Summary

  • A ProPublica investigation documents that private schools accepting public voucher funds systematically decline enrollment of students whose disabilities exceed standard classroom capacity, while most state legislatures have not required participating private schools to accept those children.
  • Families of children with severe disabilities return voucher funds at multiples of the general-program rate — 8% versus 1% in Florida, 30% versus 6% in Louisiana — because private schools they can reach decline to enroll their children.
  • Most current voucher recipients do not have disabilities, and two-thirds of the students who joined Florida’s universal-eligibility program in 2023 were already attending private schools before applying for public funds.
  • Twenty-seven years after Florida’s McKay Scholarships launched the first disability-targeted voucher, lawmakers have neither required participating private schools to accept children with disabilities nor funded specialized providers at sufficient scale, because keeping mandates minimal is the operational premise on which voucher expansion has rested.
  • Donella H. Meadows, in “Thinking in Systems,” described a reinforcing loop in which public schools lose enrollment, lose per-pupil revenue, and lose capacity to serve high-need students, becoming the school of last resort for the children whose needs justified the original programs.

A ProPublica investigation finds that private schools accepting public voucher funds frequently decline to enroll the children whose stories state lawmakers cited when creating the first disability-targeted voucher programs in 1999. About 8% of the $1.7 billion Florida distributed last year through its disability voucher program was returned — eight times the 1% return rate for the state’s general program — and many of the families who returned the funds reported they could not find a private school that would accept their child. Louisiana’s first-year disability-voucher return rate reached 30% against 6% overall; Texas students with disabilities are opting out at slightly higher rates than other students. The reporting, drawing on data from four states, shows that twenty-seven years after Florida Sen. John McKay launched the McKay Scholarships as the nation’s first statewide disability-targeted voucher, the cohort whose needs justified the original program is, by ProPublica’s account, the same cohort the now-$10.6-billion universal voucher system most reliably shuts out.

The return-rate gap

The most direct evidence in ProPublica’s reporting is the differential between disability-voucher and general-voucher return rates — the share of state money placed in family-controlled accounts that goes unused. Florida disbursed about $1.7 billion through its disability program last year; 8% of that total was returned. The state’s general program returned 1% for the same reason. Families return money for multiple reasons, but many indicated recipients remained enrolled in public school rather than finding a suitable private school, according to the nonprofit that disburses most of the funds, and disability funds returned at multiples of general-program rates.

Louisiana, which launched its education savings account program in the 2025-26 school year, delivered $43.1 million to parent-controlled accounts; 6% was returned overall because families did not spend the money, but among students with disabilities the return rate reached 30%. In Texas, where data on the first year of the state’s $1 billion voucher program is still being collected, students with disabilities are opting out at slightly higher rates than all other students. Florida is one of three states where ProPublica obtained data permitting the disability-versus-general comparison, and the pattern runs in the same direction in each.

The reasons families cannot spend the money are documented in the cases ProPublica assembled. Angela Johnson, a physical therapist outside Pensacola, returned her daughter Amelia’s roughly $10,500 award after a weeks-long hunt. Amelia, 12, requires a feeding tube, a brain shunt, and a nurse trained for medically fragile children. Johnson reported that private schools she contacted initially said they could help, then declined once she described her daughter’s needs. “The vouchers are basically designed to make you go to private school,” Johnson said, “but my kid can’t access private school.”

Marguerite Lane, an Atlanta-area mother of a daughter with autism, received a written rejection from a small, arts-infused private school: “We don’t believe we can support her needs and help her thrive in this class, and we also don’t want to compromise the educational experience of the other students.” Diavonni Jordan, a single mother of an 8-year-old son with ADHD, was told by one Christian school: “The support that we could give in our 1st grade would not be enough for him this year.” A second Christian school wrote: “Unfortunately — it doesn’t look like we would be a good fit for your family.”

Christy Black, an advocate with Disability Rights of West Virginia and mother of a daughter with Down syndrome, said she “called every private school from Huntington to Charleston because that’s as far as we could drive her and work. But no one would accept her because she has a cognitive disability.” A small Baptist academy in rural Georgia invites applicants to “join our family” but declares in its handbook that it is not equipped to educate children with learning disabilities or ADHD: “If such students are accepted in our school, it is up to the parents to provide extra services other than what the classroom teachers would do in the confines of the classroom.” In 2023, Taylor Cordes, a former special education teacher, reviewed handbooks for 660 North Carolina private schools and found that only about a third indicated they would accept students with even limited disabilities; more than 1 in 10 declared they would not.

Chris Roe, director of state policy at the Council of Parent Attorneys and Advocates, summarized the pattern: “Families are at the mercy of the schools — and schools choosing their kid — versus the other way around.”

The geographic mismatch

Even where specialized private schools exist, ProPublica found, geography concentrates capacity in ways that leave large areas unserved. More than a dozen Florida counties have no private schools that serve children with disabilities or have only one school with limited offerings. More than half of Florida’s private schools say they do not serve students with any type of disability. More than half of Utah’s counties have no private schools that take vouchers for students with disabilities; in Georgia, 71% of counties have none. In Louisiana last year, among about 350 private schools, only 14 reported enrolling a single child with a disability; eight of those admitted five or fewer.

Where specialized private schools do operate, they tend to charge tuition above the standard voucher amount, leaving families who can afford the gap with options and those who cannot without them. Jinny Kim, managing attorney at the Disability Rights Education and Defense Fund, described the structural mismatch: “It’s touted that this program is an equalizer, and then people have choice. But then if you really look at the details, it’s not actually how it happens.”

Who the money reaches

The reporting also documents a constituency shift in who the voucher money actually serves. Today, at least 18 states have voucher-style programs with universal eligibility or are phasing them in, channeling roughly $10.6 billion in public money. Florida eclipses the others: more than a half-million Florida schoolchildren — roughly 1 in 6 — use a voucher today.

When Florida opened its voucher program to all children in 2023, two-thirds of the newly enrolled students were already attending private schools. Two-thirds of states now have some kind of voucher-style program on the books. ProPublica characterizes this as a population shift: most current voucher recipients do not have disabilities, and they did not need to hunt for a private school that would enroll them — they were already enrolled when they tapped this new windfall. About 3 million school-age children with disabilities live in states with universal voucher eligibility, and the disability cohort that justified the original programs has not moved into private schools in proportion to the expansion of public funds.

The enrollment-loss cascade

The mechanism by which the gap deepens is structural. Federal law — the framework commonly identified with the Individuals with Disabilities Education Act (IDEA) — requires public schools to provide a free appropriate public education to children with disabilities. Private schools accepting public money are generally not bound by the same obligations. As voucher programs expand and divert enrollment from public schools, public schools lose per-pupil revenue that supports the specialized therapists, aides, and equipment on which disabled students depend. The same enrollment loss generates political pressure for further voucher expansion, deepening the cycle.

West Virginia state Sen. Patricia Puertas Rucker, a Republican who helped author her state’s universal voucher-like program, acknowledged the cost dynamic: “It is very expensive to have all this special equipment and to even build the buildings for children that have very specialized special needs.” Rucker said she regrets that her colleagues approved the payments without additional money for students with disabilities, as some states offer — funds that could help families afford the higher tuition often charged by specialized schools and make it more viable for operators to open them. Beyond raising those payments, Rucker said she does not think the state should do anything to cajole or require private schools to enroll children with disabilities: “The minute you start doing that, you are basically disincentivizing private schools from participating, and then you’re denying parents. I do not want to do that.”

The pattern aligns with what Donella H. Meadows, in “Thinking in Systems,” described as a reinforcing loop: public schools lose enrollment, lose per-pupil funding, lose capacity to serve high-need students, and thereby become the school of last resort for the children whose needs justified the program’s existence. Three cross-domain effects follow. On fiscal policy, public dollars subsidize tuition for students already enrolled in private schools, while disability-targeted funds return at higher rates than general funds. On civil rights, the federal disability-education framework obligates public schools to provide a free appropriate public education; private schools accepting public money have no comparable obligation under that framework. On family economics, weeks-long searches, multiple school visits, and the cost of repeated rejection are borne almost entirely by the families least equipped to absorb them.

The framing history

The disability-targeted voucher programs began in 1999 with Florida’s McKay Scholarships. Ohio followed Florida’s lead, then Utah, Arizona, Georgia, Louisiana, Oklahoma, and others. As politicians debated those first programs, a common refrain across statehouse chambers described the plans as narrow: “It’s a very limited scholarship. This really is about special needs,” a Utah representative said in 2004. “It’s not, as I heard some of you say to me, the beginning of vouchers. No, ladies and gentlemen, it’s a program to help a group of students that need our help,” a Georgia House member told colleagues in 2007. Florida’s sponsor framed disability vouchers as a way to “give any disabled student whose parents are dissatisfied with their assigned public school the ability to attend the public or private school that best fits their needs.” An Oklahoma lawmaker pledged recipients “can just take the scholarship and go to private school.”

Behind the political scenes, advocates described a two-step strategy in writing. As Arizona adopted its first disability program in 2006, Clint Bolick, then president of the Alliance for School Choice and now an Arizona Supreme Court justice, published an essay titled “Toe-Hold Strategies.” Bolick wrote that voucher advocates were “acting smarter” by pushing smaller programs for students “whose needs were harder for politicians to oppose.” “The strategy makes sense from a moral perspective, for it focuses assistance on the neediest schoolchildren,” Bolick added. “It also works politically, because choice begets choice: Once the Rubicon is crossed and legislators vote to adopt a school choice program — no matter how small or targeted — it becomes easier to support a new one, or expand the old one, the next time around.”

Sarah McKenzie, executive director of the University of Arkansas’s Office for Education Policy, wrote in 2016: “Special education private school choice programs are often seen as a ‘foot in the door’ for school choice laws. Once some success has been shown to the public, more laws can be passed to expand these programs.” At a 2011 “school choice” conference, Howie Beigelman, then deputy director of public policy for the Orthodox Union’s advocacy arm, said: “Special needs is a very easy way to get into this business. Anyone, no matter how partisan they are, it’s very hard to say special needs is not where people need ‘special’ education.” Beigelman subsequently told ProPublica by email that his remark was his opinion and he had “never seen anything ‘intentional’ about using special needs scholarships for anything more than what they are plainly for: helping educate children with special needs.”

The defender’s case

Policy defenders — state legislators who have authored or expanded programs, school-choice advocacy organizations, and the private schools that have grown under voucher funding — press four objections.

First, the return rate is an imperfect proxy. Voucher administrators have noted funds go unspent for reasons unrelated to school capacity — families moving, choosing homeschooling, deferring expenses, or failing to complete administrative steps. The disability-versus-general differential is large enough to be informative but does not, by itself, prove systemic exclusion.

Second, the documented cases are individual and cannot represent the universe of voucher-using families. Florida’s program alone serves more than half a million children. The families ProPublica selected — children with severe disabilities, single mothers, families in counties with no specialized schools — may not be representative.

Third, the strategic-history statements are 15 to 20 years old; current legislative intent is a poor inference from founding-era strategic writing, and Beigelman has explicitly denied intent in correspondence with ProPublica.

Fourth, the autonomy argument. Former West Virginia Delegate Amy Summers, also a Republican, said: “Getting the government involved in private schools isn’t the answer.” Rucker said requiring private schools to enroll children with disabilities would be “disincentivizing private schools from participating, and then you’re denying parents. I do not want to do that.” From this perspective, the structural gap is the predictable cost of a policy whose principal virtue is the absence of government mandates. The autonomy argument is the defender’s strongest ground, identifying a trade-off the reporting does not fully engage. The strategic-history rebuttal is the weakest: the same framing appeared in writings and remarks across two decades and several states, weakening the “isolated quote” objection.

The network

The reporting traces a hub-and-spoke arrangement with state legislatures at the center, channeling public money through nonprofit disbursers to private schools and families. Spokes include disability-rights advocacy organizations that pushed for the original programs, school-choice organizations that built on disability access as an entry point, the private schools themselves, and the federal government, whose disability-education framework applies to public schools only. Around this hub, family advocates, homeschooling cooperatives, and emerging microschool operators add further density.

Rucker described microschools as a parent-led alternative in which families pool voucher funds to “band together” and create small schools, allowing them to “foot tuition bills” for settings the existing private market has not provided. Microschool operators sit at a structural pinch point: they absorb families the existing private network has declined, while drawing on the same public funding stream that produced the original gap.

Two non-obvious connections structure the network. First, the “toe-hold” relationship between voucher advocates and the disability community: disability access functioned as a politically defensible entry point that built legislative coalitions and normalized public funding of private education. “Choice begets choice,” Bolick wrote, and once legislators cross the Rubicon of voting for any voucher program, subsequent expansions become easier. Second, the feedback loop between public-school enrollment loss and reduced services for the students the original programs were meant to help. Where voucher money flows primarily to families already in private schools, public schools absorbing disabled students face rising per-pupil costs and falling revenue.

A third, less visible connection runs between the universal-voucher expansion and the original disability programs’ political legitimacy. The McKay Scholarships and their successors were sold on the basis that disabled children deserved options. The options are, for many such children, unavailable. The political cost of that gap will accrue to the coalitions that built the programs; whether it produces policy correction or political defensiveness is the question the cascade leaves open. North Carolina state Sen. Lisa Grafstein, a Democrat and a civil rights lawyer who advocates for people with disabilities and opposes her state’s voucher system, said: “The thing that genuinely infuriates me is when people with disabilities are used as pawns.” That characterization is one legislator’s view, not a documented strategy in force today.

The mechanism not deployed

Two decades after states began adopting disability vouchers, lawmakers have done little to improve access to private schools for these children, even as they have expanded programs to all comers, ProPublica found. Most legislatures have not required private schools that take public money to accept even a few children with such needs, nor have they created effective incentives to help operators open more specialized private schools, which can be costly to run. That is largely because one of the driving forces behind vouchers is the goal of keeping rules to a minimum so private schools will take the public money.

A verifiable projection

The figures ProPublica presents are consistent with a verifiable projection: in states where voucher expansion has proceeded furthest and the political rationale has shifted from disability access to universal eligibility, return rates for disability funds should rise and private-school participation among disability recipients should remain flat or fall. The available figures — Florida’s 8% disability-return rate against 1% for the general program, Louisiana’s 30% against 6% overall, Texas’s higher opt-out rates among students with disabilities — fit that pattern. Whether the pattern continues will be visible in the next round of state reporting.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Consequences & Sequels
Plays a decision forward to its first- and second-order consequences.
Red-Team Advocate
Argues the adversary’s case in full to expose what a plan underrates.
Relationship Mapping
Extracts the network of ties among people, institutions, and entities.