Tuition at many of New York City’s top private schools is over $40,000 a year. That’s more than Harvard. It’s grown almost 50 percent in the past decade–faster than private university tuition. It goes without saying that parents who shell this money out believe that their children will be getting a better education in exchange for all that money.
But what if they’re wrong?
A book published last year by two professors at the University of Illinois-Champaign, titled The Public School Advantage: Why Public Schools Outperform Private Schools, offered a new analysis of two nationally representative datasets–the NAEP test, known as the “Nation’s Report Card,” and the Early Childhood Longitudinal Study, which followed a group of 17,000 children from kindergarten through eighth grade. The researchers, Chris and Sarah Theule Lubiensky, chose to look specifically at math scores, in the belief that math is a purer measurement of the influence of school because students tend to learn very little math at home. They found that after controlling for the known effects of demographics, particularly wealth and social class, private schools (and independently run charter schools) demonstrated no advantage over public schools. In fact, in several grades, students with similar demographic backgrounds did better in math when they attended public schools.
Other independent studies have had similar results. This separate 2006 analysis of NAEP scores found after adjusting for demographics, fourth grade students did better in math in public schools, and eighth grade students did better in reading in private schools, while all other differences were not statistically significant. And several different studies of charter school performance have similarly found that their quality is uneven, with a tiny positive gain in reading over that achieved by similar students at public schools and an equally small disadvantage in math.
In recent days, the authors of Public School Advantage have been defending their findings and research methodology against a series of critics who advocate school choice reforms. Most of the evidence on the other side comes from small local studies of outcomes where vouchers were given to some students to pay for private schools; the Lubienskis argue that these studies are small and nonrepresentative.
For the parent or layperson trying to draw conclusions about private vs. public vs. charter, what stands out in all this is the paucity of evidence. No one should choose a school based on math scores alone. Measures that everyone would agree would be more relevant, such as graduation and college persistence rates, are not available. Schools that emphasize teaching to the test may demonstrate gains that are offset by great losses elsewhere. And even though researchers are correct to control for the effects of demographics on average, on an individual basis they are anything but irrelevant. As a parent from a less advantaged background, getting your kid into a more economically diverse school has a high likelihood of improving her circumstances and opportunities. Nobody disputes that schools full of rich kids do better. Indeed, the same large charter school study found that when you look specifically at the outcome for poor kids, charter schools do have a small advantage, which is highly relevant because charter schools have a greater concentration of poverty than regular public schools.
In fact, the biggest take-home for parents looking at the evidence on school performance should be this: family income and education level has far more statistical impact on a child’s performance in school than any characteristic of that school. If you are in a position to consider private school for your child, you probably don’t need to.