arguments The legal counterargument to the proposal to dispose of tuition tax credits is that the law still prohibits states from directly funding religious schools. One basis for this is the Establishment Clause, and another is state no-aid provisions. Neither source of law, however, soundly supports that the government may not fund religious schools but may allow tax credits to fund them. 1. Establishment Clause First, the Establishment Clause does not prohibit states from funding religious schools—indeed, sometimes, it mandates it. This must be true after Carson, in which the Supreme Court required Maine to fund religious schools because it did so for non-religious private schools.221 The funding at issue in Carson came directly from the state; “school administrative units,” which govern public 218 Carson v. Makin, 142 S. Ct. 1987, 1994 (2022). 219 This could be abused if a school told qualified students that they would receive something in exchange for choosing that school. That would be egregious but could be hard to detect. 220 Me. Rev. Stat. tit. 20-A, § 2. 221 Carson v. Makin, 142 S. Ct. 1987, 2002 (2022) (noting that extending Maine’s program to religious schools does not offend the Establishment Clause). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 120 education in Maine, directly paid private school tuition for students living in districts without public schools.222 The Court ruled that Maine could not exclude religious private schools from this program.223 If a student in a district without a public school chose to attend a religious private school, then, Carson requires the state to pay tuition directly to that school. The Court explicitly stated that this “neutral benefit program in which public funds flow to religious organizations through the independent choices of private benefit recipients does not offend the Establishment Clause.”224 The Supreme Court’s ruling on the meaning of a constitutional provision carries as much weight as the constitutional provision itself, regardless of popular reception of the decision.225 Carson’s interpretation certainly changed our understanding of the Establishment Clause—commentators have written that Carson “called [conditions on government funding] into serious question”226 and “dismantle[s] the Establishment Clause”227—but the fact is that the Establishment Clause does not prevent direct funding to religious schools. Nonetheless, there is an argument that Carson was just wrong and that the law may, in the future, revert back to a starker separation between Church and state. If that happens, maybe tax credits would survive under state constitutional provisions and Winn, but vouchers would not be permitted to directly fund religious schools. States may thus think that sticking with tax credits instead of direct vouchers is the conservative option that is less likely to face legal obstacles in the future. Carson, Trinity Lutheran, and Espinoza were indeed landmark decisions, but there is no evidence that the law will revert back. The Court viewed Carson itself merely as an extension of the earlier two cases228 as did many commentators.229 And, since Carson was decided in June 2022, there has been an explosion of reliance. States immediately began expanding their school choice programs. In 2023, Florida, Iowa, and Utah approved expansions to their school 222 Id. 223 Id. 224 Id. at 1997. 225 Cooper v. Aaron, 358 U.S. 1, 18 (1958) (“[T]he federal judiciary is supreme in the exposition of the law of the Constitution.”). 226 Ira C. Lupu & Robert W. Tuttle, Carson v. Makin and the Dwindling Twilight of the Establishment Clause, AMERICAN CONSTITUTION (June 23, 2022), https://www.acslaw.org/expertf orum/carson-v-makin-and-the-dwindling-twilight-of-the-establishment-clause/ [perma.cc/P2KE-B VW3]. 227 Marci A. Hamilton, The Supreme Court Further Dismantles the Establishment Clause, Empowers Religious Parents to Obtain Taxpayer Funds for Sectarian Schools, and Ignores the Rights of the Children in Carson v. Makin, VERDICT (June 22, 2022), https://verdict.justia.com/2 022/06/22/the-supreme-court-further-dismantles-the-establishment-clause-empowers-religious-par ents-to-obtain-taxpayer-funds-for-sectarian-schools-and-ignores-the-rights-of-the-children-in-cars on-v-makin [perma.cc/XC63-8BPB]. 228 Carson v. Makin, 142 S. Ct. 1987, 1997 (2022) (“The ‘unremarkable’ principles applied in Trinity Lutheran and Espinoza suffice to resolve this case.”). 229 Justin Driver, Three Hail Marys: Carson, Kennedy, and the Fractured Détente over Religion and Education, 136 HARV. L. REV. 208, 225, 233–34 (2022) (describing Carson as “consistent with the terms of détente that emerged at the beginning of the twenty-first century” as well as with Meyer v. Nebraska); Mark L. Rienzi, Religious Liberty and Judicial Deference, 98 NOTRE DAME L. REV. 337, 384 (2022). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 121 choice programs.230 That reliance is important, because if an Establishment Clause challenge ever came before the Court offering an opportunity to reverse Carson, it would weigh against overturning that precedent. Another reason that the law will not revert back is that the Court’s declaration is supported by history. At its conception, the Establishment Clause was meant to prevent the establishment of a national church and did not necessarily invalidate public aid to religious institutions.231 Some churches received state support when it was ratified.232 It is a recognized, defensible position that “the original Establishment Clause embraced no substantive conception of the proper relation of church and state, but merely reflected a determination that the issue be settled locally.”233 It is undeniable that the Establishment Clause’s history and original meaning remain “sharply contested,”234 and plumbing the complicated depths of that disagreement is outside the scope of this Article. The only point relevant here is that Carson was indeed a divided decision about which reasonable minds could disagree, but it was not completely without historical and jurisprudential support. Carson thus defeats the argument that the Establishment Clause strictly prohibits state aid to religious schools. Even accepting that Carson holds that the Establishment Clause permits direct aid, it may still be argued that direct aid must reflect private choice. Carson’s text characterizes Maine’s program only as “a neutral benefit program in which public funds flow to religious organizations through the independent choices of private benefit recipients.”235 Direct aid without tax credits can meet this standard, much like it does in Maine: a state can directly fund private schools that students eligible for scholarships choose to attend. Private choice of the student is still the only way that public money gets to a religious school, and there is no need for tax credits. No one really doubts that voucher aid—a fixed amount to every student in every school—is now constitutional. More fundamentally, reading Trinity Lutheran in conjunction with Carson demonstrates that private choice is no longer a requirement. The Trinity Lutheran majority held that Missouri could not exclude religious schools from eligibility for state grants for playground surfaces. Applications for grants came from the schools themselves and were scored on criteria “such as the poverty level of the population 230 J. David Goodman, A Well of Conservative Support for Public Schools in Rural Texas, N.Y. TIMES (Apr. 14, 2023), https://www.nytimes.com/2023/04/14/us/texas-school-vouchers.html [perm a.cc/U5DM-8WZB]; Bush, supra note 130, at 100 (“[B]ills to establish universal choice are moving in more than a dozen states, including Indiana, Ohio, New Hampshire, Texas, and Virginia. Oklahoma [leaders] are committed to creating a universal ESA program. . . . South Carolina [leaders] have committed to fighting to expand educational opportunity.”). 231 Vincent Phillip Muñoz, The Original Meaning of the Establishment Clause and the Impossibility of its Incorporation, 8 U. PA. J. CONST. L. 585, 590 (2006); Justin Driver, Three Hail Marys: Carson, Kennedy, and the Fractured Détente over Religion and Education, 136 HARV. L. REV. 208, 225, 233–34 (2022) (describing Carson's consistency with Meyer v. Nebraska). 232 John C. Jeffries Jr. & James E. Ryan, A Political History of the Establishment Clause, 100 MICH. L. REV. 279, 294 (2001). 233 Id. at 293. 234 Vincent Phillip Muñoz, The Original Meaning of the Establishment Clause and the Impossibility of its Incorporation, 8 U. Pa. J. Const. L. 585, 585–86 (2006). 235 Carson v. Makin, 142 S. Ct. 1987, 1997 (2022). https://www.nytimes.com/2023/04/14/us/texas-school-vouchers.html COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 122 in the surrounding area and the applicant’s plan to promote recycling.”236 Only Justice Sotomayor’s dissent acknowledged that Trinity Lutheran did not address the private choice requirement, but the result, requiring a state to allow a benefit to a religious institution based on its application alone, implies that no private choice is required.237 Accordingly, states may fund private schools in ways that do not reflect private choice. That would look like a funding program that allocates money on a basis other than per-student, and would allow states to allocate their funds available to private schools more equitably. As an example, a state could administer a program like Missouri’s that enables schools to apply to the state for funding. A poor private school that charges minimal tuition could demonstrate its need and receive supplemental funding. Nowhere in the process would a student or donor have to demonstrate her choice that public funds to which she is entitled be directed toward a religious school. Under Trinity Lutheran, such a system would not violate the Establishment Clause.238 In short, the current Supreme Court has clarified that the Establishment Clause does not prohibit direct aid. Carson was indeed a controversial decision, and the Establishment Clause’s history is contested, but the Supreme Court has not only permitted, but required the direct funding of religious institutions. That “interpretation . . . is the supreme law of the land.”239 Accepting this as the law will allow states that want to fund private schools—secular and parochial—to do so in a way that allows reasonable regulation and proper supervision, like all governmental programs ought to. 2. No-aid Clauses No-aid clauses purport to restrict interactions between the state and religion more than the Establishment Clause does.240 Therefore, even though the Establishment Clause permits direct funding of religious schools, it may still be argued that state no-aid provisions disallow it. No-aid clauses cannot be the reason 236 Trinity Lutheran Church of Columbia, Inc. v. Comer, 582 U.S. 449, 455 (2017). 237 Id. at 2029 n.2 (Sotomayor, J., dissenting) (citing Zelman v. Simmons-Harris, 536 U.S. 639, 649 (2002)) (“Because Missouri decides which Scrap Tire Program applicants receive state funding, this case does not implicate a line of decisions about indirect aid programs in which aid reaches religious institutions ‘only as a result of the genuine and independent choices of private individuals.’”). 238 There is also an argument that some tax credit systems in existence today do not reflect private choice. Recall Carson’s language that Maine’s system was “neutral benefit program in which public funds flow to religious organizations through the independent choices of private benefit recipients” Carson 142 S. Ct. at 1997. In a system like Georgia’s, where donors indicate the school to receive their donations, the benefit recipient does not determine where the money goes. Applicants must apply for tax credit-funded scholarships directly with Georgia private schools, as opposed to receiving the scholarship and then choosing where to direct it. Schools receive the money from the scholarship organizations and may allocate it to scholarships for students, unrestrained by wealth- based eligibility requirements. 239 Cooper v. Aaron, 358 U.S. 1, 18 (1958). 240 See, e.g., Carson v. Makin, 142 S. Ct. 1987, 1991 (2022) (“Maine’s decision to continue excluding religious schools from its tuition assistance program after Zelman thus promotes stricter separation of church and state than the Federal Constitution requires.”); Espinoza v. Mont. Dep’t of Revenue, 140 S. Ct. 2246, 2260 (2020) (“The Montana Supreme Court asserted that the no-aid provision serves Montana's interest in separating church and State ‘more fiercely’ than the Federal Constitution.”). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 123 that a state excludes religious schools from a benefit they would otherwise be qualified for, however. Carson held that enforcing a no-aid provision in this way violates Free Exercise.241 Carson does not necessarily render all no-aid provisions unconstitutional, however. The majority wrote “[a]s we held in Espinoza, a ‘State need not subsidize private education. But once a State decides to do so, it cannot disqualify some private schools solely because they are religious.’”242 In other words, enforcing a no-aid provision complies with Free Exercise as long as it does not end up requiring excluding religious schools from a benefit they would otherwise qualify for. A state that does not provide any aid to private schools may constitutionally enforce a no- aid provision. We neither support nor discourage such aid in this Article: we merely argue that once a state sets out to aid private schools, direct aid is a better system. The interpretation of a state’s no-aid provision is for that state to make. Federal courts do not have authority to say that state courts interpreted their state’s law incorrectly. Georgia courts may hold that Georgia’s no-aid provision permits tax credit scholarships but prohibits direct funding,243 and Montana courts may hold that Montana’s no-aid provision prohibits both.244 The state’s no-aid provision must submit to constitutional requirements; however, the Supremacy Clause requires that state courts “must not give effect to state laws that conflict with federal law.”245 No matter how a state interprets its no-aid clause, if it requires excluding religious schools from a benefit they would otherwise qualify for if they were secular, then the no-aid clause is unconstitutional as applied.246 In short, if a state’s no-aid provision prevents it from directly funding religious schools, it is not inherently unconstitutional, so long as the state responds to that interpretation by not funding private education at all.247 If that state provides aid to private schools, however, and relies on its no-aid provision to deny that aid to religious private schools, the no-aid provision cannot stand under Free Exercise. The argument that no-aid provisions prevent direct funding can only be used by states that do not provide aid to private schools at all, in which case, the no-aid provision does not do much work, at least in the area of education. As soon as a state relies on its no-aid provision to actually single out religious schools, however, the provision is unconstitutional and cannot be enforced. The argument that no-aid 241 See Carson, 142 S. Ct. at 2002. 242 Id. at 2200 (citing Espinoza, 140 S. Ct. at 2261). 243 See Gaddy v. Ga. Dep’t of Revenue, 802 S.E.2d 225 (2017). 244 Before Espinoza, this was the Montana Supreme Court’s interpretation of Montana’s no-aid clause. Espinoza, 140 S. Ct. at 2251 (“[T]he Montana Supreme Court struck down the program [of tax-credit funded scholarships, relying] on the ‘no-aid’ provision of the State Constitution. . . .”). 245 Armstrong v. Exceptional Child Ctr., Inc., 575 U.S. 320, 324 (2015) (quoted by Espinoza, 140 S. Ct. at 2262). 246 See Carson, 142 S. Ct. at 1998 (“The State pays tuition for certain students at private schools— so long as the schools are not religious. That is discrimination against religion.”); Espinoza, 140 S. Ct. at 2260. 247 Note that if a state already operates a system of aid for private schools but excludes religious ones, its courts must order the inclusion of religious schools if a Free Exercise challenge is brought. The legislature can terminate the entire program, but state courts cannot respond to a challenge of a no-aid clause by ending all private school aid. See Espinoza, 140 S. Ct at 2246. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 124 provisions prevent states from directly funding religious schools is thus valid only in narrow circumstances: where the state funds no private schools at all.248 States were not unreasonable in believing that there were legal barriers to their direct funding of religious schools. In 1995, Justice Thomas famously described the Supreme Court’s Establishment Clause jurisprudence as being “in hopeless disarray.”249 The Court has recently taken some opportunities to clarify it, though, and it is clear from those decisions that the Establishment Clause does not forbid direct funding, and that no-aid provisions cannot justify funding only non- religious private schools. 3. Other State Constitutional Issues State constitutions may pose other obstacles to direct vouchers. In some cases, state provisions that are perceived as preventing direct vouchers may not do so in fact. This is what happened with state no-aid clauses—governments believed no-aid clauses forbade them from subsidizing private religious schools, and it turns out that no-aid clauses, when used that way, are actually unconstitutional. The Gratuities Clause contained in Georgia’s constitution is another example.250 This clause prohibits the state from “grant[ing] any donation or gratuity,”251 meaning it may not allow the free use of state labor or property when the state receives no substantial benefit.252 In other words, the clause prevents the state government from giving gifts.253 It has been suggested that direct vouchers in Georgia would violate this clause.254 Because providing public schools that do not charge tuition does not violate the Gratuities Clause, however, it is doubtful that vouchers would. Vouchers are not a gift more than they are a public benefit provided by the government well within its power to govern education. If the state wishes to provide its school-age residents with free education, it can do so through public schools or by abating the cost of private schools; neither of those benefits would be considered a gratuity. This is our first reply to objections that other state laws forbid state governments from providing direct vouchers: ensure that the provision itself, not just its longstanding perception, actually prohibits vouchers. We do not intend this paper to be a deep exploration of state law, however, and it may well be that some state provisions genuinely prohibit vouchers but allow tax credits. We urge that 248 This is close, but not quite equivalent to, saying that no-aid provisions are worthless in light of Free Exercise. It is still possible to follow and enforce a no-aid provision in a constitutional way, such as if a state with a no-aid provision relied on it as the reason not to provide aid to any private schools (knowing both that it could not aid religious schools under its no-aid provision and that it could not aid only non-religious private schools under Free Exercise). In most cases, however, it seems that no-aid provisions which are followed have the effect of excluding religious schools, rendering them unconstitutional under Carson. 249 Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 861 (1995) (Thomas, J., concurring). 250 We would like to thank Professor Fred Smith Jr., Charles Howard Candler Professor of Law at Emory University, for bringing our attention to Georgia’s Gratuities Clause. 251 Ga. Const. of 1983. art. III, § VI ¶ VI. 252 See Garden Club of Ga., Inc. v. Shackelford, 266 Ga. 24, 24 (1995); 1993 Op. Att’y Gen. U93- 14. 253 See Garden, 266 Ga. 24 at 24; McCook v. Long, 193 Ga. 299, 302–03 (Ga. 1942). 254 See Gaddy v. Ga. Dep’t of Revenue, 802 S.E.2d 225, 231, 234 n.14 (Ga. 2017). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 125 such provisions are simply bad law—they purport to allow school choice, but through a counterproductive method to the exclusion of a good one. Accordingly, if a state already seeks to subsidize school choice but feels it cannot do so directly because of a state law, legislators should consider creating an exception to whatever state law stands in the way. If the state law prohibits direct funding but allows tax credits, for example, it allows school choice through what we argue is the worst means. The legislature should either remove the obstacle and allow effective administration of school choice or decide against school choice entirely. Laws that permit only tax credits and prohibit direct vouchers are the worst of both worlds; they narrowly restrict the administration of school choice to a poorly regulated and flawed method. 4. Reforming, Instead of Replacing, Tax Credits Finally, we understand that legislators may still wish to honor the state laws that prevent direct funding but allow tax credits. We have been advised that tax credits are advantageous politically—they constitute social spending favored by liberals, but conservatives can deny that they are actually public money. If state legislators wish to maintain the tax credit as their vehicle for school choice, for political reasons or otherwise, reform to the tax credit can still address its major issues. The tax credit form is not the heart of our objection to tax credits—the main problems are actually the inequitable use and distribution and the lack of regulation. These problems can be addressed while maintaining the tax credit structure. Inequitable distribution could be addressed by forbidding the designation of a specific school or student on a donation. Florida, for example, does not allow donations to be designated to specific schools, which is a step in the right direction in terms of equity. Inequitable use requires more reform—it could be addressed by making the tax credit refundable, meaning that donors would be refunded by the state in excess of their tax liability. Tax filers with no state tax liability would receive a check from the state for the amount of their contribution. Families in lower tax brackets, then, would have just as much incentive to contribute as affluent families do, so the benefit is not concentrated on the wealthy.255 The issue that remains is that benefitting from the tax credit still requires parting with a sum of money for part of the year: the donor would have to donate to the scholarship organization and wait to receive their refundable tax credit at tax time the following year.256 This is obviously much easier for the wealthy, who have more disposable income. To fully 255 A refundable tax credit is still an obstacle more easily surmounted by the wealthy than by the poor. The wealthy already have a reason to be filing and paying attention to state taxes. People who know they will have no state tax liability do not. Georgians who make less than $8,100 in 2022, for example, are not required to file a state income tax return at all; see Filing Requirements, Ga. Dep’t of Revenue, https://dor.georgia.gov/filing-requirements [perma.cc/4RGH-GQMM]. Those who know they do not need to worry about state taxes at all will probably not go hunting down a tax credit, but those who are already filing will have more exposure to it. Thus, the wealthy remain in a better position than the poor to take advantage of even refundable tax credits. 256 See Andrew T. Hayashi & Justin J. Hopkins, Charitable Tax Deduction and Civic Engagement, 2023 U. ILL. L. REV., 1179, 1220 (2023) (“[Earned Income Tax Credit] recipients often have significant debt that accumulates during the winter holidays that they only repay after receiving their tax refunds, making them vulnerable to delays or garnishment of those refunds.”). COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 126 address the inequitable use of tax credits, then, the refund would have to be issued simultaneously with the donation. That way, recipients of the credit would not be required to part with their money for months, expanding the option to poorer families. Of course, if the tax credit is issued simultaneously with the donation, it looks more like a voucher than a tax credit. If legislators would prefer to call it a tax credit for political reasons, however, we have no objection—the issues are with the inequitable distribution and the lack of regulation, not the label. Addressing inequitable distribution of tax credits may also require means testing tax credit recipients, so that the tax credits are actually distributed more generously to those who need them to attend private school than to those who can afford private school on their own. This would be a step in the right direction, but it would require much more scrutiny of how schools and scholarship organizations distribute funds. This could be done with either tax credits or direct funding, but we expect it will be easier with direct funding, which allows more accountability and does not share tax credits’ long history of being insulated from the government. Even with this reform, tax credits will begin to look more like vouchers because the state will have more control over how they are distributed.257 Retaining the tax credit structure would also require reform in the area of regulation. As with the issues described above, there is no reason that regulation must come from outside of the states’ tax functions. If the state agency that currently administers the tax credit can adequately increase oversight of how the funds are used and awarded, then it should implement that oversight, by all means. Whether the agency that regulates tax credits is new or old is not the problem—the lack of regulation is. The long tradition of tuition tax credits being characterized as completely separate from the government (due to the need to validate their use by religious schools), is likely to make adequate regulation an uphill battle. As we have discussed, tax credits also seem a cumbersome, overly complicated mechanism for distributing public benefits, hindering accountability by hiding social programs behind a complex tax code.258 While it has been argued that this is just the price to pay for tax expenditures that support personal choice, then we aver that school choice is not the place for tax expenditures. Similarly, some may view the lack of regulation as a positive feature of tax credits; tax credits are less likely than direct funding to expose private schools to state regulations such as non-discrimination requirements. However, since tax credits are effectively public subsidies of private education, that kind of regulation is necessary and warranted as part of the nature of public funding.259 Nonetheless, the critic might continue, if a state agency can overcome these regulatory obstacles, the required regulation could be applied while maintaining the 257 Thanks to Professor Hillel Levin for bringing our attention to means-tested tax credits. 258 Thanks to Professor Alex Zhang for his expertise on the tax debate. For more on this, see Alex Zhang, Pandemics, Paid Sick Leaves, and Tax Institutions, 52 LOY. U. CHI. L.J., 383, 399–401 (2021) (“Scholars advocating a comprehensive tax base have contended that distributing government resources in the form of a tax concession is inefficient, poses difficulties in administrability, and shields the government from accountability by hiding spending in the tax code.”). 259 Tax credits’ serving as a shield against regulation is analogous to how they used to serve as a shield against Establishment Clause issues. Tax credits were separate enough from government funds to comply with the Establishment Clause. 2023] THE WORST CHOICE FOR SCHOOL CHOICE 127 form of a tax credit.260 Likewise, the mere form of a voucher alone will not solve the regulation problem; the vouchers we propose must be accompanied by regulation and accountability. We contend that vouchers, since they are not shadowed by a history of necessary separation from the government, are more readily regulated, but failure to regulate them would come with many of the problems we see with tax credits today.261 Migrating the tax credit system over to vouchers will not wholly solve the problems of tax credits; the vouchers must be done right. In short, states can choose to start fresh with direct funding, or they can reform their tax credit programs to address the major issues discussed here. Different states will have different views on which option is more complicated or politically advantageous. Adequate reform may make the tax credit resemble a voucher, but if the issues of tax credits are addressed, that label is a non-issue. B. Policy Arguments The sections above make plain that neither the Establishment Clause nor no-aid provisions need prevent states from directly funding private schools. This section addresses the question that remains open: even though a state could fund schools directly, what if it chooses to use tax credits instead? We cannot argue that doing so would be unconstitutional. In a state like Georgia, for example, where the state supreme court has held that the no-aid provision allows tax credit scholarships at religious schools, the practice is certainly constitutional. We have laid out above why interpreting a no-aid provision as prohibiting direct funding but allowing tax credits is overly formalistic. Nonetheless, states continue to do so, and federal courts may not correct their interpretations of state law.262 This Article urges that states should not use tax credits merely as an avenue around a state no-aid provision, however. Tax credits are no longer needed to circumvent the Establishment Clause, and there is no good policy reason to use them to circumvent no-aid clauses (which are unconstitutional anyway if a state provides direct aid to private schools). Once those justifications are stripped away, no policy reason to choose tax credits remains, other than a desire to subsidize private education for the wealthy. Tax credits do not well fund schools serving poorer populations and disorganized communities. Instead, they fund almost exclusively wealthy private schools—which could be a goal, but one we suspect most do not support. With direct funding as an available and stronger alternative, the downsides of the tax credit mechanism are too weighty to justify continued use.263 260 Thanks to Professors Samuel Brunson and Hillel Levin, as well as the participants in the joint faculty workshop between the law schools of Emory University and the University of Georgia for emphasizing the possibility of adequate regulation of tax credits. 261 Mervosh, supra note 166 (expressing skepticism regarding voucher regulation and accountability). 262 Nor would they probably because Winn held that tax credits do not implicate the Establishment Clause. See Arizona Christian Sch. Tuition Org. v. Winn, 563 U.S. 125, 144–145 (2011). 263 Some may argue that the downsides of tax credits—worsening educational inequality, in particular—do not outweigh the need to preserve parental choice. Doing away with tax credits does not destroy parental choice, however. Direct funding is an available, and better, alternative. COLUMBIA JOURNAL OF TAX LAW [Vol: 15:1 128 To illustrate this point, consider three scenarios created by school choice in its different forms. First, consider a state with no school choice funding at all. People who can and want to pay for private school go to private school, and everyone else goes to public school. If the public school needs improvement, its students and their families will advocate for it. In a state that permits tax credits but not vouchers, however, rich families in the struggling public school are actually incentivized to leave rather than advocate for better public schools. They can go to private school and have part of their tuition covered by the government. The public school, meanwhile, will be left struggling with fewer advocates and resources; poorer students furthermore do not have an option to leave. Consider, then, the alternative where a state uses vouchers and not tax credits. People at the struggling public school, regardless of their wealth, can stay at the public school or decide to use a standard amount of government money to abate the costs of private school. Going to private school does not require drumming up donations, private schools are not incentivized to recruit the rich, and the state is covering the costs of private school equally for each student. The second option, one with tax credits but without vouchers, is the worst of both worlds—it leaves public schools struggling and subsidizes private school for the rich to the exclusion of the poor. That scenario will exacerbate educational stratification. If state law prohibits vouchers but allows tax credits, it is just bad law unless tax credits are seismically reformed to the point where they resemble vouchers. The only policy reason that supports using tax credits in their current form is a desire to subsidize private school tuition for the rich. If this is what state legislators truly wish to achieve, they should directly say so.264 Otherwise, they are using a counterproductive system to support school choice for all. Another important question for states is whether they should fund private schools at all, but that is outside the scope of this Article. VI. CONCLUSION This article is easily summarized in four crisp points. ● Federal Establishment Clause jurisprudence now clearly permits (and sometimes mandates) direct aid to private religious schools along the same lines as any private school. ● Tuition tax credits are an unsupervised, indirect governmental aid program for private schools that only aids schools that have parents who pay Considering whether the downsides of tax credits are outweighed by the need for parental choice is therefore irrelevant; there is a way to preserve parental choice without the downsides—direct funding. 264 See, e.g., Matt Barnum & Alicia A. Caldwell, Vouching Helping Families Already in Private School, Early Data Show, WALL ST. J. (Dec. 3, 2023), https://www.wsj.com/us-news/education/vo uchers-helping-families-already-in-private-school-early-data-show-47ced812 [perma.cc/N2DP-RJ 7M]. This is, of course, unlikely to be politically wise. Facially neutral policies that actually favor the wealthy are likely to see even more limelight in the near future as cases concerning legacy admissions make their way to the Supreme Court. In the 2023 case invalidating affirmative action, Justice Gorsuch wrote that policies favoring children of donors, alumni, and faculty “undoubtedly benefit white and wealthy applicants the most.” Students for Fair Admission v. Harvard, 600 U.S. 181, 301 (2023). 2023] THE WORST CHOICE FOR SCHOOL CHOICE 129 significant state income tax. Schools that appeal to poorer students and communities do not get enough aid. ● By design, tuition tax credits lack sufficient governmental supervision and are both fraud-prone and ineffectual. Direct governmental aid comes with a regulatory framework that ensures that funds are properly used. ● States that want to subsidize private schools should do so through direct aid to schools. The path for financial aid to private religious schools has been a long and crooked one. It is time to straighten out the sidewalk to the front door of private school aid.