June 22nd, 2023 25 Minute Read Report by Nicole Stelle Garnett, Michael Q. McShane

Implementing K–12 Education Savings Accounts

Executive Summary

This is a pivotal moment for the parental-choice movement in the United States. Over the last two years, seven states have enacted new universal education savings account (ESA) programs or have expanded existing programs to universal or near-universal eligibility. Parental-choice advocates, who have labored for decades to make incremental inroads toward achieving greater educational freedom for families, have every reason to celebrate. But advocates, reformers, and regulators must not lose sight of the fact that the hard, but essential, implementation efforts that follow enactment will determine whether these programs achieve their transformational potential.

Implementing ESAs poses particular challenges because they are more than simply voucher or tax-credit programs. They are expansive parental-choice programs that provide students with public resources that they can use for a wide array of educational expenses in addition to private school tuition. As new ESA programs come online, thankfully, many people are intently focused on implementation. This report examines the challenges of implementing ESAs and makes suggestions about how to overcome implantation pitfalls. Specifically, we focus on five key priorities. First, parents must be informed about their options. Second, schools and other providers must prepare for, and respond effectively to, the opportunities and challenges provided by ESAs. Third, regulators must thoughtfully and deliberately address the need for effective program regulations. Fourth, program administrators must establish effective operational policies and procedures. Fifth, states and public-interest legal advocates must prepare for legal challenges.


“The work of destruction is quick, easy, and exhilarating; the work of creation slow, laborious, and dull.”

Sir Roger Scruton, How to Be a Conservative[1]

In 1973, Jeffrey Pressman and Aaron Wildavsky published one of the great works of modern political science. Titled Implementation, it was nominally a study of a federal port expansion project in Oakland, California.[2] But it became so much more. The pessimistic, unwieldy subtitle of the book tells the tale: How Great Expectations in Washington Are Dashed in Oakland; or, Why It’s Amazing that Federal Programs Work at All, This Being a Saga of the Economic Development Administration as Told by Two Sympathetic Observers Who Seek to Build Morals on a Foundation of Ruined Hopes.

This report reflects on the implementation of education savings account (ESA) programs. ESAs, which provide students with public resources that families can use for qualified education expenses to fit the educational needs of children (including private school tuition but also much more), are a definitive move beyond school choice to parental choice.

In this context, the slow, laborious, and dull work of implementation is critical. Implementation is challenging, unpredictable, and often ignored by policy advocates once legislation is passed. In some cases, regulators who are unenthusiastic or outright hostile to the ESA program may attempt to undermine its effectiveness through the rulemaking procedures, as has happened with some existing private school-choice programs.[3] In other cases, well-meaning government officials and program administrators may struggle to overcome flaws or bungle important tasks, such as paying providers on time, as has occurred in the rollout of Arizona’s ESA.[4]

Already this year, four states (Arkansas, Florida, Iowa, and Utah) have either created new universal ESAs, in which all K–12 students are eligible, or have expanded existing ESA programs to near-universal levels.[5] In May 2023, South Carolina adopted a means-tested ESA that will ultimately expand to serve students from families earning less than 400% of the federal poverty level.[6] With these new programs, there are seven nearly universal ESA programs (West Virginia’s universal ESA program launched in 2022, and Arizona expanded its existing ESA program to universal eligibility in 2022) and millions of eligible students. Five other states have ESAs but with more limited eligibility.[7]

Attention to the challenges of implementing these programs is critical because universal ESAs are new, uncharted territory. Universal ESAs are distinct from earlier, more modest, private school-choice devices such as vouchers and tax-credit scholarship programs in two ways.

First, while all previous programs—including a handful of ESA programs—limited eligibility in some ways (for example, limiting participation to low-income students and students with disabilities), all but South Carolina will eventually extend eligibility for participation to all or most students in a state. However, in Iowa[8] and Arkansas,[9] eligibility is phased in over time; and in Utah[10] and Florida[11] low- and moderate-income students are prioritized if there is excess demand for participation.

Second, while voucher and scholarship tax programs fund scholarships to enable students to attend private schools, ESA programs allow students to use public funds for a range of educational expenses other than private school tuition.

It has been exhilarating to watch the decades of advocacy for school choice finally reach the legislative mountaintop, but now the real work is getting started. After programs are created legislatively, authority moves to the state government agencies charged with administering the program. A rulemaking process puts meat on the bones of legislative text and clarifies what programmatic details will mean in practice. Regulations are drafted, and from those regulations flow policies and procedures that determine the day-to-day operation of the program. While it is important to get legislative text right and to win political battles in state legislatures, the rulemaking and policy-implementation phase is equally, if not more, essential to developing quality programs. Also important is the reaction of the schools and other providers participating in the programs.

Although heartened that regulators and advocates clearly understand the need to attend to implementation, we are concerned that the current conversation about implementing ESA programs is in danger of missing the forest for the trees. Many observers, policymakers, and critics are much more focused on “accountability” than functionality. Some have raised concerns about questionable spending in ESA programs, with parents purchasing items with a tenuous connection to an educational program.[12] However, that issue and the bad press that it brings are minor, compared to the issues that will crop up if ESA program implementation founders.

The scenario that we fear is not one in which 100,000 families participate happily in an ESA program and 100 make ridiculous or inappropriate expenditures. The other 99,900 families will provide more than enough counterweight in the media and in any state capital that threatens the program. Instead, we fear a program in which 100,000 families want to participate but cannot log in to the payment platform, or cannot track their expenditures, or cannot promptly pay the educational providers helping their children. Those families would form a formidable political bloc of unhappy people and an existential risk for an ESA program. That situation is also entirely avoidable.

Our arguments and recommendations about ESA implementation are based on interviews conducted in spring 2023 with program administrators, parents and parent advocates, researchers, and policy advocates. We kept our interviewees anonymous and have generalized their comments in gratitude for their candor in sharing both the successes and the struggles of ESA programs.

We offer five areas of interest that policymakers, advocates, educators, and philanthropists should think about when implementing ESA programs. We argue that ESA programs cannot succeed unless first, parents are informed and involved; second, schools and other providers prepare for and respond to the opportunities and challenges of ESAs; third, states proceed thoughtfully and deliberately in adopting program regulations; fourth, program administrators establish effective operational policies and procedures; and fifth, states and public-interest legal advocates prepare for legal challenges. We’ll discuss each briefly in turn.

Inform and Engage Parents

“If you build it, they will come” may be a memorable movie quotation, but it is a poor substitute for parental-choice policy planning. Parents are not only busy people, but they often are unaware of the publicly funded educational options available to their children. Particularly when eligibility calculations are complicated and programs are limited, it is very easy for parents to think that attempting to enroll is not worth the effort. Parents also need access to prompt customer service. When navigating new payment platforms or enrolling their children in ESA programs, they will inevitably have questions (and confront technological glitches and other roadblocks). Parents need good information, and they need to have their questions answered promptly.

Several polls have documented this accessibility problem. A 2017 survey of Indiana parents found that 39% of traditional public school parents and 36% of charter school parents had never heard of the state’s private school-choice programs. Looking specifically at low-income traditional public school parents (the population most likely to be eligible for the program), the percentage rose to 51%.[13] EdChoice’s 2022 Schooling in America survey found that 17% of parents had never heard of the term “school choice” in general and 31% had never heard of ESAs in particular.[14] A 2023 EdChoice poll found that parents are frequently confused as to whether their state has an ESA program; just as many parents in states that do not have programs think that they have one as parents in states with a program thinking that they do not.[15]

Luckily, there are steps that states can take to encourage and facilitate parental participation as they implement ESA programs. Based on conversations with parent advocates, we’d like to highlight three.

Handbooks, Webinars, and Training

A common adage tells us to never assume malice when ignorance is a sufficient explanation. This applies to parental participation in ESA programs. Parent advocates repeatedly told us that the reason some parents make purchases that are not permitted by the ESA program or that are on the border of inappropriate is that no one has clearly explained to them what is and is not allowed. As we will discuss below, common words in statutes and regulations, including “curriculum,” “devices,” and “aids” can encompass all manner of things. Defining these vague terms on a website is a necessary, but insufficient, step in informing parents about their choices and limits.

Parents need specific training on which expenditures are, and are not, permissible, how to use whatever payment platform the state has set up, and how they are supposed to choose and interact with providers. To date, states have published handbooks to explain how parents are supposed to navigate the program, but the handbooks are large and unwieldy. Parents need a more user-friendly solution. In addition to streamlining parent handbooks, training sessions specifically for parents, in person and online, augmented by a curated set of easy-to-understand webinars could be the solution. These aids ought to be easy to find, available on YouTube and other social media platforms, and should provide step-by-step guidance for what parents are expected to do, what they are allowed to do, and what they are not allowed to do. If aids are available, the likelihood of misspending will decrease significantly.

Beyond simply navigating the program, parents must be informed about the various options enabled by the programs. Frankly, families that get as far as asking “Is this permitted?” are way ahead of the game. Many parents don’t even know what some of the things that ESAs are hoping to encourage are (including microschools, learning pods, unbundled coursework, and hybrid homeschooling). States or private ESA administrators should develop clear, accessible, and short videos explaining various options for parents.

Parent Navigators

Navigating private school-choice programs can be challenging, frustrating, and isolating for families. States should take steps to make that process easier. Enter parent navigators.[16]

Organizations like Families Empowered[17] and EdNavigators[18] meet with parents, explain their options, and hold their hands as they apply to schools. They could expand their brief to cover educating families on options that are available in the ESA program, what they can and cannot spend their money on, how to use the payment platform, and what other parents around them are doing and how well that is working. It is important that these navigators respect their role as parent supporters and do not become overly prescriptive in their advice. There should be a diverse set of navigators, just as there should be a diverse set of school options.

Such programs could be funded by philanthropy, by direct state appropriation, or as an allowable use for ESA dollars. Concierge-style services have cropped up across numerous industries to ease the friction of making large and important purchases. Similarly, education navigators can help parents choose the educational providers and products that will work best for their children.

Predictability over Perfection

While not a specific policy or program, parents and parent advocates told us that predictability is more important than perfection. ESA programs are not going to launch in their perfect, final form—if such a thing even exists. Adjustments will need to be made and processes will need to be improved, but parents will bear the brunt of these changes. If states choose a new provider for the payment platform, parents will have to learn to navigate a whole new system. That system might be better—but states must be clear about the transition costs.

This underscores a point that we make later in this report: it is essential to get the request for proposal (RFP) process for key vendors right. Getting thousands or tens of thousands of parents set up in the system is a massive undertaking—one that states are not going to want to do again and again. There has been tremendous difficulty, for example, in managing the debit-card system that Arizona’s ESA program set up for participating parents. Some parents have been issued debit cards, others have been denied debit cards, and some are worried that their debit cards are going to be taken away.[19] All that uncertainty undermines confidence in and support of the program.

Schools and Providers Must Be Intentional and Self-Reflective

Some of us remember the great craft beer boom of the 2010s. Over the course of a few years, hundreds, and then thousands, of new microbreweries cropped up all over the country. It seemed as though some cities saw a new brewery open each week. According to the Brewers Association, between 2006 and 2020, the number of U.S. breweries grew from 1,460 to 9,025.[20]

Why did this happen? As Derek Thompson of The Atlantic explained back in 2018, part of the story is the evolving tastes of Americans, who seemingly got tired of the same beer options that they had. But regulatory reform played a role as well. The federal government liberalized home-brewing laws in 1978, and states changed laws to allow taprooms to sell directly to consumers, creating space in the market for small upstarts.[21]

Today, the craft beer market, while still growing, has developed a healthy churn. In 2021, for example, while 550 new breweries opened, 200 closed.[22] This is a healthy market winnowing out lower-quality entities and moving more customers toward better options.

Something similar can happen when states scale up ESA programs. There is pent-up demand from frustrated teachers, educationally inclined tech entrepreneurs, and proactive families, andhopefully—talented educators are going to try their hand at running schools. If the system is allowed to work, it will, over time, promote better options and eliminate worse ones.

We see this in states with thriving charter sectors. Education researcher Matthew Ladner has written for years about the competition that Arizona’s large charter sector places on schools. Critics decry the state as the “Wild West” because it authorizes so many charter schools and grants 15-year charters for operation. But if Arizona is the Wild West, Ladner argues, Arizona parents deliver “frontier justice.”[23] In reviewing data on Arizona’s charter sector, he found that charter schools that close are open for only four years, on average, and have only 62 students in their final year of operation. The unsuccessful schools, in other words, shut down quickly and students enroll in better options.

Potential schools or providers thinking about participating in an ESA program need to consider four crucial points. First, existing private and religious schools and other education service providers need to acknowledge the growth of ESA programs and get prepared. Second, participating private schools must resolve to use the new resources to welcome new students and to improve academically, as well as to think about ways that they can customize their offerings to take advantage of the flexibilities of ESA funding. Third, in recognition of inevitable efforts by ESA skeptics to limit their scope and effectiveness, schools must be ready to protect their autonomy. Fourth, ESAs aim to create a competitive education marketplace. Schools and other education providers must therefore provide parents with price and service transparency.

How Catholic and Faith-Based Schools Can Prepare

We are both staunch Catholic-school supporters. Private school-choice programs like ESAs have the potential to answer the prayers of generations of Catholic-school leaders who have watched schools close in response to changing demographics.[24] In 1960, 5 million students attended more than 12,000 Catholic schools. A half-century later, that number dropped to 1.9 million students in 6,500 schools.[25]

Like the man who met Saint Peter after drowning in floodwaters, having refused help from a truck, a boat, and a helicopter because he was a man of faith and trusted in God to save him, school leaders need to take assistance when it is offered. In our conversations, we learned that some private—and especially faith-based—schools are wary of participating in ESAs because of the specter of potential regulations that might come with public funding. We understand those concerns, but we believe that they can and should be addressed by legally protecting the autonomy of participating schools. Catholic schools and many other faith-based schools have argued on equality grounds for public funding for more than 150 years. Finally, these arguments are bearing fruit.

Moreover, these programs will inject competition into the U.S. educational landscape. Many other schools and educational providers will take advantage of the resources provided in ESA programs, hopefully incentivizing the creation of new schools and new schooling models. Resistance to participating will hamstring existing schools’ ability to serve children, whose options will continue to expand.

Many private, faith-based schools are accustomed to thinking that their problems are primarily financial and that parental choice is the answer to all their prayers. Experience in states with existing private school-choice programs provides abundant evidence that this simply is not true.[26] Parental choice is good education policy but is not a panacea. Participating in choice programs, schools in many states have already learned, is (or ought to be) hard work. ESAs will be no exception.

Schools must resist the temptation to simply use ESA funds to shore up shaky finances by filling empty seats. To participate successfully, schools must recruit new students—which necessitates telling parents a compelling story about why they are the best option for their children. When these students arrive, they will need to be acculturated to the school’s expectations and culture; many will require substantial academic remediation. Schools must, on that latter point, be clear-eyed about their own limitations and resolve to use ESA funds to improve academically.

Over time, successful schools should consider expanding, and successful school operators should consider opening new schools and forming new school networks. In some cases, especially in the faith-based sector, expansion may require rethinking existing school-governance models and delegating operational authority to private school-management organizations (akin to charter-management organizations).

Consider Hybrid or à la Carte Options

Since January 2021, EdChoice’s monthly polling has found that 40%–50% of school parents prefer some kind of hybrid schedule for their children, with students attending formal classes for some part of the week and learning from home for the rest.[27] This trend has continued even after schools reopened and the Covid-19 pandemic faded away. Utah’s new ESA program provides part-time public school students with a fund worth half the ESA amount ($4,000) to customize their education.[28] Private schools should consider whether to offer part-time public school students the option of taking classes with these funds. The same is true for private schools in other states, where “full-time” ESA students may wish to divide the use of their resources between attending brick-and-mortar private schools and other educational options.

This represents another opportunity for enterprising schools to offer differentiated educational programs for different families in their communities. Prestonwood Christian Academy, located in the Dallas suburbs, is a great example of a school that has already started differentiating its instructional program. It has two in-person campuses, offers hybrid enrollment, and has an online option. Bill Wendl, principal of PCA hybrid (as the hybrid program is known), put it this way:

People are hungry for a Christian education primarily because [of] what’s going [on] in our world [and] in our culture. ... They’re [also] looking at new trends. They’re looking at new ways to do school. ... They also like flexibility.[29

Some students might participate in a traditional, 8 am–3 pm, five-day-a-week option. Others might attend for some days and not others; some coursework might be placed online for students to study from their own home. There is clearly a demand for this kind of school organization. Enterprising schools will take advantage of the demand.

Of course, embracing hybrid schooling models is not cost-free for private schools. It will be harder to build community and reinforce school culture when some students are not present throughout the week. Part-time students may see themselves as consumers rather than students. Schools, quite reasonably, might be concerned about turning into à la carte curriculum providers. But they should be willing to engage the inevitable trade-off questions and to consider how to address the challenges of accepting part-time students if they embrace this option.

Zealously Guard Autonomy and Key Freedoms

As religious-school supporters, we are well aware that concerns about participating in private school-choice programs are nothing new. We’ve heard school leaders express concerns for years. They worry, quite fairly, that taking public money means accepting public control and that they might lose their unique character or, even worse, have the government impose regulations that undermine their religious ethos. Concerns about regulatory overreach have historically led many homeschooling families to oppose participation in any program offering public support—and, in some cases, to oppose the creation of new programs politically.

We want to set their minds more at ease, especially about the concern that participation may threaten religious freedom and operational autonomy. To date, private schools participating in choice programs have enjoyed substantial autonomy, and regulations attendant to participation have not raised religious-liberty concerns. (Arguably, two exceptions are provisions banning discrimination on the basis of LGBTQ status in Maryland and Maine’s voucher programs, both of which have been challenged on free-exercise grounds.)[30] With the exception of a handful of programs for special needs, ESA programs tend to have a lighter touch regulatorily than voucher programs. Advocates and government policy officials should explain this to schools concerned about overregulation.

Lawmakers and regulators should be aware of, and take proactive steps to address, the legitimate concerns of faith-based schools and other providers, as well as of participating families. To encourage participation initially and to guard against the dangers of regulatory creep, statutes should clearly express the intent to protect the autonomy and religious liberty of participants.

Recently enacted programs contain language expressing this intent. For example, Arizona’s ESA statute states that “this chapter does not permit any government agency to exercise control or supervision over any nonpublic school or homeschool.”[31] Florida’s specifies that “the inclusion of eligible private schools within options available to Florida public school students does not expand the regulatory authority of the state, its officers, or any school district to impose any additional regulation of private schools beyond those reasonably necessary to enforce requirements expressly set forth in this section.”[32] West Virginia’s ESA statute clarifies that the actions of service providers are not considered state actions and that any provider is not required to alter its creed, practices, admissions policy, hiring policy, or curriculum in order to participate.[33] South Carolina’s ESA states: “This item shall not be interpreted to preclude any independent or religious educational provider from exercising an exemption allowed under federal law.”[34]

When a new program is enacted, inevitably there is pressure for additional regulations. But demands for “accountability” often mask other motives, including a desire to undermine a program by burying it in red tape. Religious freedom, in particular, should be guarded. Lawmakers and regulators should resist pressure to adopt regulations that raise religious-liberty red flags. Private schools should be fully informed about the scope of preexisting religious-liberty protections, such as the well-established “ministerial exception” doctrine, which immunizes religious organizations from the legal scrutiny of certain employment decisions, as well as about what steps can and should be taken to ensure that they enjoy these protections.[35]

Ultimately, school leaders must make the decision that is right for their institution, and schools that do participate will need to be on guard and oppose any regulations, rules, or policies that impair their ability to stay true to their foundational ethos. Careful planning can ensure that participating schools have the financial freedom to exit programs if regulations become untenable.

Price Transparency for a New Market

Functioning markets require information. Consumers need to know what they are buying and need easy ways to compare what they are buying with other options. Prices are the best tool that we have to convey information, but in “experience goods”—products that can be evaluated only after they have been used—other sources of information can help consumers make better decisions.

With respect to prices, providers should be transparent with parents about how much their services cost. Many educational providers are used to contracting with schools or districts, for whom agreements are hammered out by boards or principals rather than by individual parents. ESAs change the customer, and providers need to change accordingly. Some providers (for example, test-prep tutors) are accustomed to operating in a market fueled entirely by private funds. The infusion of public funds into such a market obviously carries with it the risk of price increases. The best solution to that threat is to help providers publicize what they are charging parents, and for what service.

Regardless of whether disclosure is legally mandated, schools should be transparent about the educational program that they are providing. What are they teaching students, and how are they teaching it? What curricula do they use? What books do students read? How do they use technology? What are their academic and conduct expectations? How do students perform on standardized assessments? Other providers should similarly be transparent in letting parents know what their children will be doing when using their tool or service, how their child’s progress will be measured, how their digital privacy will be respected, and how that progress will be communicated to parents and any other interested party.

After Legislation Come Regulators and Regulations

While legislation constructs the skeleton of an ESA program, regulations put meat on the bones. In our conversations with individuals who have been involved in ESA implementation, they stated several important themes about these regulations and regulatory processes. These include:

Program Management

The first, and perhaps the most critical, implementation decision to be made is who will administer the program. There are two separate program-administration questions. The first question, always answered legislatively, is which governmental department will oversee and regulate the program. In many cases, this job falls on state education departments, although, in some states, legislators choose to give oversight authority to another agency—for example, treasury. This approach is often motivated by a concern that the state department of education or the education secretary may seek to undermine the program regulatorily. This approach obviously is context-specific but should not be driven by a concern that a particular education secretary (who, in some states, is elected) opposes ESA, since subsequent personnel changes may alter that fact. The bottom line is that the seat of implementation authority should be determined with an eye to the long-term future of the program, not short-term political maneuvering.

The second question is whether to engage a private entity to administer the day-to-day details of the program (approving expenditures, paying schools and other providers, etc.). Some ESA legislation mandates that the state issue RFPs leading to an agreement with a private entity to administer the program. For example, Utah requires the state to engage a private program administrator,[36] and Florida provides that the program should be administered through qualified “scholarship granting organizations.”[37] Some other states empower, but do not require, the state to engage a private program administrator.[38] Other programs are silent on the question. The first is the best approach. Although, as discussed below, there remains a need for more development in the program administration/platform space, the existing platforms are still undoubtedly preferable to entrusting details of program administration to government bureaucrats.

After a program is enacted, the state should undertake a transparent RFP process to engage a private entity to administer the program. The state should also set aside sufficient resources to ensure quality program administration. The RFP should make clear the state’s expectations for the private administrator—for example, is an administrator’s engagement limited to the operation of a platform for the submission and approval of expenditures, or does the state expect the program administrator to answer customer-service questions about program rules and permissible and impermissible expenses? The existing companies in the ESA administration space have different policies. Customer service is time-consuming and expensive but necessary. If the program administrator does not answer the questions of parents and providers, that task will fall on the state.


The agency charged with program administration must also undertake a transparent and thorough rulemaking process to establish the regulations governing the program. This initial rulemaking must communicate the spirit and intent of the program enacted by the legislature. For instance, some ESA programs are primarily private school-choice programs. Iowa’s ESA is really a “voucher-plus program,” requiring participating students to enroll in an accredited private school.[39] ESA laws in other states allow (and encourage) students to customize their education outside of brick- and-mortar private schools. These laws vary in the degree of detail specifying permissible and impermissible educational expenditures. To the extent that the goal of the program is to enable expansive parental choice and the customization of student education, the regulations should make that clear. Otherwise, risk-averse providers may simply choose not to participate, stifling innovation.

For example, if the regulations require some form of accreditation for schools to participate, parents will be impeded from forming innovative new schooling models, such as learning pods and microschools. If the statute requires accreditation—as in Iowa—regulators should consider developing alternative accreditation paths. If the aims are narrower—for example, the state intends for the ESA funds to be used at existing private schools and for homeschooling curriculum—the regulations should make this clear. Whatever the goals of the program, they should be clearly stated and easy to find.

The rulemaking process needs to address many important questions:

Who? Who is charged with approving schools and other providers for participation? The state agency with oversight authority or the private program administrator? Should providers other than schools be required to register/get preapproval at all? (Some programs do not preapprove providers other than schools.) Who will be charged with determining the range of qualified expenses? Who will define key terms such as “curriculum” and “instructional technology”?

What? What criteria will be used for making these determinations? What processes will be in place to ensure that parent/provider questions are answered promptly and consistently? What criteria will be used to exclude providers from future participation (because of, for example, parent complaints or concerns about fraud or overcharging)?

How? How will the agency charged with oversight collect feedback from parents about providers and the program administrator? Will there be a parent advisory council? A rating system for providers? Will the amount of money that may be charged for various services or equipment be capped? If so, using what criteria?

Many individuals with whom we spoke expressed concern that a lack of clarity is leading to uneven ad hoc decisions that may open the states to legal challenges about arbitrary and capricious actions. This approach can generate resentment in parents and providers when they learn, perhaps through social media, that an expense was approved for one family but not another. Such inconsistency is especially concerning to parents of children with disabilities, who may make legitimate purchases for debatable items—for example, sensory therapy items that look like toys and equipment to prepare students for life skills that look like household items.

As discussed previously, regulations can promote clarity by establishing categories of expenses that are always approved (tuition, homeschool curricula) and never approved (video-game consoles) and those that are subject to administrative review. The list of presumptively approved items may be different for students with disabilities, or regulators may give the program administrators the option of customizing the list of approved items for students with demonstrated special learning needs.

Academic Accountability

Most (but not all) private school-choice programs require participating schools to administer some standardized test—although typically not the state test—and report those results to parents and state regulators. With a handful of exceptions, schools are not excluded from participating in school-choice programs because of low test results.[40] Most ESA programs follow this model.

Iowa’s ESA will require schools to administer the state assessment;[41] Arkansas’s will require participating schools and providers to administer the state test or another nationally normed assessment approved by the state;[42] Florida’s requires the administration of a nationally recognized standardized assessment; Utah’s will require the administration of a standardized assessment or the submission of a portfolio of student work; West Virginia’s requires participating schools to choose from a menu of standardized assessments but permits parents to administer any nationally recognized norm-reference test;[43] and Arizona’s ESA does not have a testing mandate.[44]

The question of academic accountability is particularly complex. Demands for accountability may be legitimate but can mask illegitimate motives to suppress participation. Testing requirements—especially the imposition of the state assessment on participants—can scare off private schools and homeschooling families who use other methods of assessing progress. There is, in fact, empirical evidence that testing mandates have led some private schools to eschew participation in private school-choice programs, undermining their overarching goal of expanding the number and quality of educational opportunities available to participants.[45]

In the ESA context, several individuals we spoke with cautioned that regulators should tread lightly with the imposition of academic accountability rules; how lightly is a difficult question to answer. At a minimum, we are of the view that ESAs should allow a wide range of assessment options to fit the needs of the participating families and schools. This should include, as is the case in most states with private school-choice programs, a range of permissible standardized tests, not just the state test. Another model is provided by the Utah ESA’s option of choosing between a standardized assessment or the submission of a student portfolio.

Moreover, to encourage the participation of homeschoolers, regulators should, to the extent permitted by ESA law, regulate participants using ESA funds for purposes other than private school tuition in the same way as homeschooling families. Homeschooling families are not required to administer any standardized tests in many states and may refuse to participate in programs that impose a testing mandate.[46] One advocate whom we spoke to suggested that the state statute make explicitly clear that homeschooling families that do not participate in the ESA program are not subject to any additional accountability regulations imposed on ESA participants. Arizona’s program accomplishes this by clearly establishing that homeschooling students are a separate category from students participating and using ESA funds to customize their education.[47]

Financial Accountability

As discussed in greater detail below, many individuals we spoke to were concerned that regulators are misordering their attention by focusing excessively on preventing fraud by identifying and disqualifying inappropriate expenditures rather than on ensuring that ESA programs are working smoothly for participating families.

For example, if regulators (or private program administrators) spend time reviewing many or most purchases, they have less time to answer parents’ legitimate questions. An excessive focus on expenditure review will result in parents needing to ask more questions. If a backlog develops, or if parents feel that they are constantly being required to ask permission, families may grow frustrated and leave the program altogether. It is critical, therefore, that regulators avoid these pitfalls in the initial rulemaking process and periodically reassess whether existing regulations are working for families and providers.

Building on Successes

There are, of course, several states that already have limited ESA programs. While we have pointed out the aches and pains of creating large, universal programs, it is worth recognizing the good things that are happening for programs to build on.

In Arizona, organizations like Love Your School have sprung up to support ESA families and advocate for the program with policymakers.[48] (Love Your School also expanded into West Virginia.) It has an easy-to-navigate website with resources for ESA program parents, staff who can field phone calls and e-mails from concerned parents, a weekly webinar, and a lively Facebook group to connect and support families.

Advocates also applaud the efficiency and effectiveness of the New Hampshire Education Freedom Accounts ESA Program. It was created with minimal extraneous regulations and has been easy for families to join. It saw growth from 2,061 students in its first academic year (2021–22) to 3,110 in its second (2022–23).[49] The primary complaint about the program has been its low value, not anything to do with its operation.

Utah is giving itself 18 months to implement its new ESA program, which begins in 2024. This was another step that advocates praised. Given all the components and the need to get things right, taking a bit more time to build the necessary foundation is a reasonable and prudent step.

The Nitty-Gritty of Policies and Procedures

A central insight of Pressman and Wildavsky’s book Implementation is that projects often rise or fall on the details that do not make for interesting headlines or stirring political slogans.

Implementing ESA programs means drafting a set of policies and procedures that bring families and providers onboard, link them to one another, and allow funds to transfer from one to the other. Different states at different times have come up with different solutions to this problem. In the early days of ESA programs, families were reimbursed for their expenses. They bought the education service, submitted their receipt to the state, and were then cut a check to recoup the cost. As one might imagine, this process had several problems.

First, families had to have the money to pay for the good or service. Not every family has the money to cover a semester of private school tuition or to buy a computer. Second, it was administratively onerous. If a program had only a few hundred families, it might be manageable to make each reimbursement, but as the program grew, it became impossible to reimburse families in this way. And funding programs through reimbursement is a recipe for fraud. While at first blush, it might appear that individually reviewing each purchase before reimbursing is a way to prevent fraud, all a family would have to do is make up a fake tuition bill or an invoice for an educational service. For all these reasons, an alternative to reimbursement was required.

States have enlisted private vendors to create a two-sided platform: parents have an account, and providers have an account. Funds flow from one side to the other. In theory, the platform would look, for parents, like many existing HSA platforms or even payment processors like Venmo. Providers would receive money in much the same way, with clearly stated logs of where the money came from, when, and why.

Sometimes these platforms are not as simple. We heard from frustrated parents and providers that current solutions are clunky and glitchy, and they frequently malfunction. Parents don’t know how much money remains in their ESA accounts. Vendors claim that they aren’t getting paid or that it is taking weeks or months for funds to transfer. Interfaces are not user-friendly.

Much of the education commentariat is fixated on how providers will be held accountable and on ensuring that money is never misspent. But terrible administration will destroy a program long before accountability becomes a problem. If parents cannot get approved to participate in the program, if the payment platform is too clunky to be usable, or if providers opt out because they are not getting paid, programs will wither and die.

To resolve this problem, we offer three steps that states can take to draft policies and create procedures to make it easier for parents and vendors to participate in the program.

Simplifying Parent Onboarding

It should not be difficult for parents to get signed up, qualified, and enrolled in the ESA program. What can we do to make it easier to allow parents to enter?

The first step is determining parent eligibility. The difficulty varies based on the eligibility criteria or prioritization that states place on participation. In a truly universal program, all families need to demonstrate is that the child is a resident of the state, which can be accomplished with one or two documents. If individualized education program (IEP) status or income determines eligibility, or gives a priority spot, more information is required from parents.

To the extent practicable, parents should not have to provide information known by the government. The state’s treasury department knows its residents’ incomes. And state departments of education should know (or be able to confirm) students’ IEP status. Rather than putting the burden on parents to produce, scan, and upload these documents for program administrators to review, states should explore the possibility of developing systems that allow the various departments to talk to one another and collect and link this information. Vendors can help with this process. Parents, if they wish, could sign data-privacy waivers to ease data transfers.

The RFP Process

Everyone we spoke to agreed that there should be more competition in the ESA payment platform industry. Part “govtech” and part “fintech” solutions should be able to straddle the worlds of private industry and government to make payments fast and fluid. RFPs are the primary tools that states have to communicate their desires and requirements to potential vendors. It is vitally important that states take this process seriously, work hard to identify potential vendors, operate in an open and transparent way, and hold firm on key elements that platforms must have.

What are these key elements? As we have stated, platforms must be able to navigate several state agencies to collect the necessary data to rapidly bring parents onboard. Vendors should provide a user-friendly, parent-side interface that allows parents to track, in real time, their ESA balance, as well as pending and settled transactions. There should be a robust customer-service apparatus to help parents when challenges inevitably arise. Vendors should quickly tell parents what is or is not allowed, so that parents are not left waiting and thus lose access to educational options.

On the provider side, vendors should commit to clearing the vast majority of transactions (95+%) within two business days. Some questionable transactions might need to be more carefully reviewed, but those that are clearly allowed (such as private school tuition) should be approved and paid quickly. Additionally, for the sake of speed, payment platforms should pay each provider per individual service, not in batches, as appears to have happened with some existing platforms.

Purchase Reviews

A payment vendor told us that about 98% of ESA purchases clearly fall into the “always approved” or “never approved” categories. The remaining 2% require review. I.e., most parent expenses are simple to process. Paying tuition at a private school: always approved. Buying food at the grocery store: never approved. Product codes embedded in the processing architecture send clear signatures to payment systems as to the nature of what is being purchased. Most of this is clear and simple to work through. Yes, there should be some kind of appeals process when allowable expenditures are accidentally or inappropriately denied, but those are the exception, not the rule.

To minimize these problems, regulators can specify permissible and impermissible expenditures and authorize program administrators to automatically approve or disapprove expenditures falling into one of those two categories. All expenses listed in the legislation creating the ESA program should be automatically reimbursable. These lists should be periodically reviewed but not frequently amended because the latter would create confusion.[50]

For the 2% of expenses that are not easily categorized, there should be a process to ensure that the purchase is aligned with an educational program. For example, some families purchase exercise equipment for physical education, and it is not clear whether the family is buying it for school or for personal enjoyment. That might take additional verification. But again, this is a small minority of purchases, and the payment platform could filter them out so that the purchase could be reviewed by the party empowered with that authority.

There is a broader point here. If preventing misspending is the biggest goal for program administrators, that will lead to a huge administrative burden. Every government program, from food stamps to WIC to Section 8 housing, has some measure of waste, fraud, and abuse. None of these programs attempts to stamp out every potential misuse of funds because they realize that it would paralyze the system. The same is true with ESAs.

After filtering out and escalating the questionable reimbursement requests while allowing the vast majority to be automatically accepted or denied, states could also choose to conduct regular audits of spending, instead of trying to manage every single transaction. Every month or quarter, the state could select some percentage of participating families at random (or using risk-management strategies that identify more likely cases of malfeasance) and go through their expenses. Post-hoc audits would provide accountability for taxpayers and make the entire audit process easier and align program administration with standard practices in similar government programs..

In 2002, the U.S. Supreme Court effectively foreclosed federal constitutional challenges in Zelman v. Simmons-Harris, which held that the First Amendment’s Establishment Clause does not prohibit states from allowing students participating in parental-choice programs to attend faith-based schools.[51] Since then, parental-choice programs have typically been challenged in state courts, with opponents claiming that they violate various state constitutional provisions. Thankfully, most challenges have been rejected by state supreme courts,[52] although notable exceptions include the Kentucky Supreme Court’s 2022 invalidation of an ESA.[53] Despite this track record, parents and education providers must be prepared to face legal challenges.

Planning for Disruptive Litigation

In the short run, legal challenges can disrupt ESA implementation. Opponents may strategically move to block program implementation in court just before the funds become available and after parents and providers have applied to participate. Unfortunately, lower courts may endorse the legal argument and issue injunctions, preventing timely implementation and disrupting parents’ and students’ schooling plans. State officials should anticipate and prepare for this possibility by laying the groundwork for successful and rapid implementation after injunctions are lifted. Schools and other providers should consider how to respond if a program is temporarily enjoined, by, for example, raising private funds for stopgap scholarships in the short term.

Minimizing Legal Risk

Careful legislative drafting can avoid many common legal pitfalls. Legislators should take advantage of the wise counsel provided by EdChoice, the American Federation for Children, and the Institute for Justice, all three of which proactively review legislation to avoid legal landmines. To summarize some key areas:

Avoid Restricted Funding Sources

Many state constitutions require the state to set aside certain funds for the support of public schools, which are often called the “common schools funds.” ESA programs have been challenged for tapping into these restricted funds. For example, in 2016, the Nevada Supreme Court permanently enjoined using common school funds for an ESA program but upheld the use of other funding sources for the program.[54] And in 2013, the Louisiana Supreme Court ruled that funds from the state’s Minimum Foundation Program for public schools could not be used to fund vouchers, forcing the state to find a different way to fund the program.[55] Currently, New Hampshire’s ESA program is being challenged for relying partly on lottery revenues allegedly reserved for use by public schools.[56]

Legislators crafting ESA programs must ensure that they are funded by unrestricted revenue sources (e.g., the general revenue). That said, if ESAs are dependent on annual appropriations from the general revenue, they are vulnerable both to the whims of future legislatures, which may zero out funding, and to the governor’s line-item veto. The best strategy for avoiding legal challenges and preserving continuity of funding will vary by state; but in many cases, legislation can specify that ESAs should be funded using the same formula as public schools (e.g., ESAs are commonly funded at 90% of the state’s per-pupil share of district school funding) but drawing upon unrestricted funds.

Geographic Eligibility Limitations

Most state constitutions technically require state laws to apply uniformly through the state by prohibiting “special and local legislation,” which singles out specific local governments for special benefits or burdens.[57] ESAs that limit eligibility to students in particular geographic regions—for example, the largest cities or school districts—have faced challenges under these provisions. In fact, the voucher program ultimately upheld by the Supreme Court in Zelman v. Simmons-Harris was initially invalidated because it was limited to Cleveland.[58]

More recently, the Tennessee Court of Appeals invalidated an ESA because only children in Memphis and Nashville were eligible to participate, although the state supreme court ultimately reversed that decision.[59] Kentucky’s ESA was also challenged as invalid special and local legislation, although it was invalidated in 2022 on other grounds.

In some cases, at least initially, securing political support for an ESA program may necessitate limiting its geographic scope to the largest city or cities in a state. As a general rule, geographic eligibility restrictions are valid, provided that they rely on requirements that are reasonable (e.g., educational options are most needed in large cities) and crafted so that other cities or districts may qualify in the future (e.g., by gaining population).[60]

Single-Subject and Clear-Title Requirements

ESA programs enacted in recent months were paired legislatively with other education reform measures, such as teacher-salary increases. The omnibus nature of this legislation makes it vulnerable to challenges under state constitutional provisions that require legislation to have a “clear title” and to cover only a “single subject.” Most state courts interpret these provisions broadly and will accept the argument that legislation combining various education reform measures cover the single subject of “education.” State courts considering clear-title challenges also tend defer to legislatures.[61] While legislators should attend to these restrictions when drafting bills, opponents may hesitate to raise them in court because they would call into question the validity of the entirety of the legislation, not just the ESA program.


Legislation often includes a “severability clause” informing courts that if one or more provisions are invalid, the remainder of the legislation should be upheld. But sometimes the opposite is true, and the legislation specifies that all the provisions stand or fall together by directing courts not to sever an invalid provision of the law.

For example, an “in-severability” clause was included in 2017 Illinois legislation that paired a tax-credit scholarship program with major reforms to the public school-funding formula and funding to bail out teacher pensions. The clause provided that the judicial invalidation of any provision rendered the entire statute invalid, thereby warding off legal challenges to the parental-choice program.[62] Short of an in-severability clause, legislation might tie the fate of various pieces of a reform package together. For example, Utah legislation provides that, if the ESA is funded fully, public school teachers will receive an additional $8,400 per year; but if it is not funded, they will receive only an additional $4,200.[63]

Addressing Unavoidable Legal Challenges

While steps can be taken to minimize the risk of legal challenges, not all legal challenges can be avoided legislatively. Next, we consider the three most common challenges and the best responses to them.


“Standing” is a legal doctrine that asks whether the party challenging a government action has been sufficiently harmed to seek redress in court. In federal courts, a plaintiff must have suffered a concrete, individualized, “injury in fact”—rather than a general inchoate objection to the government’s action—in order to have standing to sue. While some states have more expansive standing rules, not all do. Standing arguments should be deployed whenever possible because when the party lacks standing, a court must dismiss the case.[64]

Blaine Amendments

After Zelman made clear that private school-choice programs do not violate the Establishment Clause, opponents began asserting that these programs violate state no-aid provisions, or “Blaine Amendments,” which prohibit the funding of “sectarian” schools and have an undisputed anti-Catholic pedigree.[65] Opponents argued (usually unsuccessfully) that Blaine Amendments require a greater degree of church–state separation than the federal Establishment Clause. Thankfully, a series of recent U.S. Supreme Court decisions—Trinity Lutheran Church and School v. Comer,[66Espinoza v. Montana,[67] and Carson v. Makin[68]—take this argument off the table.

These cases make clear that the federal Free Exercise Clause prohibits a state from excluding religious schools from a parental-choice program (or invalidating one to prevent public funds from flowing to religious schools). As the court admonished in Espinoza, “when the [Montana Supreme] Court was called upon to apply” a no-aid provision “to exclude religious schools from [a parental-choice] program, it was obligated by the Federal Constitution to reject the invitation.”[69]

This nondiscrimination principle does not apply to the handful of no-aid provisions that prohibit the public funding of all private schools, regardless of whether they are religious, as is the case in South Carolina and Arizona. Even in these states, however, ESAs should withstand judicial scrutiny—and, indeed, may prove the ideal parental-choice mechanism—since it is indisputably clear that ESAs fund students, not private schools. After all, unlike voucher programs, ESAs can be used for a wide variety of education expenses other than tuition. Promisingly, ESA and tax-credit scholarship programs have withstood scrutiny even by state supreme courts that previously invalidated voucher programs as running afoul of no-aid provisions, including in Arizona.[70]

Public Education Guarantees and Public Purpose Clauses

Opponents frequently allege that parental-choice programs violate state constitutional provisions that: (1) guarantee a right to a public education; and (2) limit public expenditures to “public purposes.”

There is not much to be said about these arguments except that they are wrong. All state constitutions require the state to establish and maintain a system of public schools, although the language of these provisions varies dramatically.[71] Roughly half of state supreme courts have found that these provisions create a judicially enforceable right to a public education, with most of the rest holding that the meaning of public education guarantees is a non-justiciable “political question” reserved for legislatures to decide.[72] Only one state’s high court has endorsed the argument that a parental-choice program violates a state education guarantee.

In the 2006 case Bush v. Holmes, the Florida Supreme Court invalidated a voucher program on the grounds that it undermined the uniformity of the state’s public school system.[73] While school-choice opponents have consistently advanced similar arguments since then, with some success in lower courts, no other state supreme court has endorsed the argument that a state education guarantee prohibits states from funding educational opportunities outside the traditional public school system.[74]

More recently, opponents of parental choice argued that large-scale parental-choice programs are unconstitutional because they drain public resources from, and undermine the quality of, public schools.[75] While universal ESAs may be attacked on these grounds, these claims are frivolous. ESAs result in a loss of public funding for district schools only to the extent that districts no longer receive the state share of public education funds allocated to students that they are no longer educating. The requirement that a state fund maintain a public school system should not be interpreted to require the state to hold public schools financially harmless when students exit. In any event, many recent reform packages that include ESAs increase public school funding through teacher-salary increases and other mechanisms.[76] Since many aspects of public education funding are fixed or rely on local revenues, ESAs will typically result in districts having more funding per pupil. States defending these programs should readily be able to rebut “sky is falling” claims about the fiscal effects of ESAs.[77]

Although many state constitutions include provisions that prohibit public expenditures to advance “private” purposes (or limit them to “public” ones), the claim that parental-choice programs do not serve a public purpose has never been judicially endorsed. State courts are extremely deferential to legislative determinations of what a “public purpose” is.[78] Even if they were not, private schools’ record of excelling at the important task of education young people, including those most in need, directly rebuts the argument that parental-choice programs do not serve a public purpose.


Winston Churchill was quoted as saying of World War II in 1942: “This is not the end, it is not even the beginning of the end, but it is, perhaps, the end of the beginning.” It is hard not to feel similarly about the passage of the last two years’ ESA legislation. This is a new era for school choice and for the millions of families that are now eligible for financial support to find the best education for their children. It is a time of great hope and great promise but one that advocates, policymakers, and educators need to take extremely seriously.

ESA programs are not going to stand up on their own. It is going to take thoughtful management from both the public and private sectors to ensure that the rules, regulations, policies, and procedures drafted to implement the programs reflect the desires and intentions of the legislators who created them. It is going to take innovation in the tech space to create the kinds of platforms that will make navigating the onboarding families and use of ESA dollars seamless and straightforward. It is going to take schools and other providers stepping up to the plate to provide a quality education as well as to understand the immense opportunities that these new programs represent and their responsibility to the creation of a healthy marketplace. And it is going to take lawyers to protect these programs from their inevitable challenges.

There is much work to do, but the potential rewards are immense. A decade from now, we will look back with regret if the Implementation of the 2030s is a saga of great expectations dashed in Arizona, Arkansas, Florida, Iowa, South Carolina, Utah, and West Virginia.

About the Authors

Nicole Stelle Garnett is a senior fellow at the Manhattan Institute and contributes regularly for City Journal. She is the author of Lost Classroom, Lost Community: Catholic Schools’ Importance in Urban America (University of Chicago Press, 2014) and Ordering the City: Land Use, Policing and the Restoration of Urban America (Yale University Press, 2009), as well as numerous articles on education policy, urban development, and land-use planning.

Garnett is the John P. Murphy Foundation Professor of Law at the University of Notre Dame, where she is also a fellow of the Institute for Educational Initiatives. Prior to joining the faculty at Notre Dame in 1999, she served as law clerk for Associate Justice Clarence Thomas of the Supreme Court of the United States and Judge Morris S. Arnold of the U.S. Court of Appeals for the Eighth Circuit. Garnett also practiced law at the Institute for Justice, a nonprofit public-interest law firm in Washington, D.C., where she helped defend the inclusion of faith-based schools in private school-choice programs. She holds a B.A. in political science from Stanford University and a J.D. from Yale Law School.

Michael Q. McShane is director of national research at EdChoice. He is the author, coauthor, editor, or coeditor of 11 books on education, and his analyses and commentary have been published widely in the media, including in the Wall Street Journal, National Affairs, USA Today, and the Washington Post. McShane has been featured in education-specific outlets such as Teachers College Commentary, Education Week, Phi Delta Kappan, and Education Next. His academic work has been published in Education Finance and Policy and Journal of School Choice.

A former high school teacher, McShane earned a Ph.D. in education policy from the University of Arkansas, an M.Ed. from the University of Notre Dame, and a B.A. in English from St. Louis University.


Online appendix available here.


Please see Endnotes in PDF

Photo: Ibrahim Akcengiz/iStock


Are you interested in supporting the Manhattan Institute’s public-interest research and journalism? As a 501(c)(3) nonprofit, donations in support of MI and its scholars’ work are fully tax-deductible as provided by law (EIN #13-2912529).