By Matthew Ladner & Jason Bedrick |
It is said that a lie travels halfway around the world before the truth can get its pants on. But when the truth finally catches up, it tends to arrive with receipts.
In recent weeks, anti-school-choice activists have accused Arizona’s popular Empowerment Scholarship Account (ESA) program of having a ludicrously high rate of misspending. That narrative, repeated endlessly despite being repeatedly debunked, has now collided with an inconvenient reality: the Arizona Department of Education’s own data tell a starkly different story, one that exposes the “unaccountable ESA” myth for what it always was: a misleading talking point masquerading as journalism.
On Thursday, the Arizona Department of Education published the results of a random audit of the ESA program, finding very low rates of misspending relative to other publicly funded programs and even less fraud. Less than 2% of ESA funds were spent on unallowed items, and 0.3% of the funds were spent on items considered “egregious” or fraudulent. The department is in the process of recouping those funds.
The department’s audit has punctured the gross distortions of Phoenix-based journalist Craig Harris of Channel 12, who had misrepresented the ESA program as being rife with fraud. The department called Harris’s claims “ridiculous,” “reckless,” and “a total misinterpretation of data provided by [the department] to Channel 12.”
Harris had claimed that 20% of ESA purchases represented misuse of funds based upon an examination of only a small portion of total ESA purchases—384,478 of the 1.8 million total ESA transactions since December 2024, approximately 20% of the total.
But it was not a random sample. The Arizona Department of Education had selected these purchases under a risk-based audit for additional scrutiny, so it was not a random sample that one could use to extrapolate rates of misuse in the ESA program. Instead, the risk-based pool was far more likely to contain misuse than the average purchase.
In other words, among the 20% of ESA purchases flagged for additional scrutiny, the department found 20% to be misspending. Harris, however, extrapolated the risk-based results on to the entire universe of ESA purchases. This represented a blatant distortion because the remaining purchases outside of the risk-based sample were far less likely to involve misspending.
Think of it this way: imagine that a reporter read a study showing that 20% of Americans were obese and that 20% of obese Americans had diabetes, then he ran a story claiming that 20% of all Americans had diabetes. Such a claim would be a complete distortion of the data, a conclusion that is entirely unsupported by the facts.
Harris was warned both publicly and privately that his “analysis” was deeply misleading. Nevertheless, he continued to repeat his “20% misuse” claim on television and social media.
The Arizona Department of Education decided to set the record straight by auditing a random sample of thousands of Arizona ESA purchases. The audit’s conclusion: “About 2% of purchases are unallowable expenses and only 0.3% represent fraud or egregious purchases.”
In short, Harris’s “analysis” on misspending was off by a factor of 10, and his accusations of fraud were off by nearly a factor of 100.
And although Harris made it seem like one in five ESA parents were buying diamond rings, the reality is that egregious purchases represented a vanishingly small percentage of ESA spending. Most of the unallowed items appear to have been innocent mistakes.
The Arizona Department of Education explained the difference between unallowable purchases and egregious purchases/fraud in a press release:
The submission of a purchase that is deemed unallowable does not constitute fraud. Most are innocent mistakes, such as an error in a form that must be resubmitted, or educational items that are not on the allowable list but that the user could have in good faith believed were permitted. Some examples would be backpacks, lunch boxes and water bottles.
A ridiculous figure of 20% fraud has been circulating concerning ESA purchases which resulted from a total misinterpretation of data that we provided to Channel 12.
Cracking down on misspending is important, but so is keeping things in perspective. The rate of improper payments for the Arizona ESA program, at 1.9%, stands far below a variety of programs which ESA opponents support, such as Medicaid (7.4%), food stamps (9.3%), and unemployment insurance (14.4%). Moreover, the Arizona Department of Education actively recovers misspending, and refers serious cases for criminal prosecution to punish criminal activity and to deter fraud.
Indeed, the only reason we know about the miniscule level of unallowed purchases in the ESA program is because it is so transparent—far more transparent than district schools, which do not report transaction-level data to state officials, let alone the public. Given the rash of recent scandals in Arizona’s district school system, one can only imagine what we would find if given access to information about every purchase that district schools have made, as we have for the ESA program.
In the interests of transparency and accountability, state lawmakers should require the district schools provide the same level of information about purchases as ESA families.
Harris has not yet retracted his false reporting. Instead, he has doubled down on his errors, erroneously claiming that the department’s audit was “largely skewed” and based on “a miniscule sample.” This claim is particularly ironic given that Harris erroneously treated a risk-based sample—which is inherently skewed—as though it was representative of the entire population. A random sample, by contrast, is representative of the whole.
Facts are stubborn things. So too, apparently, are anti-school-choice activists who can’t let go of the false narrative they pushed in the face of overwhelming evidence to the contrary.
Matthew Ladner is a Senior Advisor for education policy implementation and Jason Bedrick is a Senior Research Fellow at the Heritage Foundation’s Center for Education Policy.
