Florida Already Answered the Voucher Question

Austin ISD is running a $181 million deficit while sending $821 million a year in local property taxes back to the state through “recapture,” a Robin Hood mechanism where property-wealthy districts send their surplus tax revenue to the state for redistribution to poorer ones. The per-student funding allotment has been stuck at $6,160 since 2019 (HB 2 added a temporary $55 bump for 2025-27, but it reverts), and enrollment has been declining steadily as families move from central Austin to suburban districts and charter schools draw from AISD’s remaining zone. AISD’s share of Travis County’s school-age population has dropped from 47% to 38% since 2013, driven by suburbanization (to both other Travis County districts and adjacent counties like Williamson and Hays) and school choice. (For the full breakdown, including Census and IRS migration data, see our earlier analysis.)

Now there’s a new variable. In 2025, Texas signed SB 2, creating the Texas Education Freedom Account (TEFA) program: a universal voucher worth roughly $10,000 per student (the exact amount varies by year), $1 billion in funding for the first two years, launching fall 2026. Any school-aged child eligible to attend Texas public schools can apply. Because demand exceeds the initial funding cap, slots are awarded via a priority lottery that favors low-income families (up to roughly $156,000/year for a family of four), students with disabilities, and siblings of current participants, with 80% of funding reserved for those priority groups. TEFA is an education savings account (ESA), meaning the money goes into a parent-controlled account that can be spent on private school tuition, tutoring, curriculum materials, or other approved education expenses. When we wrote the AISD piece, we said to watch the application numbers to see if it would matter.

274,000 Texas families applied in the first window. Three-quarters were already in private school or homeschool; just 25% came from public schools, and applicants skewed whiter and wealthier than the public school population they’re drawn from. (These are first-window application numbers; the funded-participant mix may shift as later waves are processed.)

But Texas doesn’t have to guess what happens next. Florida passed an almost identical law, House Bill 1 (HB 1), in March 2023: universal eligibility, ESAs worth $8,000 to $10,000 per student, no state standardized test required for recipients, and private schools retain full admissions control. Florida is two full school years into its experiment, with a third underway. The data is in.

This piece uses Florida’s actual numbers, pulled from FLDOE datasets on OpenData, to project what the next few years look like for Austin ISD and Texas public schools. We’re not forecasting; we’re looking at what already happened in a state that did the same thing two years earlier.

The same policy, different state seal

Line them up and the overlap is hard to miss:

The cost trajectories track each other too. Florida’s voucher spending now runs about $4 billion a year (roughly 8% of the state’s K-12 budget), according to the Florida Policy Institute and Education Law Center. Texas’s Legislative Budget Board (LBB) projects TEFA will grow from $1 billion in year one to $4.7 billion by 2029-30, with at least $805 million in annual lost public school funding by that point, driven by enrollment declines that reduce how much state funding each district receives. Florida’s actual spending validated the independent projections that legislative proponents initially dismissed as too high, and Texas’s LBB numbers are following the same curve.

Florida’s Year 0 figure ($1.1 billion) reflects the pre-HB 1 means-tested voucher spending through the FTC and FES programs. Texas had no comparable program before SB 2, so its Year 0 is zero. The convergence by Year 3-4 is the part that matters: both programs land in the same $4-5 billion range once universal eligibility reaches saturation.

What happened to Florida’s enrollment

Florida’s public school enrollment tells a story with a clear inflection point. From 2016 through 2023, enrollment grew steadily, recovering fully from the COVID dip and hitting an all-time high of 2.87 million in 2023-24. Then it reversed. The decline started the year HB 1 took full effect.

In the two years since HB 1 took full effect, statewide enrollment dropped by 79,355 students, erasing the post-COVID recovery and falling back to 2020-21 pandemic levels. The timing tracks with HB 1’s launch, though charter growth, migration, and housing costs also accelerated in the same period, and isolating the voucher-specific share of the decline isn’t possible with the data Florida publishes. We’ll come back to that gap.

The kindergarten numbers are a leading indicator, because those students never entered the public system at all. Kindergarten enrollment dropped 9.3% in just three years:

18,417 fewer kindergartners in three years. These aren’t students who transferred out of public school; they’re kids who started their education outside the public system entirely, many of them funded by the state for the first time under HB 1’s universal ESA.

The losses concentrate in Florida’s largest urban districts:

One detail reframes those numbers. According to the Florida Policy Institute and Step Up for Students data (the private scholarship organization that manages over 99% of Florida’s voucher funds), roughly 70% of new voucher recipients under HB 1 were never enrolled in public schools. They were already in private school or homeschool. Only about 30% represent actual exits from the public system. Texas’s TEFA data shows a similar pattern: only 25% of 274,000 applicants came from public or charter schools.

The districts that broke first

Enrollment decline on paper becomes school closures, layoffs, and community destruction in practice. Florida’s largest districts are providing a real-time preview.

Broward County (the 6th-largest district in the US) lost 10,360 students in a single year, with 91% of the loss from traditional district schools. The district is now evaluating 34 schools for potential closure, with 58 schools operating below 70% of building capacity.

Duval County (Jacksonville) projects an enrollment deficit of 10,000 students compared to five years ago. To maintain the state-mandated 3% reserve (below which the state can trigger a financial takeover), the district eliminated over 700 positions and plans to close up to 18 schools over five years. Four of the six schools slated for immediate closure have student populations that are over 90% non-white.

Miami-Dade reported an unexpected decrease of 13,060 students (4% drop) for 2025-26, responding with hiring freezes and aggressive promotion of magnet and career academy programs to compete for enrollment share.

Even small districts aren’t immune. In Flagler County, a projected drop of just 432 students instantly created a $2.5 million deficit.

The pattern is the same one Austin ISD is already living through from charter growth and recapture, but accelerated: enrollment drops, fixed costs (buildings, buses, pensions) spread across fewer students, budget shortfalls open, schools close, and remaining students concentrate in fewer, more stressed campuses.

Projecting the math onto Austin ISD

Florida’s data gives us a range, not a point estimate, for what Austin ISD might face. Applying statewide averages to a single district is methodologically weak, so let’s use a band derived from Florida’s observed urban district experience, constrained by what SB 2’s spending caps actually allow.

The cap matters in year one. SB 2 limits total TEFA spending to $1 billion in the first two years. At $10,474 per voucher, that’s roughly 95,500 vouchers statewide. AISD represents about 1.3% of Texas’s 5.4 million public school students, so a proportional allocation gives the Austin area roughly 1,200 vouchers. With only 25% of TEFA applicants coming from public schools, that’s about 300 AISD students leaving in year one. The cap is the binding constraint, not demand.

After year one, the cap ramps fast. The LBB projects total TEFA spending reaching $4.7 billion by FY2029-30 (see cost trajectory chart above for the year-by-year ramp). At those levels, the cap stops being the constraint. Florida’s urban district experience becomes the better model. Hillsborough was relatively stable through 2023-24, then lost 10,749 students (~2.4% per year) in the two years after HB 1 took full effect. Miami-Dade lost 14,331 in a single year (2024-25 to 2025-26), a 4% drop. Across comparable urban districts, the post-HB 1 acceleration above pre-existing trends runs roughly 1-3% of enrollment per year. Applied to AISD’s ~72,000 students, that’s 700-2,100 students per year in voucher-driven losses once the program reaches scale.

The range: 300 to 700 students lost in cap-constrained year one, accelerating to 1,000-2,100 per year as the cap lifts. Cumulatively, 1,400 to 3,500 by year three. But the headcount projections are only half the story. What matters is what each departure does to the budget.

These are scenarios, not forecasts, based on what happens if Austin tracks comparable Florida urban districts, constrained by SB 2’s spending caps in year one. The actual number depends on TEFA uptake, private school capacity in Travis County, and legislative adjustments. But 274,000 applications in the first window suggest demand will outstrip the cap, and the ramp to $4.8 billion by year four means the constraint is temporary.

One thing these projections don’t capture: AISD is already losing students to charter schools, and that pressure isn’t going away. Statewide, charter enrollment has grown to nearly 500,000 students, and charters now absorb 15.3% of state K-12 funding, nearly double their enrollment share, because they draw almost entirely from state funds rather than local property taxes. AISD’s enrollment dropped from 86,000 to 72,000 over the past decade, and charter growth was a significant driver. Voucher exits are additive to that existing trend, not a replacement for it. The projections above model only the voucher-driven acceleration. The charter draw continues underneath.

What each lost student actually costs AISD

To understand the funding impact, you need to know how Texas school finance works, and for a district like Austin, the mechanics are counterintuitive.

The state sets a “formula entitlement” for each district: the amount it’s supposed to spend per student. Start with the basic allotment ($6,160 per student, frozen since 2019, with a temporary $55 bump from HB 2 for 2025-27 that reverts after). Then multiply by a cost-of-education index that adjusts for local wages (~1.12 for Austin) and by weighted student counts that add funding for special education, bilingual, and economically disadvantaged students (~1.20 average across AISD). The math works out to roughly $8,300 per student for AISD.

Here’s the catch. Districts like Austin collect far more in local property taxes than their formula entitlement. AISD collected $739 million in maintenance and operations (M&O) property taxes in 2025, but its formula entitlement (what the state says it needs) is much less than that. The difference goes to recapture: the state takes the excess and redistributes it to property-poor districts.

Think of it like a salary cap. The state says AISD is entitled to spend $X based on its student count. Austin taxpayers generate way more than $X. The overage gets clawed back. In 2025, AISD paid $772 million in recapture (PEIMS accounting; AISD’s own budget documents put the total at $821 million including tax base assignments). That’s more than half of every property tax dollar Austin schools collect.

Now watch what happens when a student leaves for a voucher school. The district’s formula entitlement drops by ~$8,300 (one fewer student). But Austin’s property values don’t change. Local tax revenue stays the same. The amount AISD is allowed to keep from its own property taxes just shrank by $8,300, so that money is swept directly into the district’s recapture payment to the state. The funding loss isn’t a theoretical line-item on a state ledger; it’s real money leaving the district’s operating budget through a higher recapture bill.

TEA’s 2025 financial data shows how this plays out in real dollars:

That’s before TEFA takes effect. And in a recapture district, every student who leaves makes the math worse, not better: entitlement drops, local revenue holds, recapture grows.

Meanwhile, the expenditure side barely moves. When a student leaves for a voucher school, the school building doesn’t get cheaper to heat, insure, or maintain. Bus routes don’t shrink by one seat. AISD’s only real lever is cutting staff and closing campuses to match declining enrollment, and they’ve already been doing both: 10 school closures and hundreds of layoffs so far. But voucher exits are harder to forecast than demographic trends. The district can watch enrollment data for the upcoming year and identify which schools are declining fastest, but by the time that data is clear, staffing decisions for the next school year may already be locked in. The result is a lag: revenue drops immediately when students leave, but costs take a year or more to catch up.

That squeeze, revenue dropping fast while costs drop slow, is something every district faces when enrollment declines, but most districts at least have a buffer. In Florida, when a student leaves for a voucher school, the district loses state per-pupil funding but keeps its local property tax revenue as a cushion. AISD doesn’t have that cushion. Recapture already takes more than the district keeps.

Applied to the enrollment projections above, at ~$8,300 in formula entitlement per student:

The low scenario ($39 million annual loss by 2030) assumes AISD’s enrollment decline tracks the lower end of Florida’s urban district experience, with year one constrained by SB 2’s $1 billion spending cap. The high scenario ($62 million) assumes an urban premium on voucher uptake and a faster ramp from 25% to 30% public school exit rates. Both are cumulative: each year’s new departures add to the prior year’s losses, because once a student leaves for a voucher school, they typically don’t come back.

Stack that on the existing $181 million deficit:

This assumes no cost-cutting measures are implemented. AISD has already closed 10 schools and cut hundreds of positions, and more cuts are planned. The chart isn’t projecting the actual deficit; it’s showing the additional burden voucher losses place on a budget that’s already underwater. In the high scenario, that burden pushes the total gap to $243 million by 2030, roughly a quarter of AISD’s $933 million operating budget, without a single additional student leaving for charter schools or any change to the recapture formula.

These projections are also linear. They assume a steady rate of voucher exits each year based on Florida’s experience. They don’t model what happens when the losses start feeding on themselves: funding drops, programs get cut, class sizes grow, schools close, and the district becomes less attractive to the families still in it. That accelerates the exit rate beyond what a straight-line projection captures. Florida is already showing early signs of this. Broward lost 10,360 students in one year and is now evaluating 34 schools for closure. Each closure concentrates remaining students into fewer campuses, which pushes more families to consider alternatives. Duval cut 700 positions and is closing up to 18 schools over five years, with four of the six immediate closures in schools that are over 90% non-white.

AISD already closed 10 schools in November 2025 to address its pre-TEFA deficit. If voucher losses accelerate the same pattern, the projections above are conservative. The actual trajectory could look more like a curve than a line, with exits compounding as the system visibly degrades. We don’t model that acceleration because we can’t quantify it yet, but the feedback loop section below lays out why it’s the most likely direction.

The $5 billion question nobody can answer

Texas has built expensive infrastructure to measure whether public education spending works: STAAR testing (the state’s standardized exam), A-F accountability ratings, state interventions for chronically failing campuses. Charter schools get the same treatment. If you’re spending $10,000 per student in public money, you need to know if it’s working.

SB 2 spends roughly $10,000 per student in the same public money, then exempts recipients from every one of those accountability systems except a commercially available standardized test (like the MAP or SAT) reported to the Comptroller.

The accountability gradient makes the asymmetry concrete:

Three tiers of schools, all receiving the same taxpayer dollars per student, with accountability requirements that decrease as you move from left to right.

SB 2 does require private schools to administer a commercially available standardized test (like MAP or SAT). Parents report scores to the Certified Education Assistance Organization (CEAO), which passes aggregated results to the Comptroller for an annual report. School-level results aren’t published. There’s no A-F rating. The state can remove a school for compliance violations (fraud, safety, failure to administer tests), but there’s no mechanism to intervene based on academic outcomes. A private school producing terrible test results keeps receiving $10,474 per student, because the state has chosen not to collect the school-level data that would tell it otherwise.

Florida proves this isn’t theoretical. The structure is similar: private standardized tests instead of the state exam, no state assessment for voucher students, results reported to a private scholarship organization rather than the education department. Two years in, FLDOE stopped publishing annual reports on participating private schools. The Florida Auditor General found compliance failures even within the minimal oversight that exists, including late payments, failure to correctly prioritize existing Unique Abilities scholarship students during the transition, and illegal accumulation of unspent funds exceeding $50,000 per student in some accounts.

This doesn’t tell you whether school choice is a good idea. It tells you Texas is about to spend $4.7 billion a year on a program it has deliberately chosen not to measure.

Who actually gets the money

Universal vouchers create two separate fiscal problems that are easy to conflate. They’re both real, but they work through different mechanisms and hit different budgets. The headline number is the one to sit with: 75% of Texas TEFA applicants were already in private school or homeschool.

The 70-75% who were already paying their own way are the bigger fiscal story. Before universal vouchers, the state spent $0 on these families. Now it’s paying $8,000-$10,000 per student for kids who were never in the public system. This money comes from the state general fund, not from public school budgets directly. Texas has a surplus today, so supporters argue it’s affordable. But the LBB projects the program growing from $1 billion to $4.7 billion by 2029-30, mostly driven by this group, and surpluses don’t last forever at that burn rate. It’s worth asking where that surplus comes from. A significant chunk is recapture revenue: Austin ISD alone sends $821 million a year in local property taxes back to the state, and other property-wealthy districts (west Texas oil country, Dallas, Houston) contribute billions more. The state is effectively collecting local tax dollars from public school communities, then redirecting them to subsidize private school tuition for families who were never in the public system.

There’s a physical constraint that sharpens this. Austin’s top-tier private schools generally operate at or near full capacity, and several are opting out of TEFA entirely to preserve independent governance and admissions control. That means there aren’t many open seats at these institutions for an influx of public school transfers. The program structurally functions more as a financial subsidy for families already occupying existing private school seats than as a mechanism for public-to-private enrollment shifts.

The 25-30% who leave public schools are the direct threat to districts. Each exit costs AISD roughly $8,300 in lost entitlement through the funding formula. That’s the enrollment cliff modeled in the previous section: fewer students means less revenue, but costs don’t drop at the same rate. This is a dollar-for-dollar transfer from public schools to private ones.

The distinction matters. The larger group commits billions in state revenue to subsidizing families who were already paying their own way, revenue that could otherwise go toward teacher pay, facilities, or the funding formula itself. The smaller group defunds districts directly.

Texas’s 75/25 split means even less of the program is doing what proponents advertise (giving public school families a choice) and more is subsidizing families who already had one. Iowa is further along the curve: only 6.8% of voucher recipients in 2024-25 actually transferred from a public school. The pattern is consistent across every state that has tried universal vouchers.

There’s one more financial layer Florida literally cannot answer yet: what happens to private school tuition after vouchers arrive?

In higher education, federal aid inflates tuition by as much as 60 cents per dollar (per the Texas Public Policy Foundation’s own analysis). The logic applies directly: if parents were already willing to pay $12,000 a year, and a $10,000 voucher shows up, the school raises tuition to $22,000. The voucher becomes a transfer to the institution, not a savings for the family.

With top-tier Austin private schools (tuition upwards of $30,000) largely operating at capacity or opting out of TEFA entirely, the market segment most exposed to this dynamic is mid-tier and parochial schools charging $9,000 to $15,000, right in the range where the ~$10,000 voucher closely matches current tuition. These are the institutions most susceptible to the Bennett Hypothesis: the voucher doesn’t expand access, it raises the floor.

We can’t measure this in Florida because the state doesn’t require private schools to report what they charge. But states where researchers built their own datasets tell a clear story. A Princeton study of Iowa’s voucher program found kindergarten tuition jumped 21-25% after vouchers launched, while pre-K students (ineligible for vouchers) saw zero increase at the same schools. A Tulane study across 11 voucher states found 5-10% more tuition inflation than in non-voucher states. In Arizona, the Hechinger Report found nearly all of 55 surveyed private schools raised prices after ESA expansion, with half increasing tuition by 10% or more.

Follow the money one more step. If tuition inflates to absorb the voucher, families already paying for private school get a $10,000 taxpayer subsidy while low-income families still can’t afford the gap between the voucher and the new sticker price. School closures concentrate in low-income, majority-minority neighborhoods. The families who can’t navigate the application process, can’t provide transportation across town, or whose children have needs that private schools can decline to serve, are left in a system with fewer resources per student.

The feedback loop

Here’s how the pieces connect:

1. Universal vouchers pass. The state begins paying ~$10,000 per student for private school attendance.
2. ~70% of voucher dollars go to families already in private school, stretching the finite state education budget.
3. Public schools lose students (the 30%) AND funding. The LBB projects $805 million in reduced public school funding by 2029-30.
4. With less funding, public schools cut programs, increase class sizes, and close campuses.
5. Degraded public schools push more families toward vouchers. “Look, public schools are failing” becomes the justification for further expansion.
6. The only system being publicly graded (STAAR, A-F ratings) is the one being defunded. The degradation of public schools is visible. The performance of voucher schools is not.
7. Repeat.

This is why the enrollment and deficit projections earlier in this piece are conservative. They assume a constant exit rate based on Florida’s first two years. If the cycle above takes hold, exits accelerate as the system degrades, and the curve steepens beyond what a linear model captures.

Florida’s kindergarten numbers are the clearest warning sign. When kindergarten enrollment drops, that smaller cohort doesn’t just affect one year. It rolls forward: fewer first graders the next year, fewer second graders the year after that, and so on through the entire pipeline. A kindergarten dip today compounds into a systemwide enrollment decline over the next decade. Florida is already showing this pattern, and Texas should expect the same trajectory once TEFA is fully operational.

You can argue about whether this cycle is an unintended consequence or a feature. The math doesn’t care.

For Austin ISD, there’s an additional layer that makes the cycle worse than anything in Florida: recapture. AISD sends $821 million a year in local property taxes to the state through the Foundation School Fund. Each recaptured dollar offsets a dollar the state would otherwise spend from general revenue, freeing that revenue for other purposes, including TEFA vouchers at private schools with no STAAR requirement, no public reporting of results, and no obligation to accept all students.

The most concrete illustration: in November 2024, Austin voters passed Proposition A, explicitly raising their own property taxes to fund Austin ISD. 76% of those new Prop A dollars go straight to recapture. Those recaptured dollars flow into the Foundation School Fund, offsetting the state’s own contribution and freeing general revenue for other uses. TEFA is funded from that same general revenue. The connection isn’t an earmark; it’s a substitution effect. Every recaptured dollar from Austin property taxes is a dollar the state doesn’t have to spend from its own funds, and those freed-up funds now include voucher payments to private schools statewide, while several of Austin’s own top-tier private schools are opting out of the program entirely and Austin ISD is simultaneously closing campuses. That is not what Austin voted for.

Florida’s test scores tell an incomplete story

Given all that financial upheaval, how are Florida’s public school students actually performing? Better than you’d expect.

Florida’s proficiency rates on the FAST (Florida Assessment of Student Thinking, the state standardized test that replaced the older FSA in 2023) have held steady or improved:

School choice proponents point to the FAST-era improvement as evidence of “competitive effects.” A 2026 synthesis from the American Federation for Children found gains equivalent to 120 additional days of learning over 15 years in competitive markets.

But the comparison is fundamentally broken. Voucher students aren’t taking the FAST. They’re taking tests of the school’s choosing (MAP, Stanford-10, etc.) that measure different things on different scales. Florida can show you how public school students are doing. It cannot show you whether voucher students are doing better, worse, or the same, because the system was built to make that comparison impossible. Iowa requires voucher students to take the same state assessment as public school students, proving this is a policy choice, not an inevitability.

There’s also a timing problem. Two years isn’t long enough to see the full impact of school closures, staff layoffs, and demographic concentration on test scores. The effects are structural, and they lag. Broward is evaluating 34 schools for closure right now. Duval is cutting 700 positions. Those changes haven’t fully filtered into the classrooms yet, and the students most affected are disproportionately in the earliest grades. Go back to the kindergarten data: 18,417 fewer kindergartners in three years, kids who never entered the public system. Those children won’t hit FAST testing grades (3rd grade) until 2027 at the earliest. The cohorts currently being tested went through a pre-HB 1 system that was still fully staffed and fully funded. The cohorts shaped by closures, program cuts, and larger class sizes are still coming through the pipeline. Flat scores today don’t tell you what scores will look like in three to five years, once the compounding effects of annual enrollment loss, reduced funding, and campus consolidation reach the tested grades.

The longer-term research is less encouraging. Brookings found large-scale voucher programs in Indiana, Louisiana, and D.C. produced negative impacts on math and reading achievement, with effect sizes actually more severe than COVID-era learning loss. A Florida State University evaluation of Florida’s own scholarship program found no significant test score gains, and students who used vouchers but returned to public school performed worse than peers who never left.

One more context point on the school grades data: Florida had 1,181 A-rated schools in 2018-19 and 1,299 in 2023-24. The shift is overwhelmingly explained by the FSA-to-FAST assessment transition (a new test with new cut scores), not by voucher effects. We mention it only because you’ll see it cited in the debate, and it’s important to know what’s actually driving the numbers.

What the data does and doesn’t prove

What the data shows. Florida’s voucher costs landed where independent analysts predicted, not where legislative proponents said they would. Enrollment losses accelerated in urban districts, the accountability gap is structural, and the primary beneficiaries are families already in private education. The LBB’s own projections show Texas on the same cost trajectory.

The strongest arguments for vouchers, honestly engaged:

Florida’s public school test scores held steady or improved after HB 1. Proponents attribute this to competitive pressure. But the survivorship bias explanation (struggling students leave, the average rises) doesn’t hold up against the program’s own demographics: the 70/30 data shows voucher recipients skew toward wealthier families who were already in private school or who can navigate the application process and cover tuition gaps. The students leaving public schools are, on average, from higher-income households with more engaged parents. If anything, losing those students should lower averages, not raise them. The fact that scores held steady despite that compositional shift is arguably the more notable finding.

Parental satisfaction surveys consistently show voucher families are happier with their children’s education. But satisfaction data has a selection problem: it only captures families who successfully used the voucher. It doesn’t capture the family earning $45,000 a year who looked at the voucher, looked at the actual cost of private school after tuition hikes, and stayed in a public system that now has less funding per student because of the program.

Some children were in chronically failing schools. For those families, a voucher may be the best available option right now. The Urban Institute found some voucher programs produced meaningful improvements in college enrollment rates, though not in degree attainment. But SB 2 prioritizes 80% of slots for low-income students and students with disabilities, and priority access doesn’t solve the math: if the voucher covers $10,000 and tuition is $18,000, a low-income family still can’t afford the gap. For students with disabilities, there’s an additional barrier: private schools are not bound by the federal Individuals with Disabilities Education Act (IDEA) and legally retain the discretion to deny admission based on a student’s special education needs. The state can prioritize these students for voucher funding, but it cannot compel the private market to accept them. The result is that public schools will continue to serve the highest concentration of students requiring specialized, cost-intensive services, even as per-student funding falls with each enrollment loss. The students most likely to benefit from leaving a failing school are the least likely to have the resources, or the admissions pathway, to actually use the voucher.

What the data doesn’t prove. That every voucher student is worse off, that public schools would have spent the money better, or that the competitive effect isn’t real. We genuinely can’t know, because the system was designed not to measure it. The strongest defense of vouchers (“competition improves everyone”) can’t be evaluated because the accountability infrastructure was deliberately excluded from the law.

For Austin ISD: the district was already in a structural funding crisis before SB 2. Recapture, a frozen allotment, declining enrollment. Vouchers add another draw on the same constrained pool, and 274,000 applications in the first window suggest it won’t be a rounding error.

Why this article exists

I started pulling on this thread because the AISD article ended with a “wait and see” on SB 2, and then 274,000 Texas families applied for TEFA vouchers in the first window.

This article was different from the others on this platform. I went looking for voucher outcome data and hit a wall. Florida doesn’t publish voucher student test scores, doesn’t track private school tuition, and the Auditor General was denied access to scholarship organization records. FLDOE stopped publishing annual reports on participating private schools after 2022-23. There is no dataset that would let you compare voucher student outcomes to public school outcomes in Florida, because the data was never collected.

That gap is the story. When public schools underperform, you know immediately: STAAR scores, A-F ratings, intervention reports. When $4 billion a year flows to private schools through vouchers, you get nothing. Iowa requires voucher students to take the same state test, so measurement clearly isn’t the obstacle. Florida and Texas chose not to.

The Florida data that does exist was scattered across state agencies and advocacy group analyses on both sides. We loaded the FLDOE data into OpenData so the comparison could at least be made with the numbers available. Every chart above links to its source data. If you disagree with a conclusion, you can run the same query and check.

SB 2 reports voucher test scores to the Comptroller, not TEA, and doesn’t require school-level publication. In three years, someone will try to write this same article about Texas, and they’ll hit the same wall I did with Florida. The reporting structure is already law.