Editorial Note: This article and its series are part of a collaborative project with 13 Arizona newsrooms covering two important topics in education. Funded by the Arizona Local News Foundation and Report for America, all stories in the project can be found here.
Sitting in the KAWC studio, retired Yuma School District One Superintendent Jamie Sheldahl paints a picture.
Imagine a public school with 20 class sections and 30 kids in each section. For easy math, assume the school gets $10,000 per kid.
If three kids per section leave that school using an Empowerment Scholarship Account (ESA), Sheldahl says, that’s 60 kids out and the school district’s lost $600,000 in revenue — but not enough revenue to reduce any kind of expenditures.
“You still need a teacher for 27 kids, just like you needed a teacher for 30 kids,” he said. “It's not enough to reduce your custodial staff. It's not going to cut into your busing.”
In Sheldahl’s experience, this is how ESAs affect public schools at the local level.
And that example illustrates why one of the most common claims in Arizona’s ESA debate — that an ESA costs roughly $7,000 while a public school student costs considerably more — is much harder to evaluate than it sounds. Because savings don’t occur smoothly, one student at a time. They occur in chunks, which makes the question of what one student actually “costs” more complicated than it first appears.
So What Does A Student Actually Cost?
Well, it depends. If you look at the Arizona Auditor General’s site, district level reporting breaks down what districts actually spend across different areas.
These include instruction, student support, instruction support, administration, operations, food service and transportation.
Summing these together and dividing by a district’s enrollment produces an average per-student spending figure. In Arizona, school enrollment is commonly measured through average daily membership, or ADM. In these tables below, the Arizona Auditor General breaks down how much Yuma and La Paz districts spent per student across different areas in Fiscal Year (FY) 2024-2025.
The statewide average total, for example, is the figure ESA proponents like Matt Beienburg, director of education policy at the Goldwater Institute, use when they compare per-pupil spending.
“The fairest way, I think, to look at this — I mean, I guess there are a couple fair ways; one is, ‘What is the total cost per kid?’” he said.
By comparing this to the average ESA award amounts, Beienburg concludes ESAs cost taxpayers less per student — an important contention in the ESA debate.
“I cannot stress enough that a kid who is in the ESA program costs taxpayers less than it would if that kid were in the public district school system,” he said. “Where those funds come from could change; that's true. It could be more from property taxes. It could be more from the General Fund. That is true. But it's the same model that applies to public charter schools, and it is less money per student.”
According to the Arizona Department of Education (ADE), the most recent average ESA award amount for non-disabled students is $7,500. For students with disabilities, the average award amount is $10,800. But award amounts do vary as seen in the figure below.
On the face of it, average ESA award amounts are less than the roughly $16,000 Arizona public school districts spend per student statewide. But public school officials say there’s an important distinction to be made: that figure includes federal, state and local money.
“It's important to really understand what goes into that number,” said Denis Ponder, Yuma School District One superintendent. “And so when we're talking about the funding and whether it may detract from or provide challenges for public schools and ESAs and charter schools, it's important to note that we're specifically talking about the General Fund.”
“ESAs, like charter schools, are 100% dependent on the General Fund,” added Dale Ponder, Crane Elementary School District superintendent. “There's no taxation, so there's no tax population where they can assess a property tax in order to fund a portion of that. Whereas school districts — some school districts have 100% of their needs met by the local property taxpayer and others have smaller percentages as they go.”
Since ESAs draw from the Arizona General Fund, folks often turn to state revenue for an “apples-to-apples” comparison when the question turns to savings.
Do ESAs Save the State Money?
This is where public schools distinguish where their money comes from. If an individual points to the $16,000 per-pupil average to argue that the state saves money spending $7,500 per-pupil, the problem becomes that the state itself isn’t spending every dollar from that $16,000.
As Beienburg explained, “the way the funding formula works in Arizona is it basically has a hybrid system where districts have a formula that says, ‘Here’s how much you’re going to get, and you’re going to raise the first part of that from local property taxes,’ and then the state kicks in on top of that.”
This is why some districts might receive zero dollars in state equalization assistance as Dale Ponder noted.
In this table below, we examined the ADE’s local education agency reports to get a better sense of how much Yuma and La Paz public schools draw from the General Fund.
That state contribution is called equalization assistance. It is meant to help fill the gap between what a district’s formula says it should receive and what its local property tax base contributes.
In the table above, you can see each district's unweighted ADM. That’s essentially a district’s total enrollment as calculated by average daily membership. Because the state has identified some student populations as requiring more support, they weigh some students differently.
For example, students with disabilities, English language learners and kindergarten students might require more support and resources, so they’re given a multiplier. This results in a weighted ADM.
Weighted ADM is multiplied by the state’s base level amount which, for FY 2026, is $5,113.26.
Transportation and district additional assistance, which accounts for capital needs, are also added to the base funding amount.
Once the district’s local property-tax contribution is subtracted from its equalization base, the remaining amount is its state equalization assistance. Divide that by the unweighted ADM, and we can see how much districts draw from the General Fund per student.
Through this method, we can see that Yuma County public school students draw about $6,500 from the state’s General Fund whereas La Paz County public school students draw about $3,200 on average.
The state also funds public charter schools. These schools don’t receive local property-tax revenue, so their equalization assistance is funded by the state and consists of base support funding plus charter additional assistance.
But is this a clean apples-to-apples comparison? Beienburg argues it isn’t. One example he gave is the Classroom Site Fund (CSF), a state sales tax fund that provides additional money for things like teacher pay, class-size reduction and student support services. For FY 2026, the tax rate is 0.6%, and ADE lists the distribution at $842 per weighted pupil.
“We all pay for it. Every time, virtually any purchase you make, there's a six-tenths of a cent sales tax, and it goes into a big pot of money that goes to Arizona public schools, and currently that kicks out about a thousand dollars for every kid,” he said. “That's state tax money that goes to every public school kid, and zero dollars of that go to each ESA kid.”
But including that money doesn’t necessarily tell us how much the state saves when a student leaves a public school. Unlike equalization assistance, CSF is a dedicated pool distributed among public schools based on weighted enrollment. If one student leaves, the state doesn’t necessarily stop spending that student’s proportional share of the fund; those dollars may instead be redistributed among the remaining public school students.
Denis Ponder also argues that CSF is different from ordinary General Fund spending because voters created it as a dedicated revenue stream for specific public school purposes. He compared it to a local override — money that exists because taxpayers approved an additional tax for designated uses.
These comparisons answer two different questions: how much public tax revenue supports a student and how much state spending actually changes when that student leaves.
One might ask if there’s any merit to adding the local property-tax contribution back in to get an idea of how much state and local tax money supports students through the district funding formula. In the table below, we added the local levy to equalization assistance and divided that amount by unweighted ADM. We also divided the local levy by unweighted ADM to show how much of that formula funding comes from local property taxes.
Together, these figures provide a fuller picture of the state and local tax dollars supporting the formula. But it’s worth noting the two sources work differently. State funding draws from a statewide tax base, so money supporting a Yuma County student, for example, may come from taxpayers elsewhere in Arizona, like Scottsdale or Flagstaff for instance. Local property-tax funding, by contrast, is raised within the district.Furthermore, one student leaving doesn’t mean property owners in that district immediately pay one student’s share less in taxes.
So that leaves perhaps the central question: what happens to state spending when a student actually moves from one education system to another?
What Happens When a Student Switches?
Whether an ESA saves the state money depends in large part on where the student was educated before entering the program.
A student leaving a district school, a student leaving a charter school and a student who was already in private school or being homeschooled do not have the same effect on the state budget.
In 2023, the Joint Legislative Budget Committee actually tried to illustrate that difference by comparing the formula costs of district schools, charter schools and ESAs. Their analysis can be found here, but we’ll take a look at some of the findings.
In these tables below, JLBC compared the state’s formula costs for hypothetical pupils in grades 1-8 and high school. If you recall, weighted ADM takes special populations into account, including kindergarten.
Those are known as Group B weights, so these tables show the formula cost for a pupil who does not qualify for any additional Group B funding, in both districts that need state aid and those that do not.
In that comparison, JLBC found that ESAs draw more from the state’s formula than public school districts but less than public charters.
As previously discussed, some populations might require more support than others, so JLBC conducted a comparison for pupils with moderate intellectual disabilities.
Here, they found ESAs cost less than both district and charter schools:
The cost differences are notable, however, and more detailed analysis can be found in an older 2021 fiscal analysis for Senate Bill 1452, which proposed expanding ESA eligibility to children who are eligible for Free and Reduced Price Lunches as well as children receiving Title I services.
That analysis projected General Fund savings in the first two fiscal years before costs emerged in the third. Beyond that, they found the impact would be more speculative as it would become more difficult to estimate the number of new entrants to the program.
It estimated that each non-disabled child shifting enrollment to the ESA program from a public school would generate $473 of state costs at the time if the child was previously enrolled in a school district — unless that was a non-state aid district. If the district wasn’t taking state aid, that switch would generate $6,844 of state costs for that time.
If that same child was previously enrolled in a charter school, however, the switch would generate a state savings of $757 at the time. This generally tracks with the 2023 analysis but with different numbers.
That analysis identified another important consideration, however: additional costs for students who never enrolled or intended to enroll in a public school.
“In that circumstance, the cost of the ESA is the full estimated cost of the award, as there are no offsetting savings for lower public school enrollment,” the analysis said.
This is a criticism ESA opponents have raised. Students who would not have qualified for ESA eligibility before universal expansion and who otherwise would have remained outside the public school system represent costs the state would not previously have incurred.
Today, it’s difficult to identify the percentage of ESA participants who have never attended a public school, as ADE does not publish that specific information. Its quarterly enrollment reports do, however, provide a narrower measure: where new participants were enrolled immediately before entering the program.
In the following table from ADE’s FY 2026 third-quarter report, we can see how many participants switched from public to ESA statewide through universal eligibility in the first year it was available and how many switched in the following year:
The data show that an increasing share of new universal-eligibility participants are coming directly from public schools. But that does not mean every universal ESA participant represents a public-school cost that disappeared: some entered from private school or homeschool, and the reports only identify where students were enrolled immediately before joining ESA.
That distinction matters because, as JLBC's analyses illustrate, an ESA can save the state money, increase state costs or create an entirely new state expense depending on where the student otherwise would have been educated.
That’s why Denis Ponder told KAWC, “When a child leaves a public school or a charter school and goes and takes the ESA, the argument is it saves $7,000 a kid, and really, that's not the way the math works.”
When Savings Come in Chunks
When asked how ESA universal expansion has affected Yuma public schools, Sheldahl said the impacts weren’t immediate because there were already a lot of students who had left the public school system by 2022.
What happens, he said, is essentially the illustration he offered at the beginning. Enrollment can decline one student at a time, but many district expenses cannot. A school may lose funding for several students before enrollment falls enough to eliminate a classroom, bus route or staff position.
“[ESAs] siphon off revenue, but not enough revenue to cause the schools to significantly pull back staffing and other fixed expenses in order to balance it out,” Sheldahl said.
The same dynamic can work in reverse. Wellton Elementary has been adding class sections as its enrollment grows, but Superintendent and Principal Lisa Jameson said additional students do not necessarily generate enough new revenue to cover the full cost of another teacher.
“In one way that's good — there's choice there between teachers and everything — but that makes the budget just a little bit more tight because we do have increasing enrollment, but that increasing enrollment wouldn't necessarily cover the new teachers’ pay depending on where they are at on a salary schedule,” she said. “Salaries are usually one of the more expensive things in the budget.”
Other required expenses can put pressure on district budgets as well.
“There are certain statutory and constitutional requirements that we have to do, so we can’t pick and choose which ones we’re going to follow,” said Dale Ponder. “We have to follow them.”
One example he offered is special education, an area he described as underfunded even at the federal level, noting that federal Individuals with Disabilities Education Act (IDEA) appropriations remain well below the law’s 40% maximum federal contribution benchmark.
“That is woefully underfunded yet it’s required to be provided,” he said. “In Arizona, we can't have unfunded mandates, but we have plenty of underfunded mandates.”
So what does a district do when the budget’s tight but they have needs to meet?
“When that is underfunded, you end up stripping away from other areas in the budget to meet this mandate,” Ponder replied.
He added that can leave districts shifting limited resources between programs even when students in both general and special education have unmet needs.
Capital needs add another layer. Last year, a Maricopa County Superior Court judge found Arizona's public-school capital funding system unconstitutional, concluding that it did not provide the general and uniform system required by the state constitution.
These are some of the costs and constraints public school officials point to when they argue that per-pupil comparisons do not capture the full effect of ESA expansion on districts.
ADE estimates current ESA awards at an annualized $1.1 billion. As the switcher analysis from earlier illustrates, that figure does not represent $1.1 billion in money the state would otherwise have spent on public schools. Still, public school officials look at the scale of ESA spending against needs they say remain unmet in their districts.
“If we had a billion dollars to put in the school system, our lives would be different,” Jameson said. “People have asked me, ‘Well, don't you think parents should have choice?’ Of course, I think they should have choice. Choice is important and we should always have choice.
“But I drive on this road every day. I don't have the choice to tell someone, ‘I want you to fix that one — not that one over there because I don't drive there.’ And people who have money now are saying, ‘We want the state to pay for our children to go to a private school.’ That to me is 100% unfair.”
Jameson continued to note that she doesn’t think anyone would have an issue with military and special education children using ESAs, “but it’s no longer that; it’s become universal.”
How Do You Measure a Worthy Investment?
Public school officials argue that staffing, special education, facilities and other systemwide responsibilities help explain why district spending is higher than an individual ESA award. But ESA proponents see some of those same costs differently.
“Part of the issue is that it’s not about what is the specific amount of money that you give, it’s what does that money go toward?” Beienburg said.
Beienburg doesn’t dispute that public school districts carry costs that ESA families do not, but he argues that this is part of why ESAs can provide education at a lower cost to taxpayers.
“The state of Arizona pays tens of millions of dollars to build a school building for a school district or $100-plus million every year in building renewal grants,” he said. “All of that stuff goes to supporting these buildings, which are often simply not used or way under capacity. That’s a huge drain on the district’s budget and on state taxpayers.”
Beienburg has previously pointed to underused school buildings as one example of those costs. A 2019 Goldwater Institute report he co-authored, citing Arizona Auditor General data, found district building space had grown considerably faster than enrollment over the preceding years.
And given the funding public schools already receive from their combined revenue sources, Beienburg looks to parents leaving public schools for ESAs as an indicator that the public school system is failing some families.
“If taxpayers are being asked to spend $10,000, $12,000, $15,000 for a student in the public school versus $8,000 on the ESA, and the ESA is providing an education that the family believes is better, it’s very hard to argue that it’s a resource problem,” he said.
But this might be another area where an apples-to-apples comparison gets tricky.
Sitting back down with retired superintendent Jamie Sheldahl, he described one hypothetical student growing up in a financially stable household with highly educated parents and access to a literacy-rich environment.
“The fact is, those kids, they’re not very expensive to teach because they’ve got a great background when they come into school and they’ve got a family structure and family values that really supplement their education in great ways through travel, through reading, through community sports and activities,” he said. “As far as how much does it cost to teach that kid in a year? Not that much.”
The other child he described is one born into poverty, maybe a single-parent household or a multigenerational home with very few resources and lacking a literacy-rich environment.
“The first kid might come to kindergarten with a vocabulary that's ten times as big as the second one, so how much does that kid cost to educate?” he asked. “I can put 40 kids who are above grade level in reading and math and have great support at home in a classroom and they’re all there with a good teacher and they’re all going to excel, but if I do that with 40 of the kids who grew up in poverty and maybe don’t speak English as their first language, I can pretty much guarantee that maybe none of them — or maybe just one or two of them — will come out of that environment excelling.”
Sheldahl explained that these children cost more to educate because they need more resources.
“Where do we have a concentration of students from significant poverty? Public schools,” he said. “Not private schools, not homeschool vouchers — because their parents have to go to work or they don't get a check.”
And that might be another problem with treating every per-student dollar as interchangeable: students themselves are not.
A purely cost-based comparison also leaves a broader policy question: If one education model can serve some students for less, how much weight should cost carry in deciding how Arizona funds education?
At the same time, cost is not the state’s only consideration. The Arizona Constitution also requires the Legislature to maintain a general and uniform public school system.
In a state with finite resources, the tension lies in allocation. Some ESA proponents argue for a say in that decision.
“I shouldn't have to put my tax dollars into the public school and then also pay separately for my tuition here [Southwestern Christian School],” said Brittani Ibarra, an ESA parent. “It's my tax dollars. Why can't I decide where it goes?”
Some public school officials argue the state’s priorities are out of order.
“It feels like the state of Arizona has the autonomy where they can say, ‘We're not going to do the things that we're constitutionally obligated to do, so we can focus on the nice to have’s,’” said Crane Superintendent Dale Ponder. “As a school district, we're not afforded that same opportunity.”
All of these considerations don’t establish whether any particular ESA is worthwhile or whether a family should use one. But they do show why a lower price tag doesn’t necessarily settle the question of what education should cost.
—
This reporting is supported by a grant from the Arizona Local News Foundation.
