What Happens When Public Schools, Charter Schools and Taxpayer-Funded Private Choice All Compete for the Same Children—and the Same Public Dollars
By Bear Howard and Associates
Note: this is more than an essay. It is, in many ways, a white paper, almost a thesis, to explore the topic in great detail. I apologize for the length. The topic is much bigger than most of us can comprehend. Arizona is an experiment in America, and it’s commitment to educating its children.
For most of modern American history, public education rested on a fairly simple bargain.
A community established schools. Taxpayers supported them. The schools were required to educate virtually every child who showed up at the door. Rich or poor, academically gifted or struggling, disabled or able-bodied, highly motivated or chronically absent, the children belonged to the system.
That system was never perfect. Some public schools were excellent. Some were terrible. Bureaucracies grew. Teachers’ unions acquired influence. School boards became political. Regulations accumulated. Innovation could be painfully slow.
But the underlying arrangement was clear:
The public school was the community’s educational institution of last resort—and first resort.
That arrangement is changing.
Across America, parents increasingly have alternatives: charter schools, magnet schools, open enrollment, private schools, homeschooling, online education, vouchers, tax-credit scholarships and education savings accounts.
No state has pushed this transformation much farther than Arizona.
That makes Arizona something close to a Petri dish for the future of American education.
Arizona did not simply add one alternative to traditional public schools. It layered choice upon choice.
First came one of America’s most aggressive charter-school systems.
Then extensive open enrollment allowed students to cross traditional district boundaries.
Then Arizona transformed its Empowerment Scholarship Account program into a universal program available broadly to families regardless of income.
The result is an educational marketplace in which the traditional district school still carries many of the broadest public obligations but increasingly competes for students and funding against institutions operating under different rules.
The question is no longer simply:
Are charter schools better than public schools?
Charters are public schools.
Nor is it simply:
Should parents be allowed to send their children to private schools?
Of course parents have that right.
The deeper question is:
What happens to a universal public-school system when government deliberately creates competing educational systems with greater operational freedom, allows public money to follow families into those alternatives, and then judges the traditional public schools partly by the performance that remains?
Arizona is beginning to provide the answer.
The Traditional Public School Is Still America’s School
Despite decades of discussion about school choice, the overwhelming majority of American children attending public schools still attend traditional district schools.
In the 2022–23 school year, approximately 45.1 million of America’s 48.9 million public-school students attended traditional public schools, while about 3.7 million attended charter schools. Charter schools represented about 7.6 percent of public-school enrollment nationally.
Arizona looks considerably different.
In 2024–25, Arizona reported 871,526 students enrolled in district schools and 231,660 in charter schools. In other words, roughly one out of every five Arizona public-school students was already attending a charter.
Arizona’s charter share has been far above the national average for years. NCES reported that charters represented about 20.5 percent of Arizona public-school enrollment in 2022–23, compared with 7.6 percent nationally.
Then Arizona added universal ESAs.
As of August 10, 2026, the Arizona Department of Education reported 102,098 students enrolled for the 2026–27 ESA year, while cautioning that the number would remain fluid during the opening weeks of the new contract year.
Arizona therefore has three substantial educational sectors operating simultaneously:
traditional district public schools,
public charter schools,
and publicly financed private/home educational alternatives through ESAs.
That is what makes Arizona so revealing.
Stage One: The Traditional Public-School Model
The traditional public-school district does more than operate individual schools.
It maintains an educational infrastructure for a geographical community.
That distinction is fundamental.
An Arizona district is responsible for schools, buildings, administrators, transportation, personnel systems, special education, financial controls, curriculum requirements and a broad range of state and federal obligations.
Arizona law generally requires district schools to admit resident children who meet the age and enrollment requirements.
The district cannot structure its responsibility around only the students most compatible with a particular educational philosophy.
The district has to remain there.
If a child’s parents become unemployed, the district remains.
If the child has a serious disability, the district remains.
If the family moves into town halfway through the year, the district remains.
If another school does not work out, the district remains.
If a child has behavioral difficulties, language difficulties, academic deficiencies, family instability or no parent aggressively researching educational alternatives, the district remains.
That universal availability is one of the traditional public school’s greatest virtues.
It is also one of its greatest burdens.
Stage Two: Arizona Introduces the Charter School
Charters changed the equation.
The original charter concept was not inherently anti-public-school. It was an experiment inside public education.
Give educators and organizations greater freedom.
Permit them to develop specialized schools.
Reduce bureaucracy.
Hold them accountable for results.
Allow parents to choose.
See what happens.
Arizona embraced that philosophy aggressively.
The difference in legal structure is striking.
Arizona law requires charters to meet numerous important obligations involving health, safety, civil rights, financial accountability, special education and state testing. But the law also states that, except where charter law or the charter itself provides otherwise, the charter school is exempt from statutes and rules relating to schools, governing boards and school districts.
That represents a fundamentally different operating philosophy.
The traditional district is generally told:
Here are the rules under which you will operate.
The charter is much closer to:
Here are the outcomes and essential protections we require; otherwise, you have substantially greater latitude in deciding how to get there.
Consider teachers.
Arizona law generally provides that a district teacher cannot be employed without the required teaching certificate.
Charters have considerably greater flexibility.
This can allow a charter to hire an accomplished engineer to teach physics, a computer scientist to teach programming or a working artist to teach art without requiring precisely the same traditional professional path.
That can be a tremendous advantage.
It can also permit terrible hiring.
Freedom does not guarantee excellence.
Freedom magnifies management.
Excellent leadership can use flexibility brilliantly.
Bad leadership can use the same flexibility disastrously.
That is one of the lessons of the charter movement that should matter to everyone interested in public education.
Charters Also Have Greater Ability to Define the Product
Arizona charter law specifically permits schools to emphasize particular educational philosophies, instructional styles and subject areas.
A charter can essentially say:
This is what we do.
Classical education.
STEM.
College preparation.
Fine arts.
Language immersion.
Montessori.
Back-to-basics instruction.
A particularly demanding academic environment.
A district can certainly create specialized programs too, but it does so within a substantially larger statutory, political and institutional structure.
Enrollment also works differently.
An Arizona charter must generally accept eligible applicants, but only until the capacity of a program, class, grade or building is reached. When demand exceeds capacity, the school can use the selection procedures permitted by law.
That is not the same mission as maintaining a geographically based school system capable of educating the resident population.
This distinction does not mean charters are secretly selecting only brilliant children.
Many charters educate disadvantaged populations, minority populations and children from low-income families.
Some have produced excellent results with exactly those students.
But an individual charter can determine its size, educational emphasis and configuration in a way that the community’s underlying district system often cannot.
And Some Charter Schools Really Do Perform Better
It would damage the argument to pretend otherwise.
The strongest evidence does not show that charter schools are simply a fraud produced by selective families.
CREDO at Stanford studied approximately 1.8 million charter students across multiple states and compared their progress with matched students in traditional public schools.
Its 2023 national study found that charter students experienced, on average, the equivalent of about 16 additional days of learning in reading and six additional days in mathematics per school year compared with matched traditional-public-school peers. Schools affiliated with charter management organizations showed still larger average advantages.
That matters.
Some charter schools are apparently doing something better.
The interesting public-policy response should therefore not be:
How do we protect district schools from successful charters?
It should be:
What are successful charters doing that traditional public schools should be allowed to do too?
If a regulation prevents a district principal from assembling the best faculty while a charter can do so more easily, reconsider the regulation.
If a charter succeeds because it can rapidly redesign curriculum, staffing or scheduling, ask why a district school cannot.
If charter autonomy is producing better education, then charter success should become an argument for greater freedom throughout public education, rather than merely an argument for creating more institutions outside the district system.
But there is another side to the comparison.
A School’s Test Score Is Not the Same Thing as the School’s Contribution
Suppose one school has 70 percent of students proficient in mathematics and another has 45 percent.
The first school may indeed be better.
But the scores alone do not prove it.
Children do not arrive at school as identical raw material.
They arrive carrying enormous differences in:
- prior education,
- parental education,
- family income,
- vocabulary,
- health,
- disability,
- stability at home,
- attendance,
- parental expectations,
- access to books and technology,
- tutoring,
- transportation,
- and thousands of hours of experiences accumulated before the teacher ever meets them.
This matters whenever families choose schools.
Choosing itself requires action.
A parent must learn that alternatives exist.
Investigate them.
Fill out applications.
Compare schools.
Arrange transportation.
Possibly pay expenses beyond whatever public subsidy exists.
Monitor whether the school is working.
And sometimes move the child again.
Those activities are not distributed equally throughout the population.
Federal research illustrates the relationship with parental education particularly well.
In a national 2016 parent survey, only about 20 percent of students whose parents had less than a high-school education had parents who reported considering other schools. Among students whose parents had graduate or professional degrees, the figure was approximately 41 percent. The likelihood of moving to a neighborhood for access to a particular public school also increased substantially with parental education.
This does not mean less-educated parents do not care about their children.
It means navigating an educational marketplace requires information, time, confidence and resources.
That creates an unavoidable sorting mechanism.
Choice does not randomly distribute families.
The Quiet Advantage: Educationally Engaged Families Find One Another
Imagine two otherwise similar children.
One has parents who study school report cards, attend meetings, talk to teachers, compare curricula, know other highly involved parents and are willing to change schools quickly if dissatisfied.
The other attends whatever school is nearby because nobody in the household has the time, knowledge or inclination to conduct an educational search.
Even if both children begin with equal ability, their educational environments may diverge.
Put enough highly engaged families into one school and something else happens.
Those families affect the school itself.
They reinforce academic expectations.
They volunteer.
They demand good teachers.
They complain when courses deteriorate.
Their children bring peer expectations about homework, college and achievement.
The institution begins benefiting not merely from individual parental involvement but from the concentration of educationally engaged families.
That does not necessarily mean the school caused the entire advantage visible in its test scores.
Some of the advantage walked in the front door.
That is why comparing raw school averages can become misleading.
The Traditional Public School Can Become the Residual Institution
This is where the structural issue becomes more troubling.
A traditional public school remains available for everybody.
That includes children whose families successfully navigate school choice.
But it also includes disproportionate numbers of children whose families:
cannot navigate it,
do not navigate it,
lack transportation,
live where alternatives are scarce,
move frequently,
need complex services,
or return after another placement fails.
Over time, the district risks becoming what might be called the residual provider.
Not the school system exclusively for poor children.
Not the school system exclusively for disabled children.
Not the system exclusively for less-involved families.
But the institution increasingly responsible for the population left after numerous other forms of choice have operated.
And then society compares the average performance of the remaining population against the populations that have sorted themselves into alternatives.
That comparison can be informative.
But it is not automatically an apples-to-apples measure of institutional effectiveness.
Then Comes the Financial Catch-22
School finance makes the problem considerably more serious.
Educational funding is increasingly designed around the principle:
The money follows the child.
There is an obvious logic to that.
If Johnny no longer attends School A, why should School A continue receiving all the money associated with educating Johnny?
But schools do not operate like individual checking accounts.
Suppose an elementary school has 500 students and loses 25.
Its enrollment has fallen 5 percent.
Its expenses do not fall 5 percent.
It still needs:
the building,
the principal,
the office,
the heating and cooling system,
the library,
the cafeteria,
the custodian,
the counselor,
most of its teachers,
technology systems,
insurance,
security,
and transportation infrastructure.
Perhaps it can eventually eliminate one teaching position.
Perhaps not.
If those 25 departing children came from six different grade levels, there may be no classroom that can simply disappear.
These are sometimes described as stranded costs or sticky costs.
Funding can leave one student at a time.
Costs generally disappear in chunks.
A classroom.
A bus route.
A school.
And shutting down those chunks is painful.
Arizona Is Already Seeing the Consequences of Declining District Enrollment
This is no longer theoretical.
The Arizona Auditor General reported in February 2026 that most Arizona school districts continued to experience declining enrollment and associated funding reductions. One-third of districts had experienced enrollment reductions serious enough potentially to increase their financial risk.
The Auditor General also reported that as overall district enrollment declined, the proportion of remaining students requiring special-education services increased.
A separate January 2026 financial-risk analysis found nine of 207 analyzed districts in the highest-risk category and another nine approaching it, up substantially from the prior year.
District spending on instruction fell by nearly $53 million in FY2025, and average teacher salaries actually declined in 95 districts even though the statewide average rose slightly to $65,613.
None of that can fairly be blamed entirely on school choice.
Arizona also has demographic changes, birth-rate changes, migration patterns, local management decisions, inflation and other financial pressures.
But declining enrollment and school funding are directly connected.
And now Arizona has created another mechanism capable of accelerating that decline.
Stage Three: The ESA Changes the Nature of the Experiment
Arizona began its Empowerment Scholarship Account program as a targeted program.
Then, in 2022, Arizona became the first state to establish essentially universal eligibility. By January 2026, the Education Commission of the States reported that universal school-choice programs had spread to at least 18 states.
This represents a much more dramatic policy step than charter schools.
A charter is still a public school.
It participates in Arizona’s public-school testing system.
It remains subject to numerous public accountability requirements.
It operates under a charter agreement.
An ESA is different.
Arizona deposits state education funds into an account controlled by a qualifying parent for approved educational expenses. A family can use ESA money for private-school tuition and various other permitted educational services and products. The parent agrees not to enroll the child simultaneously as a full-time district or charter student receiving the same public funding.
The Department of Education describes the principle plainly: education money follows the student to the educational setting selected by the family, including education at home.
That is a profound transformation.
Arizona has moved from:
government funding a public-school system
toward:
government financing individual educationa
The Explosion in ESA Enrollment
Before universal eligibility, Arizona’s ESA program was comparatively small.
JLBC data show:
2019: 6,450 ESA students
2020: 7,781
2021: 9,777
2022: 12,049
2023: 53,364
2024: 76,427
2025: 87,602
JLBC estimated about 101,602 for FY2026 and projected 113,602 in FY2027, while stressing that future projections remain highly uncertain.
ADE’s more current count reported 102,098 enrolled for 2026–27 as of August 10, 2026.
A program serving roughly 12,000 students before universalization has therefore grown into a program serving roughly 100,000.
That is not an experimental footnote anymore.
It is a major component of Arizona’s educ
And It Has Become a Billion-Dollar Program
The financial trajectory is equally striking.
ADE reported that the annualized value of ESA awards for the 85,195 students enrolled at the end of FY2025 was approximately $881.7 million.
JLBC’s FY2026 baseline included approximately $1 billion for original and universal ESA awards.
Its FY2027 baseline assumes 113,602 ESA students eligible for approximately $1.216 billion in awards. JLBC explicitly warns that the estimate is speculative because nobody yet knows where ESA participation will eventually reach a steady state.
This creates a different budget question from the charter-school debate
The Child Who Leaves Public School Versus the Child Who Never Attended
There are two fundamentally different ESA transactions.
Consider the first child.
A student is attending an Arizona public school.
The state is already paying toward that student’s public education.
The family takes an ESA and leaves.
Some portion of the state’s public-school expenditure disappears along with the child, and ESA spending replaces it.
There may be a state savings or additional cost depending on the exact funding circumstances, but conceptually some existing public expenditure is being redirected.
Now consider another child.
That child has attended private school for years.
The parents have always paid the tuition themselves.
The child enrolls in universal ESA.
Nothing has been saved at a district school because the child was never there.
Arizona taxpayers have simply assumed part of an educational cost that the family previously financed privately.
That distinction is critical.
Arizona’s Own Data Show Both Things Are Happening
In FY2024, only about 47.6 percent of newly enrolled universal ESA students in grades 1–12 had attended a public school immediately before entering ESA.
That means about 52.4 percent had not come immediately from public schools. That group included students from private education and other non-public circumstances, so it should not simply be labeled “private-school students.”
By FY2025, the composition had changed.
ADE reported that 57 percent of new universal ESA participants in grades 1–12 had attended public school immediately beforehand.
That evolution is extremely important.
The early universal ESA program disproportionately created a new public subsidy for children already outside public schools.
Now the program increasingly does something else as well:
It moves students directly out of public schools and moves public funding with them.
Arizona therefore experiences both financial effects simultaneously.
The First ESA Effect: Government Assumes a Previously Private Expense
There is nothing inherently illegitimate about government deciding to subsidize private education.
Government subsidizes many activities people might otherwise purchase themselves.
The policy question is whether that is where limited public education dollars should go.
Suppose a financially comfortable family has voluntarily paid $12,000 each year to send a child to private school.
Arizona now gives that family perhaps $7,000 or $8,000 toward the cost.
The family is clearly better off.
The private school receives public-supported purchasing power.
But Arizona has not relieved a district school of the expense of educating that child because the district was not educating that child in the first place.
The ESA is therefore largely a new state expenditure.
This is where the opportunity-cost argument becomes important.
It would be inaccurate to say:
Every ESA dollar was taken directly from a teacher’s salary.
Government budgeting does not work that way.
But it is entirely fair to ask:
What else could Arizona do with approximately $1 billion of annual public resources?
Money devoted to one purpose cannot simultaneously be spent on another without higher taxes, additional revenue, borrowing or spending reductions somewhere else.
A billion dollars could alternatively support some combination of:
higher teacher compensation,
classroom aides,
school counselors,
special education,
reading intervention,
career and technical education,
building repairs,
air-conditioning replacement,
school security,
technology,
capital improvements,
or other public priorities.
That does not prove those alternatives would automatically be superior.
It does establish that ESAs involve a very real opportunity cost.
The Second ESA Effect: Money Leaves While Infrastructure Remains
When ESA participants increasingly come directly from public schools, a different problem emerges.
The funding moves faster than the infrastructure can shrink.
A district can lose 100 students without losing them conveniently from four entire classrooms.
They may disappear five from this grade, nine from another, two from a special program, twelve from a high school and one or two from dozens of other places.
The money follows the children with mathematical efficiency.
The expenses do not.
Eventually districts adjust.
They combine classes.
Eliminate teachers.
Reduce electives.
Cut administrative positions.
Consolidate bus routes.
Defer maintenance.
Close schools.
Sell buildings.
Those responses may be financially necessary.
They also change what remaining families experience.
And that can produce the Catch-22.
The Catch-22 Begins Feeding Itself
Imagine the following sequence.
A district loses enrollment.
Funding falls.
The district cuts personnel and programs.
Class sizes increase.
Electives disappear.
A neighborhood school closes.
Families become dissatisfied.
The most mobile and informed families investigate charter, private and ESA alternatives.
Additional students leave.
Funding declines again.
The district makes another round of reductions.
Its remaining population becomes more concentrated with students requiring expensive or complex services.
Reported academic performance may decline.
The declining performance becomes another reason for families to leave.
At that point, cause and effect become extremely difficult to separate.
Did the public school lose students because it deteriorated?
Or did it deteriorate partly because it lost students and resources?
The answer can easily be:
both.
That is a feedback loop.
And once such a loop begins, comparing the final district with the alternatives that helped reshape its enrollment can become profoundly misleading.
Who Uses Choice?
This question becomes especially important as Arizona moves from charter choice to universal private choice.
Advocates describe ESAs as democratizing access.
There is truth in that argument.
A family that could never afford private tuition may now be able to consider it.
A child trapped in a genuinely terrible school may acquire an escape route.
A parent may use an ESA to construct an individualized educational program better suited to the child.
These are meaningful benefits.
But universal eligibility does not mean universal ability to use the program effectively.
A $7,000 or $8,000 subsidy does not magically create a private school near every family.
It does not create transportation.
It does not eliminate tuition above the ESA amount.
It does not give a working parent time to homeschool.
It does not guarantee admission to a preferred private school.
And it does not erase differences in parental educational sophistication.
Researchers examining Arizona’s ESA participation have found that participation rates have been particularly high in wealthier communities. A 2025 Brookings analysis of universal or near-universal programs in six states found that programs without income considerations—particularly Arizona and West Virginia—were disproportionately used in wealthier areas.
Significantly, that was not merely an Arizona phenomenon.
Similar socioeconomic patterns appeared in multiple states, although program design affected their severity. States using income limits, priorities or larger awards for poorer families could reduce the regressive pattern.
This is one reason Arizona matters nationally.
It is no longer necessary to speculate entirely about whether some Arizona dynamics might appear elsewhere.
Some already have.
But Arizona Is a Petri Dish, Not a Crystal Ball
This distinction is important.
If another state duplicated Arizona’s rules exactly, it would be reasonable to expect similar incentives:
families would have stronger reasons to explore alternatives;
private schools would have greater incentive to enter or expand;
existing private-school families would have an incentive to claim available public subsidies;
public schools would lose funding when students departed;
districts would confront stranded costs;
and families with greater information, transportation and financial resources would often be better positioned to use the marketplace.
But the magnitude would vary.
A dense metropolitan area with dozens of private and charter schools is different from rural Montana.
A wealthy suburban district is different from an impoverished urban district.
A state with generous school funding is different from one with chronically low per-pupil spending.
Transportation matters.
Demographics matter.
Private-school capacity matters.
Regulations matter.
Voucher size matters.
Whether eligibility is income-tested matters.
Whether private schools receiving public money must test students matters.
So Arizona cannot tell us exactly what will happen in every state.
But it shows clearly what incentives are created when the rules are structured this way.
And that is precisely what a Petri dish is for.
There Is Also an Accountability Gradient
Arizona’s three systems operate at different points on an accountability spectrum.
Traditional district schools receive public money and operate under extensive public regulation.
Charter schools receive public money but have substantially greater operational freedom. They remain public schools and participate in statewide public-school testing.
ESA-financed private education receives public money through the parent, but children outside public schools are not part of the same universal AASA testing structure. Arizona’s statewide AASA requirement applies to public district and charter students.
ESA law requires parents to provide education in subjects including reading, grammar, mathematics, social studies and science, and permits ESA money to pay for nationally standardized tests, but this does not create the same statewide public reporting system that exists for district and charter schools.
So Arizona has progressively separated public financing from public accountability.
That makes one particular question remarkably difficult to answer:
What educational return are taxpayers receiving from approximately $1 billion in annual ESA expenditures?
We can count participants.
We can count dollars.
We can identify purchases.
What we cannot presently do with comparable statewide data is reliably compare the academic growth of ESA students with similar children remaining in district and charter schools.
That is an extraordinary informational gap for a program of this size.
The Evidence on Private-School Vouchers Should Produce Caution
Research on traditional voucher programs does not provide a simple academic success story.
Some research has found positive effects on longer-term educational attainment, such as graduation or college enrollment.
But several modern voucher studies have found significant negative effects on standardized academic achievement, particularly when programs expanded rapidly.
A review of the voucher literature described a pattern in which attainment effects sometimes appeared positive while effects on measured academic achievement were often negative.
That does not prove Arizona’s universal ESA program harms achievement.
Arizona has not produced the comparable statewide student-outcome data needed to make such a claim.
It means the opposite claim—that transferring public money to private educational choices necessarily improves academic outcomes—is also unsupported.
We are spending first and learning later.
Perhaps much later.
The Great Paradox
This produces one of the strangest features of modern education policy.
Traditional public schools are subjected to enormous accountability pressure.
Their test scores are published.
Their school grades are published.
Teacher qualifications are regulated.
Their governing boards meet publicly.
Budgets receive scrutiny.
Buildings are public assets.
Failing schools become political controversies.
Then some of those same public dollars can follow students into settings where taxpayers may receive substantially less comparable information about academic performance.
The argument is that parental choice itself provides accountability.
If parents dislike the school, they can leave.
There is considerable merit to that idea.
Markets do discipline providers.
But education has one enormous complication:
Parents cannot always accurately measure educational quality in real time.
A child can be happy at school.
Parents can love the teachers.
The environment can feel safe.
The curriculum can sound impressive.
The report card can look wonderful.
And the student can still be two years behind in mathematics.
That is why objective assessment exists.
Consumer satisfaction and educational effectiveness are related concepts.
They are not identical.
Public Education Has Become a Marketplace—But Only Partly
The Arizona model is often described as educational competition.
But it is an unusual marketplace.
Traditional districts cannot simply behave like ordinary competitors.
A struggling restaurant closes.
A struggling public-school district cannot tell the remaining children:
We have decided this market is no longer profitable. Good luck.
Government still requires somebody to educate them.
The district carries that residual obligation.
This means Arizona has created a marketplace in which one competitor remains responsible for maintaining the underlying public infrastructure.
That competitor is simultaneously told to become leaner, more entrepreneurial and more competitive.
Hence the Catch-22.
The Private-School Family Has a Perfectly Rational Perspective
None of this should be turned into an attack on parents.
Consider parents already sending their child to private school.
They pay taxes.
They pay tuition.
Then Arizona says:
You qualify for an ESA.
Why would they decline it?
From the parents’ perspective, accepting the money may be entirely rational.
Likewise, a parent dissatisfied with a district school who discovers an excellent charter should move the child.
A low-income parent who can use an ESA to give a child a dramatically better education should not be told that the child must remain somewhere worse because the district needs the enrollment.
Children do not exist to financially support institutions.
That is the strongest moral argument for school choice.
But public policy must examine what happens when millions of individually rational decisions interact.
Something can make perfect sense for every individual family while producing serious systemwide consequences.
That is a classic public-policy problem.
We May Be Creating Two Different Educational Populations
The most important division may eventually be less about public versus private than about something more subtle:
families who actively manage an educational portfolio and families who depend upon the default system.
The first group searches.
Compares.
Moves.
Applies.
Changes course.
Supplements.
Uses tutors.
Uses ESAs.
Finds transportation.
Chooses specialized charters.
Pays tuition differences when necessary.
The second group uses the school that exists.
That second group will always include families who are completely satisfied with traditional public schools.
But it will also include disproportionate numbers of families with fewer financial, informational or logistical resources.
If that sorting becomes strong enough, the traditional public school can begin inheriting a larger share of the educational problems that the marketplace does not solve.
Then its performance falls.
Then the public is shown the performance data as evidence that the institution itself is inferior.
That is the potential distortion.
We Should Also Be Honest About What District Schools Do Wrong
None of this absolves traditional public education.
School districts sometimes waste money.
They can become bureaucratic.
Poor teachers can remain too long.
Innovation can be smothered by regulation and institutional resistance.
Administration can grow while classrooms struggle.
School boards can concentrate on politics instead of education.
The Arizona Auditor General repeatedly finds large spending variations even among comparable districts and identifies opportunities for improved efficiency.
Some parents chose alternatives because the district system genuinely failed them.
That fact should not disappear from the discussion.
Indeed, if traditional public schools respond to competition merely by demanding that alternatives be restricted, they will have missed the most valuable lesson school choice provides.
People leave institutions for reasons.
Those reasons should be studied.
The Better Question Is What Rules Actually Produce Better Schools
After roughly three decades of Arizona charter experience and several years of universal ESA expansion, the debate should mature.
It should no longer be:
Public schools good. Charters bad.
Nor:
Government schools bad. Choice good.
Those are political slogans.
The useful questions are much more practical.
Which regulations actually protect children?
Which merely protect bureaucracy?
Which teacher-certification requirements improve teaching?
Which unnecessarily prevent talented people from entering classrooms?
Which charter freedoms correlate with stronger academic growth?
Which charter practices should district schools be allowed to copy?
What minimum academic information should be required whenever taxpayers finance a child’s education?
Should universal ESA subsidies go equally to a millionaire’s child already attending private school and a poor child attempting to escape a failing school?
Should some public funding remain with districts to support unavoidable system capacity and stranded costs?
Should public money carry basic nondiscrimination, financial-transparency and academic-reporting requirements regardless of whether it reaches a district, charter or private school?
Those are considerably harder questions.
They are also the right ones.
The Financing Question Cannot Be Avoided
Arizona’s ESA debate ultimately exposes a philosophical decision about what taxpayers are financing.
Are taxpayers funding:
a public-school system capable of educating every child?
Or are taxpayers funding:
the individual education of each child wherever the parent chooses?
Those concepts overlap.
They are not identical.
If society chooses the second model completely, it must confront what happens to the first.
Someone must still operate schools in rural communities where no meaningful marketplace develops.
Someone must educate children other schools do not serve.
Someone must maintain special-education capacity.
Someone must accept children in October, January and March.
Someone must keep facilities ready even when enrollment fluctuates.
Someone must remain.
That function costs money.
If public policy finances only individual students and refuses to recognize the cost of maintaining universal educational capacity, the underlying institution can become progressively weaker even while every individual funding decision appears rational.
What Arizona May Be Showing America
Arizona’s experience suggests several conclusions.
First, choice is popular because it solves real problems for real families. That political reality is unlikely to disappear.
Second, charter-school autonomy can produce genuine educational improvement. Strong charter results should be studied rather than dismissed.
Third, choice produces sorting. Families differ in information, education, mobility, transportation and financial resources, so access to nominally universal choice will never be perfectly equal.
Fourth, money following students creates financial pressure on institutions whose costs cannot fall student by student.
Fifth, universal vouchers create a new fiscal phenomenon when taxpayers begin subsidizing families who previously purchased private education without a government payment.
Sixth, once vouchers become universal, they can also accelerate movement directly out of public schools.Arizona’s own data show the proportion of new universal ESA students coming directly from public schools increased from less than half in FY2024 to 57 percent in FY2025.
Seventh, academic accountability becomes weaker precisely as public financing moves farther from the traditional system.
And eighth, traditional district schools increasingly face the possibility of being judged on a population partly shaped by everybody else’s freedom to leave.
That is the Catch-22.
Arizona Is No Longer an Exception
Arizona’s experience matters even more because other states are following.
As of early 2026, universal school-choice programs existed in at least 18 states, with additional states expanding eligibility.
Research comparing several universal or near-universal programs already finds some common patterns, including disproportionately high participation in wealthier communities when programs lack income-based restrictions.
So the proposition that “Arizona is unique and therefore tells us nothing about America” is becoming less credible.
Arizona may simply have arrived earlier.
If states adopt similar rules, similar incentives should be expected.
Not identical results.
But similar pressures.
There Is a Way Out of the Catch-22
The answer does not have to be abolishing charter schools.
It does not have to be abolishing ESAs.
And it certainly should not be forcing children to remain in bad schools merely to preserve district enrollment.
The answer begins with recognizing that competition should operate under intellectually honest rules.
If charter flexibility works, give more flexibility to district schools.
If public money follows students, recognize the legitimate fixed cost of maintaining a universal public educational infrastructure.
If taxpayers finance private education, require enough standardized information to determine whether children are actually learning.
If the purpose of ESAs is educational opportunity, consider whether unlimited subsidies to families who would purchase private education anyway are the best use of scarce public resources.
If disadvantaged families are supposed to benefit from choice, design transportation and funding structures that make choice genuinely usable rather than theoretically available.
And stop judging schools primarily by raw proficiency averages.
Follow children.
Measure where they started.
Measure how much they learned.
Study who entered.
Study who left.
Study which students returned to district schools.
Study which schools consistently add educational value regardless of the children they receive.
That would tell us far more than another political argument about whether charter schools or traditional public schools are “better.”
The Fundamental Public Obligation
There is one principle that should survive whatever educational marketplace America creates:
Every child must have somewhere good to go.
Not somewhere technically available.
Somewhere good.
A parent who finds a superior charter school should be able to use it.
A family that genuinely benefits from private education should have the freedom to choose it.
Innovators should be allowed to create better ways of teaching.
But society must also preserve a strong institution for the child whose parents do none of those things.
That child has exactly the same claim on a decent education.
Perhaps an even greater one.
The great danger in the current evolution of school choice is that America could gradually construct an increasingly sophisticated marketplace for families who know how to navigate it while allowing the default institution serving everyone else to deteriorate.
Then we would congratulate ourselves on expanding choice while quietly reducing the quality of the choice that remains available to the largest number of children.
The Arizona Experiment
Arizona has done something historically significant.
It has asked what happens when government progressively loosens the connection between public education and publicly operated schools.
The charter movement said:
Let public schools be operated differently.
Open enrollment said:
Let families cross district boundaries.
Universal ESA said:
Let the public education dollar leave the public-school system entirely.
Each step can be defended individually.
Together they amount to a transformation.
Arizona is no longer merely operating a public-school system with some alternatives around the edges.
It is building an education marketplace with a public-school system embedded inside it.
The traditional district remains the largest component.
It remains the provider with perhaps the broadest obligation.
It remains responsible for enormous amounts of physical and institutional infrastructure.
But it is now competing with organizations possessing greater operating freedom and with privately operated alternatives eligible to receive public financial support.
That changes everythi
The Question America Eventually Has to Answer
The debate should not end with whether parents deserve choices.
They do.
Nor should it end with whether traditional public schools deserve protection.
Institutions do not deserve protection merely because they are institutions.
The final question is harder:
How do we preserve parental freedom and educational innovation without financially and socially hollowing out the institution that must remain capable of educating everybody?
Arizona may be the first state forced to answer that question on a large scale.
It will not be the last.
The lesson of the Arizona Petri dish may ultimately be that choice and public education are not inherently enemies—but choice without attention to systemwide consequences can gradually transform the population, finances and performance of the schools left behind.
A good charter school teaches us that autonomy can work.
A bad charter teaches us that autonomy needs accountability.
A successful private school teaches us that government does not possess a monopoly on good education.
The ESA teaches us that public financing can empower families—but also that subsidizing private choice creates substantial new public obligations and difficult questions about equity and accountability.
And the traditional district school teaches us something easier to overlook:
Somebody still has to be there for everybody.
That obligation has a value.
It has a cost.
And it should be included whenever we compare one educational system with another.
Otherwise Arizona—and eventually America—may commit a remarkable analytical error.
We may create one educational sector with the broadest obligations, the heaviest infrastructure, significant regulatory constraints and an increasingly difficult population to serve.
We may simultaneously give competing sectors greater freedom and allow public dollars to follow families into them.
We may then compare the resulting test scores and financial condition.
And finally we may announce that the traditional public-school system has failed.
If that happens, we will need to ask whether we discovered the failure—
or designed it.