Fund the Student, Abandon the School: What Universal Vouchers Do to the Districts Left Holding Everyone

The phrase is 'funding students, not systems.' It is a good slogan because it sounds unarguable. Who, after all, would defend a system against a child?

But the policy behind the slogan has now been running long enough to examine. Over the past several years, roughly a dozen states have enacted universal or near-universal education savings account programs — Arizona led the way in 2022, followed by Utah, Iowa, Arkansas, Florida, Indiana, Ohio, Oklahoma, West Virginia, Tennessee, Alabama, Louisiana, South Carolina, Georgia, and, after a long and bitter fight, Texas, whose 2025 law phases in a program worth roughly a billion dollars at maturity. Several more have smaller or income-targeted programs. As of this year, the majority of states have some form of private school subsidy on the books.

The money follows the child. The costs do not.

The theory of the voucher is straightforward: take the state's per-pupil allotment, hand it to the family, let them choose. The problem is the second half of the sentence that never gets said. A school district's costs are not per-pupil in the short run. Buildings, buses, insurance, the principal's salary, the special education coordinator, the debt service on a gymnasium built in 1998 — these are fixed. When a hundred students leave a district of two thousand, the district does not shed a hundred students' worth of expense. It sheds a hundred students' worth of revenue and then has to find the difference somewhere: larger classes, deferred maintenance, and eventually consolidation.

That squeeze is survivable in a suburb with a growing tax base. It is a slow emergency in a rural county where the school is the largest employer, the football field is the only civic institution, and there is no private school within an hour's drive to spend a voucher at. It is worse in a high-poverty urban district where the students most likely to leave are the ones whose families have the wherewithal to navigate an application portal, arrange transportation, and absorb the gap between the voucher and tuition.

What the state-level data actually shows

Three findings recur across the states that have run these programs longest.

First, the programs cost more than projected, often substantially. Universal programs attract far more participants than the fiscal notes assumed, and in several states the ESA line item has grown faster than the public school formula it was supposed to complement.

Second, a large share of the money goes to families already paying private tuition. In Arizona, state data has consistently indicated that a substantial majority of new ESA applicants had no record of prior public school enrollment. In other words, a considerable portion of the expenditure is not an exit ramp from a failing school; it is a subsidy for a choice already made.

Third, take-up skews upward. Analyses of enrollment in several states have found ESA participants more likely than the public school population to be white and from higher-income households — a pattern that should be unsurprising given that the programs require families to front tuition and be reimbursed, or to find a school that will accept the voucher amount as full payment.

The strongest case for choice

None of this is an argument that the status quo is fine, and it would be dishonest to pretend otherwise. There are children in genuinely failing schools right now, and telling their parents to wait another generation for reform is a moral failure dressed as patience. Black families' demand for educational self-determination has deep roots running back through the freedom schools of the 1960s, and dismissing it as a corporate plot is both condescending and historically illiterate. Competition can, in principle, discipline complacent systems.

But a policy must be judged by what it does, not by what it promises. Exit without entry is not choice — it is abandonment with a receipt. In large parts of rural America there is no second option to exit to. For a child with a disability, the voucher frequently covers tuition but not the therapies, aides, and legal protections that an individualized education program guarantees in public school, where private schools are under no obligation to provide them. And the competitive-effects literature remains genuinely mixed, with the most careful studies finding modest gains for participants and little consistent evidence of improvement in the public schools left behind.

Who is left in the building

The students who remain in the district after a universal program matures are, disproportionately, the most expensive ones to educate: children with significant disabilities, English learners, students in foster care, students whose families are in housing crisis. Their schools are now asked to do more with less, and the political coalition that once defended those schools has been partially purchased out of it — a dynamic that rural Republican legislators in Texas recognised when they broke with their own leadership over the voucher bill, citing the damage to their hometown districts.

The squeeze is the point, or it should be resisted

There is a version of school choice that is pro-equity: generously funded, targeted to low-income families, with real accountability for accepting schools and a hard commitment to hold districts whole. That is not what most of these statutes do.

A state that offers a voucher to every family and a cut to every district that loses one has not empowered parents; it has arranged for the slow defunding of the only institution some children will ever have — and the families with no car, no tuition gap, and no private school down the road will be the ones who find out first.