David Osborne recently predicted academic doom for red states having recently passed universal private choice programs. “This will accelerate the process of the rich getting richer while the poor fall further behind,” Osborne asserted. Osborne problematically ignored our nation’s actual experience with universal choice programs, making his column more a litany of faith than a clear-eyed analysis.
Osborne predicts a bleak future for states with universal private choice programs, with poor families left behind. Osborne prefers a charter school model of choice, keeping choice within the public realm of regulation and accountability:
“Is there an alternative, other than the status quo of struggling public school systems? Indeed there is. States and school districts could reduce bureaucratic controls, empower educators and increase choice, competition and accountability for performance within the public school system, through the spread of charter schools. Cities that have done so, including New Orleans, Washington, D.C., Denver and Indianapolis, have produced some of the nation’s most rapid improvements in student performance.”
Arizona lawmakers created the first universal private choice program in 1997, the nation’s first scholarship tax credit program. Decades passed before another state enacted a private choice law with equally expansive eligibility. Three years earlier, in 1994, Arizona lawmakers had created two de facto public universal choice programs in the nation’s most robust charter school law and a statewide district open enrollment statute. “Large” and “relatively lightly regulated” would accurately describe Arizona choice programs, both public and private. Arizona lawmakers expanded and supplemented scholarship tax credits repeatedly; the Arizona charter sector became the largest among states, and open enrollment between and within districts dwarfed both in combination. Arizona created the nation’s first education savings account program in 2011 and expanded eligibility several times before making it universally available to Arizona K-12 students in 2022.
Given Osborne specifically cites four jurisdictions with the sort of choice programs of which he approves- Denver, Washington D.C., New Orleans, and Indianapolis, it seems in order. The Stanford Educational Opportunity Project provides academic growth data by jurisdiction (schools, counties, and states) and student subgroups for the 2009-2019 period. Comparing the rate of academic growth for low-income students in each of these four jurisdictions with those of Arizona counties in Figure 1:
Academic growth is a very important academic measure. While raw scores are very strongly correlated with student demographics, growth is much less so. Scholars widely view academic growth as the best measure of school quality. Many years into exposure to universal choice programs, Arizona’s low-income students seemed to be too busy learning to suffer Osborne’s predicted calamities. Greenlee County is a rural and remote area of Arizona with approximately 1,500 students and (alas) no charter or private schools during the period covered by the data. In this measure, a “zero” basically entails having learned a grade level worth of material per year on average, so the performances for Denver, DC and Orleans Parish are respectable, Marion County (host county of Indianapolis) less so.
The Stanford Educational Opportunity Project also measures the gap in learning rates by subgroup, which is measured by subtracting the learning rate of poor students from that of non-poor students. The four jurisdictions lauded by Osborne ranked first, second, third, and fifth in comparison to Arizona counties in terms of the amount of learning rate inequality between poor and non-poor students. There was exactly one state that had a positive rate of academic growth for both poor and non-poor students and had a faster rate of academic growth for poor students. It is the state marked “1” and spoiler alert…it is Arizona, the host of multiple universal choice programs.
Osborne’s hypothesis held that what some would regard as wild, lightly regulated “let it rip” choice programs would prove to be a disaster for low-income students, and conversely, well-regulated choice programs should advantage the poor. In practice, however, we find evidence to support the opposite conclusion. These results would not have surprised Milton Friedman in the least:
The results in the above figure also sit comfortably with the diagnosis of John Chub and Terry Moe, who identified politics as the central flaw of the American public school system. The American public school system does not do a terrific job on average in educating students, but it does a fantastic job in maximizing the political power and revenue of employee unions and their associated fellow travelers. Attempting to set up a governance structure of politically disinterested technocrats who will give families just the right amount of freedom and just the right amount of regulation comfortable for technocrats is an appealing theory. In practice, the most powerful and reactionary forces in modern American politics hijack the project easily unless a powerful, supportive constituency rises to defend the programs.
