Private School Choice Programs Should and Will Continue to Expand
本文回应Douglas Harris对私立学校选择项目的批评，认为私立学校选择（如教育券和ESA）应继续扩大，并基于现有证据和趋势论证其合理性，对政策制定者和教育研究者有参考价值。
Douglas Harris has written a cogent essay arguing that universal voucher and ESA programs are undesirable because public charter schools are better. While I disagree with Harris on several key points, I begin by emphasizing our areas of agreement. We agree that public charter schools have a strong record of improving academic outcomes, especially for disadvantaged students. We agree that accountability tests sometimes fail to align with the curriculum taught to students, complicating comparisons between groups across time and school sectors. We agree that the ability of private schools to teach religion gives them an advantage relative to secular public schools. We agree that “public accountability is about more than just test scores. It is about ensuring that schools instill common language, values like tolerance, and belief in democracy” (Harris 2025). We agree that expanding private school choice programs in the form of vouchers and ESAs increases freedom, axiomatically. We agree that Levin's four-criterion evaluation scheme is useful for assessing the desirability of expanding private school choice. We tend to disagree on the evidence that is most relevant to the application of Levin's criteria, and how to characterize and interpret that evidence. Harris claims that statewide voucher programs in Florida, Indiana, Louisiana, and Ohio provide the best evidence regarding the likely effects of universal private school choice programs. Florida certainly is highly relevant, as it serves over half a million choice students, recently converted its voucher programs to ESAs, and is a model for many states, given its 28 years of experience publicly funding private schooling. The experiences of Indiana and Ohio are less instructive than those of Florida, since the Hoosier and Buckeye states are two of only three states with universal school voucher programs, the other being North Carolina. A total of 14 states combine universal eligibility with the more flexible ESA policy model, with Arizona the earliest adopter of that now popular approach. The least relevant program for forecasting the future effects of universal ESAs is the Louisiana Scholarship Program (LSP). The LSP was a highly regulated school voucher program limited to low-income students. It was repealed and replaced with the Louisiana GATOR universal ESA program in 2025. The LSP taught policymakers how not to design and implement a private school choice program (Wolf 2019). They have taken those lessons to heart, rendering the disappointing participant effects from my team's evaluation of the LSP the least informative findings for predicting the outcomes of current and future private school choice programs. Since no private school choice program designed like the universal ESAs proliferating across the country has been rigorously evaluated, I draw the evidence for my assessment from all the voucher and ESA evaluations, via meta-analyses and systematic reviews, with a special focus on the model programs in Arizona, Arkansas, and Florida, which best reflect the policies du jour. I disagree with Harris about the proper counterfactual for assessing private school choice programs. Harris creates a false dichotomy when he states, “today's debate is really about whether we should continue to expand public school choice or expand religious and other private school choice” (Harris 2025). Prominent commentators argue we should not expand either type of choice (Ravitch 2013), we should expand public but not private forms (Osborne 2017), we should expand private but not public forms (Kingsbury and Greene 2025), or that we should expand school choice in all its forms (DeAngelis 2024). This policy debate is about whether to expand private school choice. The “main alternative” to private schooling and homeschooling is not public charter schools, as Harris claims. Charters enroll only 8% of all K–12 students. The main alternative to vouchers and ESAs is neighborhood public schools (NPS), which enroll 78% of all K–12 students. When random lotteries are held to award private school vouchers, most students who lose those lotteries subsequently attend NPS, not charter schools, even in cities like the District of Columbia with large charter sectors (Wolf et al. 2007). There is no evidence that universal private school choice programs threaten public charter schools. Growth in charter school enrollments stalled in the years preceding the pandemic, but charters are expanding concurrently with the launch and expansion of private school choice programs. Private school choice appears to be catalyzing charter school growth, as all forms of school choice grow simultaneously. I agree with Harris that public charter schools are an attractive school choice for many parents. Charters are not, however, a perfect substitute for private schools, as some parents desire that their child learn in a religious environment, and public charter schools are prohibited from teaching religion or incorporating religious activities into the school day. History does not support Harris's claim that “secular teaching” and “non-discrimination” are “longstanding schooling traditions” in the United States. The Protestant version of Christianity was taught in public schools during the first 325 years of public schooling's existence in the United States and its antecedent, the American Colonies. Secular teaching represents a relatively recent break from the longstanding tradition of religious instruction and mandatory Bible reading in public schools, banned by the US Supreme Court in 1962. Sadly, discrimination is also a longstanding tradition of US public schools, most infamously by race in the Jim Crow South but more pervasively by residence, throughout the country, and up to the present day. Harris argues that K12 education in the United States does not require “radical change.” Plenty of evidence indicates otherwise. Students have lost nearly a full grade level of learning since the pandemic, with the losses larger for disadvantaged students and showing little evidence of rebounding. Chronic absenteeism is more than twice as high now as it was pre-COVID. One out of every three students feels unsafe at school (Centers for Disease Control and Prevention 2024). The United States is second only to Luxembourg in per-pupil spending (OECD 2024). Nevertheless, US students scored between 12th and 22nd in achievement on the most recent international math and science tests, with comparatively large gaps between our high and low performers (U.S. Department of Education 2024). A major course correction seems warranted. Harris states, “While the evidence on this is limited, there are reasons for concern that vouchers will make us less socially cohesive” (Harris 2025). School vouchers subsidize private schooling. There is a deep empirical literature on the differences between public and private schools regarding the tolerance, political participation, political knowledge and skills, and community engagement of their students and alumni. The findings from that research base favor private schools. Even if we focus only on the causal studies, private schools are at least as effective as public schools in preparing tolerant, active, informed, and engaged citizens. Private schools have advanced the public purposes of education for many decades and in various countries. I see no reason to think that will change if more students attend them with the support of public funding, especially since parents report that nurturing their child's civic values is an important reason why they send their child to a private school. Like it or not, universal vouchers or ESAs are coming soon to a neighborhood near you, if they are not there already. The number of states with universal eligibility for private school choice subsidies soared from 0 to 17 in the past 3 years. Policymakers will continue to enact universal school choice programs because they are popular with voters. As these programs continue to grow and spread, we will garner more evidence of their effects, and the debate over their benefits and costs will continue.