Louisiana’s first year of universal-style education savings accounts produced a striking split.
During the 2025-26 school year, Louisiana delivered $43.1 million to parent-controlled education accounts. Overall, 6% of the money was returned unspent. Among students with disabilities, the return rate was 30%, according to ProPublica’s September 2026 investigation into voucher access for students with disabilities.
That does not mean every returned dollar represents a private-school rejection. Families return voucher money for multiple reasons. But many families interviewed in the reporting said they could not find suitable private schools that would enroll their children, which puts the gap between 30% and 6% in a very different light.
The pattern is easier to understand by going back to Florida, where disability-focused vouchers helped establish the political model.
In 1999, Florida created a pilot voucher program championed by Republican state senator John McKay, who had experienced his own family’s frustration with a public school system that struggled to meet his daughter’s significant learning needs. It became the nation’s first voucher system specifically for students with disabilities and was later named the John M. McKay Scholarships for Students with Disabilities Program.
The pitch was narrow and sympathetic: give families of children with disabilities public money they could use to find another educational setting when their assigned public school was not working for them.
For some families, that option has worked. For others, receiving the money is only the first step.
Angela Johnson learned that distinction while searching for a school for her 12-year-old daughter Amelia. Amelia qualified for a Florida voucher worth thousands of dollars, but Johnson could not find a private school equipped to meet her daughter’s needs. The voucher existed. The usable private-school option did not.
Florida’s statewide numbers show that Johnson’s experience is not an isolated administrative quirk. More than half of the state’s private schools report that they do not serve students with any type of disability. In the Florida data analyzed by ProPublica, about 8% of the $1.7 billion distributed through the disability-focused voucher program was returned because recipients remained enrolled in public school. In the parallel program open to all students, the comparable rate was 1%.
The gap between those figures is the story.
Louisiana has now produced an even wider first-year gap: 6% returned overall versus 30% among students with disabilities. Texas’s new $1 billion voucher program has also shown a smaller difference, with students with disabilities opting out of using the money at slightly higher rates than other students as its first-year data comes in.
The mechanism is more complicated than saying federal disability law simply stops at the private-school door.
Public schools have extensive obligations under the Individuals with Disabilities Education Act, including providing eligible students with a free appropriate public education. When parents instead place a child in a private school through a state voucher, federal rules treat that child as parentally placed. The child can still fall within IDEA’s child-find and equitable-services framework, but the federal IDEA regulations make clear that there is no individual right to receive all of the special-education and related services the child would receive in public school.
That difference matters because receiving a voucher does not require a private school to possess the staff, programs or capacity a particular child needs. A family can have money available for tuition and still discover that the schools within practical reach cannot accommodate the child.
Chris Roe, Director of State Policy at the Council of Parent Attorneys and Advocates, has described the resulting imbalance plainly: instead of families simply choosing among schools, schools also decide which children they are prepared to accept.
That distinction becomes more important as targeted programs grow into broader ones.
When Florida opened voucher eligibility to all students in 2023, two-thirds of the newly enrolled voucher students had already been attending private schools. As of September 2026, more than 500,000 Florida schoolchildren, roughly one in six, were using vouchers.
At least 18 states had universal voucher eligibility or were phasing it in, and about 3 million school-age children with disabilities lived in states where families broadly qualified for voucher-style programs. The political constituency and financial scale had moved far beyond the population that helped establish some of the earliest disability-focused programs.
The history was not entirely accidental.
In 2006, as Arizona adopted a program for students with disabilities, Clint Bolick wrote an essay titled “Toe-Hold Strategies.” Bolick was then president of the Alliance for School Choice. He is currently a justice on the Arizona Supreme Court.
His argument was that targeted programs for politically sympathetic populations could establish the legal and political foundation for later expansion. Once lawmakers had accepted a smaller form of school choice, broader versions became easier to pursue.
That expansion has become financially significant. In its January 8, 2026 spending-share ranking, EdChoice estimated that private school choice programs had allocated $10.6 billion during the reporting period it analyzed.
The growth changes the arithmetic for public systems as well as private ones.
Voucher financing differs from state to state, so there is no single formula for what happens to a district when a student leaves. But public schools remain the institutions required to educate eligible children with disabilities even when those children cannot find workable private placements. Enrollment and funding can shift faster than obligations for specialized staff, transportation, facilities and services do.
This is where the marketing language of school choice can diverge from what a family experiences.
The promise is personalization: every child gets a path better suited to their needs, and every parent becomes an active chooser. But personalization requires an available provider. A funding account does not create a nurse, a speech therapist, an aide, an accessible classroom or a school willing to take on a child’s particular needs.
The distinction is especially important because the Louisiana figure does not show that 30% of families with disabilities consciously rejected private education in favor of public school. It shows that 30% of the money allocated to that group came back unspent. Reporting from families provides evidence that private-school access is one reason for that gap, but it is not the only possible reason.
Arkansas offers another useful piece of the history. Its Succeed Scholarship began as a private-school choice program for students with disabilities. The University of Arkansas’s Office for Education Policy noted both practical limits on access and the argument that disability-focused programs could serve as a “foot in the door” for broader school-choice laws.
That source does not prove that every state’s voucher population follows the same trajectory. What it does show is that the possibility of expanding targeted programs was openly part of the policy discussion years before universal eligibility became common.
None of this requires assuming that private schools which decline a child are acting in bad faith. A school without the staff, facilities or specialist support a child requires may genuinely be unable to provide an appropriate placement.
The policy question is what happens when a publicly financed program promises families more choice while leaving admission and service capacity largely on the supplier side.
John McKay’s original argument was rooted in a real frustration: parents whose public school was not meeting their child’s needs should have another option. That case remains understandable. What the newer data exposes is the distance between receiving permission to shop elsewhere and finding somewhere capable of serving the child.
The scale of the experiment is no longer small. EdChoice says 75 educational choice programs are on the books in 34 states, the District of Columbia and Puerto Rico. ProPublica separately counted at least 18 states with universal voucher-style eligibility or programs moving in that direction.
Those are different measures, but together they show how far the policy has travelled from the targeted disability programs that helped establish it.
The Louisiana numbers put the unresolved problem in unusually clear form. In the same first-year program, families overall left 6% of their allocated money unused. Among students with disabilities, the figure was five times higher.
A voucher can put money in a parent’s hands. It cannot, by itself, create a school willing and equipped to take the child standing beside them.