The Impossible Standard for ESAs: Zero Misuse

Arsenic is a highly toxic chemical element and a known carcinogen. There is arsenic in some drinking water.

That sounds alarming, because arsenic can be dangerous at high levels or over long exposure. But the public policy response is not to demand that every last trace be removed from every glass of water at any cost.

The Environmental Protection Agency sets a maximum contaminant level at 10 parts per billion for arsenic in public drinking water systems. The goal is safety, risk reduction and a standard that can actually be administered.

Zero is not the rule

That is how public policy works almost everywhere. We do not eliminate every risk. We try to reduce the risk to safe levels and weigh the cost of getting closer to zero.

Except, it seems, when it comes to education savings account programs.

A bad purchase becomes a headline. A bad vendor becomes proof of collapse. A parent buys something outside the rules, and suddenly the preferred solution is to roll back the program.

So, it might be astonishing to some to see that this standard of zero risk is not applied anywhere else in government.

The federal government reported roughly $186 billion in improper payments across 64 programs in fiscal year 2025. That does not mean every improper payment was fraud. Many were payment and documentation errors.

But that is precisely the point.

Government programs are prone to administrative errors. They’re complicated. Humans make mistakes. Systems are prone to errors. Bad actors exist.

The answer is better program integrity, not abolishing Medicare, Medicaid, SNAP or unemployment insurance.

SNAP’s national payment error rate was 10.9 percent in fiscal year 2024. Medicaid’s improper payment rate was 6.1 percent in 2025. Medicare Part D’s was 4.0 percent. Unemployment insurance has had even higher improper payment rates.

For many people, these programs serve important purposes, so policymakers try to reduce errors while preserving access.

That last phrase matters: while preserving access.

Arizona tells a different story

But ESA programs should be treated the same way.

Arizona offers the clearest example. Critics circulated claims suggesting that roughly 20 percent of ESA spending was improper. My talented colleague John Kristof dug into the issue and showed why that number was badly misleading.

It came from a risk-based review of questionable transactions, not a representative sample of all ESA spending. When Arizona’s ESA spending is examined more appropriately, the picture looks very different: roughly 1 to 2 percent of dollars appear to be unallowable, and fraudulent or egregious purchases are a much smaller share, just 0.3 percent.

That is not zero. It should not be waved away.

States should police misuse but do so without strangling the program. Families using ESA dollars should expect clear rules. Taxpayers should expect serious oversight.

But “nonzero” is not the same thing as “out of control.”

The cost of chasing perfect

The inimitable economist Thomas Sowell nailed it when he wrote, “There are no solutions. There are only trade-offs.” 

There is a cost to chasing perfection. Too much bureaucracy can turn oversight into a barrier. And a program built to stop every bad purchase can start stopping good ones, too.

Policymakers have to balance oversight and access.

The right question is whether the program catches misuse without getting in the way of legitimate educational spending. A program that’s hard for families to use is a failed program.

The paperwork trap

Annie Lowrey of The Atlantic made a related point in a recent Conversations with Tyler episode. Talking about fraud and benefit programs, she said, “Improper payments are a function of the complexity of this system.”

The same principle applies to ESAs. Misuse and fraud are related, but they’re not the same thing. Fraud requires intent. Misuse can include mistakes, confusion, bad documentation or purchases that fall outside program rules. A parent who misunderstands an allowable expense is in a different category than someone inventing fake students or forging documents.

Make the rules harder to follow, and more families will make mistakes. Add more paperwork, and more parents will get tripped up. Families are bound to make mistakes if the rules create a maze and are hard to follow.

Lowrey’s broader point was that simpler rules can reduce improper payments and bureaucratic costs at the same time. The goal for ESA oversight should be clear rules on the front end, reasonable checks on the back end and consequences for actual fraud.

The point is to catch misuse without making the program miserable to use.

The same basic logic shows up across government. Zero misuse is almost never the standard. Serious oversight is.

Apply the same standard

Zero misuse is not the standard elsewhere.

Tax compliance: Every year, Americans fail to pay hundreds of billions of dollars they legally owe. The IRS calls this the “tax gap,” and its projected annual gross tax gap for tax year 2022 was $696 billion. The IRS audits and enforces the law, yet leakage occurs. But nobody has ever argued, to my knowledge, that the IRS should be abolished because of this.

Prescription drugs: FDA approval doesn’t mean a drug has zero side effects. It means the benefits outweigh the risks for approved uses under regulated safeguards. Most people already understand tradeoffs.

Airport security: Airport screening is designed to reduce the risk of threats making it onto planes. But screening still misses things and slows people down. Too little screening makes flying less safe. Too much screening, and too many people miss their flights. That’s the tradeoff.

Public schools: Yes, even public school districts sometimes mishandle taxpayer dollars. Shocker, I know! But the same people who push for pulling back or eliminating ESA programs don’t treat a district’s bad purchase as proof that public education itself should shrink. Shouldn’t ESA families get the same basic fairness?

So what should policy makers consider a “fair” standard? Is it serious oversight?

If so, then a 1 to 2 percent misspending rate can’t be treated as a crisis in an ESA program while far higher improper payment rates are tolerated across major government programs.

Public programs should be accountable. Fraud should have consequences. Mistakes should be fixed.

But families should not lose educational opportunities because choice opponents apply an impossible standard they would never apply to the rest of government.