November 19, 2026
Since Arizona opened its Empowerment Scholarship Account (ESA) program to every K-12 student in 2022, no line of attack has traveled farther or proven more durable than the charge that the program is a fiscal disaster. It reaches policymakers in other states as a few claims, usually delivered together: that Arizona’s ESA program is an unsustainable, billion-dollar-and-growing cost threatening the State budget, it is unaccountable and riddled with high-rates of misspending and fraud, and that funding ESA students is expensive and drains resources from public schools. All are typically presented as settled facts, though none survives contact with the data. This report – the latest in an ongoing Common Sense Institute series that will be fact-checking claims about the state’s program since universal eligibility – examines these arguments using real, public budget and program data.
The short version: the billion-dollar figure is real, but only about half of it comes from the 2022 universal expansion critics object to. The rest funds disability and other categories the program has served since 2011. The ESA program does not explain the state’s budget woes, and on a per-pupil funding basis universal ESAs represent the lowest taxpayer funded education option in the state. Additionally, the waste, fraud, and abuse statistics that critics cite are just plain false; misspending within the program is exceedingly low, especially when compared to other government programs.
Claim #1 — Arizona’s unsustainable ESA program costs over $1 billion
THE CLAIM: “We’ve got a spending problem when it comes to vouchers, … We’ve got a billion dollars that’s completely unaccounted for, and we’ve got public schools that are closing.” – Beth Lewis (Save our Schools, 2026)
The total $1 billion cost is accurate. The conclusions drawn from it are not. As the figure below shows, only a little more than half (about $600 million) of program cost comes from the universal expansion, even though universal students are three-quarters of enrollment – meaning universally eligible recipients represent the lowest cost beneficiaries on the program, and despite what critics claim, the ESA program did not cause the state’s budget woes.
The Facts
- The total cost figure is right. As of May 2026, over 102,000 students participate (10% of the K-12 population) and annual scholarship awards this year will total over $1 billion. Both the total enrollment and total spending figures include recipients participating in one of eleven eligibility categories, including the universal K-12 category.
- Nearly a third (29%) of ESA participants still enroll under the non-universal “legacy” categories, and this group accounts for almost half of all program spending. The universal expansion critics target is on the order of $580 million, not the full billion-plus.
- Universal expansion did not cause Arizona’s budget woes. Between FY2017 and FY2022 the state General Fund accumulated over $4.5 billion in cash surplus; more than $2 billion remained even after universal expansion.[1] The decision to spend another $2.3 billion on 316 line-items (averaging just $7.4 million apiece) in the FY2024 budget was the ultimate catalyst.
- The new spending that followed dwarfed the program. In two years (FY2023–FY2024), policymakers raised annual General Fund spending by nearly $5 billion (+39%) and committed another $2.5 billion to tax and revenue reductions. The universal ESA program was about 6% of that $7.5 billion in new commitments. Had growth been held near 4.4% per year, CSI estimated the state would still hold a roughly $5 billion surplus even with ESA’s fully funded.[2]
- Tightening state budgets is national and tracks the end of pandemic good-times — not new ESAs in Arizona. As federal pandemic aid expired and revenues normalized after a roughly 50% surge, budgets tightened nationwide. Heading into 2026, about ten states were flagged with a “challenging” fiscal outlook and another thirteen as “conditional”, and the states most cited for shortfalls do not have universal ESAs. The roughly 15 states that had adopted universal or near-universal eligibility by early 2026 are not over-represented among the distressed.[3]
- Arizona’s K-12 system is running into a demographic ceiling. Arizona’s publicly funded K-12 system is essentially saturated: adding district, charter, and ESA enrollment together, the total number of publicly funded students today (1.16 million) is about what state analysts projected way back in January 2020. Even with universal ESA, the state is paying for no more students than it expected – all that’s changed is how we are paying for them (a combination of District, Charter, and now-universal ESA).
Bottom line: The billion is real, but about half is the contested 2022 expansion and it is the cheapest funding channel per pupil the state operates. With the publicly funded system near demographic saturation and the school-age population declining, and since future growth is increasingly switchers from other public options who add little or no new public cost, continued growth is approximately cost-neutral.
Claim #2 — The ESA program is unaccountable and riddled with misspending and fraud
THE CLAIM: “A very big story from my colleague Craig Harris reveals an audit of a very large sample of ESA purchases that shows just over 20% are considered unallowable. 20%. That’s higher than any previous official estimates.” — Joe Dana, 12News (February 2026)
Since the creation of universal ESAs, critics have tried to convince the public that the program is riddled with fraud, waste, and abuse. They reach these conclusions by citing ill-conceived statistics calculated from risk-based samples that are not representative of the ESA program. The fact is, when statistical calculations and sampling are done properly, we find that unallowable spending, and more specifically fraud, is incredibly rare within the ESA program and well below the improper spending/reimbursement rates for many other government programs.
The Facts
- The “20% unallowable” claim from 12News and others relies on a risk-based audit sample, not a statistically representative sample of ESA users or ESA-funded purchases.[4] [5] Risk-based auditing is a standard practice among federal and state agencies, allowing investigators to concentrate on transactions or accounts most likely to involve unallowable spending, or those already flagged as potentially unallowable. It is not designed to support population-level inference.
- The 20% figure describes a share of transactions or items purchased – not dollars spent – a distinction that often gets lost when the finding is repeated. For example, there are three ways ESA recipients can make purchases through the program, and the Marketplace option accounts for both the largest share of ESA transactions, and incidentally, the largest share of transactions flagged as potentially unallowable. However, these purchases are typically very small in nature, and even if we assume that 20% of them are truly unallowable purchases, that would only translate to about 2.4% of total program spending.
- Actual unallowable spending is less than 2% of total program spending, and outright fraud is even less. Using a true random sample that is statistically representative of the entire ESA program, ADE found only 6.4% of items purchased were unallowable, translating to 9% of all spending. Only 0.3% of total spending went towards items that were considered “egregious” misspending.[6]
- Moreover, accounts found to have made truly unallowable purchases are suspended, and those amounts must be repaid.[7] Those found to knowingly misuse ESA funds or commit fraud are removed from the program and can be referred to the Attorney General for criminal investigation, demonstrating the existence of the very guardrails critics claim don’t exist.
- The unallowable spending rate within the ESA program is well below the rates for other publicly funded programs. The improper payment rate for Unemployment Insurance was 14.4% in 2024 – over seven-and-a-half times the rate of unallowable spending within the ESA program. Even the improper payment rate within Arizona’s own SNAP program is nearly six times higher than ESAs.[8] [9] [10]
- While critics remain hyper-focused on the relatively low unallowable spending and even lower fraud levels within the ESA program, they overlook significant misspending within the state’s district school system. An analysis from CSI found 41 cases of waste, misspending, or outright fraud across the state’s public-school districts since 2010, for a total cost to taxpayers of $81.3 million. Of those 41 identified instances, 31 involved some sort of criminal proceedings.[11]
Bottom line: The insinuation that 20% of ESA program expenditures are fraudulent is simply false. Less than 2% of program expenditures go to unallowable purchases, and those transactions that could reasonably be categorized as fraud are even more rare – 0.3%. This rate is far lower than most government programs, and the focus on ESAs overlooks the millions in actual fraud committed within the state’s district schools over the last 15 years.
Claim #3 — Funding ESAs is expensive and drains resources from public schools
THE CLAIM: “It’s just an attack on public education. Taking money out of the system, draining resources, weakening our schools.” — Marisol Garcia (President of the Arizona Education Association, February 2024)
Despite claims that ESAs are “draining” the public system, Arizona’s district schools have seen funding rise, not fall: total K-12 funding grew about 60% between FY 2016 and FY 2025, with the ESA program accounting for only a small portion of that increase. The rest went directly to traditional public schools even as their enrollment declined. Universal ESA recipients also receive less taxpayer funding than students enrolled in the public system, so a family’s decision to switch doesn’t add net cost; if anything, it frees up resources. Meanwhile, the enrollment declines critics point to predate universal ESA by years. The state shed 40,000 public students overnight in 2020-2021 and most never returned. Last school year alone (2025–2026), Arizona’s district classrooms lost another 30,000 students, the largest single-year decline since 2020–2021 and a continuation of a trend that began almost two decades ago.
The Facts
- District enrollment has been declining, and this predates universal ESA by years. District enrollment peaked around 2008 and has fallen since. The single largest drop – roughly 40,000 students – came in 2021, before universal eligibility took effect in 2022. Today there are about 75,000 fewer students in district classrooms than in 2019–2020, the share of Arizona children in district schools has slipped from about 80% a decade ago to under 70%, and the school-age population is now shrinking. ESAs followed this trend; they did not create it.[12]
- Despite declining enrollment, the district system is receiving more funding, not less. Total K-12 funding rose about 60% (+$6.3 billion) between FY2016 and FY2025; the ESA program was only about 11% of that new money. The rest went straight to traditional public, mostly district, schools despite their falling enrollments. On an inflation-adjusted basis, Arizona now spends roughly 30% more per public-school pupil than a decade ago.
- The ESA program itself has no direct bearing on traditional public K-12 funding. Arizona’s education funding formulas are driven by enrollment, so the mere existence of universal ESAs does nothing to divert dollars from district schools, absent an actual decline in enrollment. That decline still requires an explicit choice by a family to leave the traditional public system. And as noted above, declining district enrollment was already a growing trend before universal ESA eligibility existed.
- ESA students cost less per pupil than the public-school seats they replace. A typical universal-eligibility ESA costs taxpayers an estimated $8,285. These are students without special needs who are eligible under the 2022 law, many of whom left traditional schools relatively recently. Even accounting for ESA students with special needs, the average ESA across all recipients is a little under $11,000 — well below the roughly $15,000 an average district student costs, of which $13,300 are direct state and local taxpayer funds.
- Demographic shifts mean future growth is increasingly switcher-driven and close to cost-neutral. With few new entrants left, the marginal new ESA user is increasingly a student moving out of a public school; ADE reports that even now roughly 60% of new universal grade 1-12 ESA recipients attended a public school immediately before enrolling. And a switch is, on balance, a savings for taxpayers.
- CSI estimates switchers on average could free up a combined $6,985 in resources. Of this amount, we estimate an average net immediate savings of about $95 per switcher, plus roughly $6,890 per student in other “non-formula” public K-12 funding freed for longer-term savings or reallocation. Additional ESA enrollment adds little or no net new cost; policymakers should instead be watching the entire K-12 landscape.[13]
- Even in edge cases (less than 5% of students) where the state General Fund doesn’t capture the savings, local taxpayers do. In so-called non-state-aid districts, local property taxes fill the entire equalization formula bucket. But in this case, if a student switches to ESA, the new General Fund cost is fully offset by an immediate taxpayer savings (local property tax decrease). This isn’t speculative or hypothetical; it’s automatic and driven by formula.
Bottom line: ESA students cost less on a per-pupil basis than public school pupils, and the state is spending more per public-school student than ever despite declining enrollment. Moreover, A student who does move from the public system to an ESA is roughly cost-neutral to the State General Fund in the near term and a net savings over time. Policy makers concerned with the health of district schools should focus less on the existence of a particular program and more on the underlying factors contributing to the enrollment declines.
References
[1] Common Sense Institute, ESA Facts & Fictions (October 2024).
[2] Common Sense Institute, ESA Facts & Fictions (October 2024).
[3] Morgan Scarboro, “Ten States Face Budget Deficits Going into 2026,” MultiState, January 6, 2026; universal eligibility counts from Ballotpedia, “States with and without universal school choice programs” (February 2026) and EdChoice (2026).
[4] Reany, Jack, “The 20% Myth: Debunking Arizona Empowerment Scholarship Account (ESA) Misspending Claims,” Substack, March 29, 2026.
[5] Beienburg, Matt, “12News’ and Ivy League Expert’s False ESA Claims Collapes – Retraction Required,” Goldwater Institute, March 20, 2026.
[6] ADE’s ESA Study Data FY 2026
[7] Parent Handbook, Empowerment Scholarship Account Program School Year 2025-2026
[8] Centers for Medicare & Medicaid Services Fiscal Year 2024 Improper Payments Fact Sheet
[9] Office of Inspector General (OIG) Oversight of the Unemployment Insurance Program.
[10] USDA FY 2025 National and State Payment Error Rates
[11] Based on CSI analysis of Auditor General reports and news articles.
[12] Common Sense Institute, ESA Facts & Fictions (October 2024) and “Education Choice, Access & Transparency” (2026), citing ADE enrollment data and Census/State Demographer figures.
[13] Common Sense Institute, “Education Choice, Access & Transparency” (2026), “Net Fiscal Consequences of Arizona ESA Switchers,” drawing on the JLBC Overview of K-12 Funding, JLBC FY26 Appropriations Report, and ADE ESA Quarterly Report.