[Freddie deBoer] Education Doesn't Work

If environment dictates everything, then school quality would be immensely consequential. If school quality is a real, stable, and meaningful property, and the environmentalists are right about educational outcomes, then switching schools should have a dramatic effect on how a student performs in the classroom. But what parents typically find is that their child slots into the distribution just about where they were at the old school. The research record provides a great deal of evidence in this direction too.

For example, winning a lottery to attend a supposedly better school in Chicago makes no difference on educational outcomes. In New York? Makes no difference. What determines college completion rates, high school quality? No, that makes no difference; what matters is “preentry ability.” How about private vs. public schools? Corrected for underlying demographic differences, it makes no difference. Parents in many cities are obsessive about getting their kids into competitive exam high schools, but when you adjust for differences in ability, attending them makes no difference. The kids who just missed the cut score and the kids who just beat it have very similar underlying ability and so it should not surprise us in the least that they have very similar outcomes, despite going to very different schools. (The perception that these schools matter is based on exactly the same bad logic that Harvard benefits from.) Similarly, highly sought-after government schools in Kenya make no difference. Winning the lottery to choose your middle school in China? Makes no difference. All of this confirms anecdotal experiences. Did kids you know go from failures to whiz kids when they moved to a different state, a drastic change in environment? No, of course not. Because they’re the same kid.

Teacher quality perhaps exists but likely exerts far less influence than generally believed. There is no such entity as “school quality.” (I have addressed the charter school shell game at considerable length here.) There is the underlying ability of the students in a school that produces metrics that we then pretend say something of meaning about the school itself. That’s it. Zoning doesn’t make kids perform poorly by keeping them out of the best schools. Zoning creates the impression of the “best schools” by keeping out the kids destined to perform poorly.

Late last century France created an artificial cutoff date for conscription into the military. The result was a sudden influx of students into higher education; men who would have spent years in the military instead spent them going to school. But it not only made no difference in terms of wages or employment, it made almost no difference in the amount of people graduating from higher education. Conscription was screening out the marginal students from attending higher education. But removing the screen didn’t make those students any less marginal. If school matters most, going to more school (or whatever quality) should have a demonstrable impact on students. At a vast scale, in one of the wealthiest and most developed countries on earth, it did not. It simply made no difference.

Meanwhile constantly cited explanatory mechanisms, like class size, aren’t necessarily explanatory at all. Calls for smaller class sizes are ubiquitous because they are one of the few interventions endorsed by both neoliberal ed reformers and the teachers unions who are their enemies. But the quantitative effects of class size has been a research obsession for at least 40 years, and yet there is no real consensus on what exactly works, for what students, in what contexts. Claims of a small class size advantage routinely carry extensive lists of provisos, such as saying that it’s necessary for physical classroom space to not shrink along with number of students, the kind of requirement that makes implementing these efforts at scale a policy nightmare. Nor is there consistency between different studied groups; for example, at the university level class size effects vary wildly depending on who you’re looking at.