Poverty Matters, But Schools Matter Too!

Ben Adler says we should worry less about the state of education because the outcomes are driven by poverty:

Growing up in New York City I saw a lot of public schools with bad outcomes because the student population was deeply disadvantaged. But the few wealthy kids who went to those very same public schools turned out just fine (not long ago I met one who went to Yale — I guess my parents wasted their money on sending me to private school, because Yale rejected me). Meanwhile, there are “good” public schools in wealthier neighborhoods where the main difference is just that the kids come to school with a full stomach and their parents read to them before they go to bed at night. Then there are all the private schools where some of the teachers are (unofficially) tenured and are pretty unimpressive, but the students tend to turn out (usually) OK. On the other hand, students who come to school poorly rested from a night in a homeless shelter, malnourished, or with untreated illnesses tend to do poorly. All the charter schools in the world can’t solve those problems.

This is, I think, a half-truth as best illustrated by the charts I put together for this post. To see the half that’s true and the half that’s not true, you need to look at the data from NAEP’s Trial Urban District Assessment. For example, how many kids score “below basic” in the 8th grade math assessment? Well:

Boston and New York and Washington all look bad. But they also all have above-average numbers of poor kids. So what happens if we look just at the poor kids?

The New York and Boston data illustrate the half of the demographic determinism thesis that’s true. What at first glance appears to be low performing schools in New York and Boston looks, when you look just at the poor kids, to merely be a reflection of the fact that these schools have more challenging populations. On the other hand, look at the Washington data and you’ll see the half of the demographic determinism thesis that’s not true. Poor kids in Washington do much worse than poor kids a few stops north on the Acela.


And you see this pretty consistently if you look at the TUDA data. No matter how you slice the information demographically, New York and (especially) Boston are performing pretty well and Washington is doing terribly. It’s also true that no matter how you look at it, poor kids do worse than non-poor kids. Unfortunately, TUDA participation is purely voluntary and not all that widespread. So while the available information pretty clearly establishes a large, but far from complete role for demographics we can’t say much specifically about the situation that exists in the vast majority of American cities.