Did Arizona’s ESA Expansion “Blow a Hole” in the Budget?

View Post

No. A closer look at state data and student switchers reveals that the ESA program’s net costs were shrinking by Year 2.

Choice opponents and some media reports argue that Arizona created a “budget meltdown” when the state expanded its education savings account (ESA) program to all K-12 students in the state. Did the program, as one headline claimed, “[blow] a massive hole in Arizona’s budget”?

In this Informed Choice installment, I present a fiscal analysis of Arizona’s ESA program during its first two years. The analysis isolates the cost of expanding the ESA program to all K-12 students in the state. It differs from previous analyses by incorporating data from the Arizona Department of Revenue on the state’s tax-credit scholarship (TCS) programs to generate a more complete picture of switchers and, therefore, the savings that offset program costs. It also differs from an analysis by the Common Sense Institute, which analyzes the program as a whole, projecting that the program’s total cost will reach $1 billion in 2026 while state spending will be about $800 million less overall on district schools than projected due to broader enrollment declines across Arizona’s K–12 system.

I estimate that the net cost (total cost minus state savings) for ESA universal expansion was $178.3 million in the first year and $118.0 million in the second year. ESA students likely diverted from public school systems generated an estimated $146.6 million in state savings in the first year, offsetting 45% of the total cost. Switchers generated state savings which offset 71% of the second year’s total cost for universal expansion.

The results from this analysis, coupled with other rigorous fiscal analyses of the program, contradicts rhetoric that the program is a “budget-busting free-for-all” or fiscally unsustainable.

Program Cost

I do not use the program’s total cost as reported in Arizona Department of Education (DOE) reports because that would overstate the cost of expansion as more than 12,000 students were already in the program prior to universal expansion. Because the DOE reports don’t isolate the cost of universal expansion only, I must estimate that cost.

We know the number of ESA students who enrolled under universal expansion and can estimate that expansion’s cost by multiplying this number by an average ESA award amount. The mean ESA cost for all students in the program’s first year was $9,523. This measure overstates the true cost of universal expansion, however, because it includes many higher-cost students outside the universal pathway. Prior to universal expansion, 60% of ESA students had special needs.

A less biased measure is the median ESA amount because it better reflects the typical award and excludes the disproportionate influence of higher-end amounts for students with disabilities. The Arizona DOE reports that the median ESA amount was $7,194 for FY 2023 and $7,409 for FY 2024.

I estimate the cost of universal expansion in years 1 and 2 as follows:

Year 1: $7,194 x 45,157 universal ESA students = $324.9 million

Year 2: $7,409 x 55,590 universal ESA students = $411.9 million

It’s important to note that this price tag is not the same as a net fiscal impact, as these cost estimates do not include offsetting savings from switchers.

Switchers

Switchers are students who would likely enroll in public schools without financial assistance from a school choice program. These students represent savings for state taxpayers because the state no longer must pay for their public education.

To estimate switchers, I combine data from quarterly ESA reports published by the Arizona DOE and annual TCS reports published by the Arizona Department of Revenue.

The ESA quarterly reports include both the number of universal-eligibility ESA students enrolling in the program for the first time that year and the number of those students “attending public school immediately before ESA enrollment.” While the latter provides a proxy for switchers, it significantly underestimates the true total because Arizona operates four TCS programs, two which have strong prior public school enrollment requirements.

In other words, thousands of ESA students were in private schools immediately before ESA enrollment because they were participating in a TCS program at that time.

The fiscally relevant question is: Of those ESA students in a private school immediately before ESA enrollment, how many were there because of a TCS, and how many had switched from public schools earlier?

A previous Informed Choice post provides a more detailed analysis of switcher estimates during the first two years of universal expansion. The spike in ESA participation in Year 1 coincided with a 22,350 decrease in the number of tax-credit scholarships (see chart below), providing strong evidence of significant migration from TCS programs to the ESA program.

Chart Sources: Arizona Department of Revenue, School Tuition Organization Income Tax Credits in Arizona Fiscal Year 2024, March 2025, https://azdor.gov/sites/default/files/document/REPORTS_CREDITS_2025_fy2024-private-school-tuition-org-credit-report.pdf; Arizona Empowerment Scholarship Account (ESA) quarterly reports from the Arizona Department of Education, https://www.azed.gov/esa/esa-quarterly-reports

Of these 22,350 fewer tax-credit scholarships in FY 2023, 12,181 came from two programs with prior public school enrollment requirements. These students are not included in the DOE’s count of ESA students switching directly from public schools. Adding these 12,181 TCS students to the 6,157 ESA students the DOE reports as having attended public school “immediately before ESA enrollment” yields an estimated 18,338 switchers in year 1.

Repeating this method for FY 2024 yields 17,357 new switchers. Adding this to the prior year’s 18,338 gives an estimated 35,695 total switchers through FY 2024.

These estimates likely understate the true number of switchers because they assume zero switchers from the TCS programs without switcher requirements. Given that scholarship granting organizations have incentives to award scholarships to students who will fill empty seats in private schools (i.e., students switching from public schools), this assumption adds an additional layer of caution to the analysis.

To recap:

Year 1: 18,338 total switchers (new ESA students)

Year 2: 35,695 total switchers (18,338 from Year 1 + 17,357 new switchers in Year 2)

State Savings

The ESA amount is set at 90% of the state’s per-pupil base funding plus any additional assistance that would apply if the student were enrolled in a charter school.

Using the median ESA amount, the average state cost for students under universal expansion is $7,993 (= $7,194 / 0.9) for FY 2023 and $8,232 (= $7,409 / 0.9) for FY 2024.

Therefore, each student switcher under universal expansion generated state savings of $799 in year 1 ($7,993 — $7,194) and $823 in year 2 ($8,232 — $7,409).

Year 1: $7,993 × 18,338 switchers = $146.6 million reduction in state revenue for public schools

Year 2: $8,232 × 35,695 switchers = $293.8 million reduction in state revenue for public schools

Net Fiscal Impact of Choice Program

After accounting for state savings from switchers, the estimated net fiscal impacts (NFI) on the state budget from universal expansion are:

Year 1 NFI: $324.9 million (cost) – $146.6 million (savings) = $178.3 million net cost

Year 2 NFI: $411.9 million (cost) – $293.8 million (savings) = $118.0 million net cost

Chart Note and Source: Author’s calculations using data from the Arizona Department of Revenue and Arizona Department of Education

Conclusion

The net fiscal impact from expanding Arizona’s ESA to all K-12 students in the state decreased in the program’s second year as the share of ESA students who are switchers increased. This has led to a greater portion of program costs being offset by state savings. These net costs of $178 million in year 1 and $118 million in year 2 represent just 1.1% and 0.7% of total K-12 funding from all sources (local, state, and federal) for Arizona public schools, respectively. They also represent 0.2% and 0.1% of total state expenditures on all public services during those years.

Arizona’s experience demonstrates that expanding educational opportunity to every family can be done responsibly. The data show manageable fiscal effects alongside substantial gains in parental choice and flexibility. While the debate over universal choice will no doubt continue, the evidence suggests that Arizona’s program has not imposed anything close to the fiscal strain that critics predicted.

This was originally published to our Substack.