Choice and autonomy are now touchstones of U.S. education reform. Since over two decades ago, when John Chubb and Terry Moe famously argued that “choice is a panacea” and reported that more school autonomy led to better school outcomes, policymakers have been enamored with devolving authority away from school districts and creating options for families.
The number of charter schools in the United States is growing, with almost 6,000 such independent, largely autonomous schools of choice. At the same time, the market share of private schools —non-government schools that, unlike charter schools, are “not supported primarily by public funds” — is declining, creating demands for subsidies on the grounds that the 30,861 private schools do a better job of educating children.
Since 2012, states have adopted or expanded 28 voucher, tax credit or similar programs to subsidize families choosing private schools, and the charter school movement is growing with bipartisan support. These efforts are based on the popular notion that autonomy from state regulation allows schools to respond to parents’ preferences for quality education options. Autonomous schools have the freedom and incentive to adopt more effective practices in areas such as curriculum, pedagogy, staffing and management. Thus, the logic goes, they get better results.
However, our analysis of nationally representative samples of Catholic, Lutheran, conservative Christian, and other private schools — a total of 1,355 private schools — raises serious questions about that logic. We found that once we account for the fact that private schools serve families with more advantages associated with academic success—things like money and highly-educated parents—we find that public elementary schools are, on average, simply more effective at teaching mathematics. Indeed, demographic differences more than explain any apparent edge in the raw scores of private school students, and by the time they reach middle school, public school students score ahead of their demographically similar, private school peers, with differences ranging from a few weeks to a full grade level, depending on the type of private school.
Given public education’s considerable challenges, these findings are remarkable, and cut at the heart of the current reform movement. Yet they are starting to be echoed by other researchers at the Educational Testing Service, Notre Dame and Stanford universities. How should we understand these patterns?
Current school reform efforts elevate the idea of autonomy, positioning parents as expert choosers and schools as autonomous competitors that embrace effective instructional methods to attract students. This market model for education is neat, appealing, and quite possibly wrong.
Private schools have more operational autonomy, sure, but this autonomy is too often used to maintain outdated strategies that may align with parental preferences but are not particularly effective for educating students.
For example, private school students are more likely than their public school counterparts to sit in rows, complete math worksheets and believe that mathematics is “mostly memorizing facts”—a narrow view that captures neither the breadth of the discipline nor the reasoning that is central to it. In contrast, public schools have moved beyond traditional, repetitive exercises, and more often ask students to solve complex, real-world problems and to learn geometry, data analysis, and early algebra ideas, in addition to basic arithmetic.
This difference can partly be explained by the fact that public school teachers are more likely to be certified and to receive ongoing training in the field, keeping them current on research-based instructional standards and resources supported by professional entities such as the National Council of Teachers of Mathematics and the National Science Foundation. Private school teachers are rarely impelled to receive such training. And despite much criticism, teacher certification and up-to-date instructional practices are actually positive correlates of achievement, and the fact that these are more prevalent in public schools helps explain the public school advantage.
These patterns highlight some of the downsides of autonomy, especially in more competitive conditions where schools may try to play to popular demands instead of embracing professional expertise. In fact, it is not at all clear that parents choose schools primarily on the basis of academic effectiveness. School uniforms, the demographics of a school, and sports programs are easier to observe, and parents often consider these, along with religious values, to be more important than the quality of academic instruction, as consistently shown in studies of parents’ school-choice behaviors from places like the National Center for the Study of Privatization in Education at Columbia University. As schools are cast as competing businesses in current reforms, families may be influenced by image over academic substance, just as fast food marketing successfully focuses on fun and not nutrition. Professional models for education avoid this need to play to mass consumer demands, instead focusing on evidence and expertise.
It would be overly simplistic to say that parents are poor choosers when it comes to schools, since they work with the information, options, and priorities that they have. Instead, it appears that more autonomous schools—the private and charter schools so often credited with innovation—are doing a poor job of choosing effective educational strategies, of working on behalf of students, rather than parents. We agree that there are serious problems facing public education. But private models for public education do not appear to be the answer.