Don't punish schools for teaching white students too well
No Child Left Behind can be summed up in four words: good ideas gone awry. Unfortunately, one key part of the Obama Administration’s “blueprint” for overhauling the landmark federal law might perpetuate that legacy for another ten years. In particular, the President wants to intervene in schools with large achievement gaps—a well-intentioned instinct that, if implemented sloppily, could punish racially integrated schools for teaching white students too well.
Let’s start with a little context. The most consequential part of the Administration’s NCLB plan isn’t about school turnarounds, or merit pay, or even teacher effectiveness. It’s the call to focus federal accountability efforts on the very worst-performing schools in the country—and to leave the rest alone (or to the mercy of the states). This would be an enormous change, and would take pressure off the vast majority of the nation’s schools.
The proposal makes eminent sense, for, as Secretary Arne Duncan said when releasing his boss’s plan, ED can't micromanage 95,000 schools from Washington and instead needs to focus on a smarter, more targeted federal role. Obama and Duncan would still publish achievement results for all schools—disaggregated by race, income, special education status, etc.—but the typical American school would be more or less off the hook from federally-mandated sanctions.
But there’s an obvious downside to this approach: It pulls back from the notion that all schools should be held accountable for the performance of all of their students. It would allow schools with decent average test scores to skate by even if they post terrible results for their African-American or Hispanic or low-income kids. (This has not gone unnoticed by NCLB stalwarts like former Secretary of Education Margaret Spellings.)
To address this concern, the Administration added one more wrinkle in its blueprint: Along with the lowest-performing schools in the country, it would also sanction those with large, persistent achievement gaps, which would “be required to implement data-driven interventions to support those students who are farthest behind”; they would also get the dreaded label of a “challenge school.”
But what if some of those schools have large gaps not because their minority students are performing poorly, but because their white students are doing really well? Wouldn’t this needlessly demoralize the staff and students of schools that are actually quite good?
To find out if we were worried about nothing, we dug into the California test score data for math in 2007-2008. (Our data comes from the Academic Development Institute’s Achievement Profile Analytic Tool; we chose California because of its rigorous test and high bar for “proficiency.”) We identified all of the elementary schools with more than a handful of black and white students (there were 720). Then we pondered: Depending on how we define a “large” black-white achievement gap, how many schools would unfairly get the “challenge school” label because of high white student performance rather than low black performance?
We assumed that the Administration is focused on the “proficiency gap”—the difference between the percentage of white students who reach the proficiency bar on the state test and the percentage of minority kids who do. (Ideally, we’d look at white and black average test scores rather than proficiency rates because these are more precise indicators of achievement; unfortunately, states are not required to report these data, and any federal policy would most likely center around proficiency rates.) And then we defined a “large gap” as one that is bigger than average. (In California, that means a black-white gap larger than 23 percentage points.) That gave us 310 schools with above-average black-white gaps.
We crunched some numbers and here’s what we found: More than a third of those schools serve black students who are scoring above-average compared to other black students in the state. These schools aren’t “failing” their black students; if anything, they are doing relatively well by them. But their white students do even better. We ran the numbers for the Hispanic-white gap and got similar results; more than a third of the schools with above-average gaps serve Hispanic students with above-average achievement—but white students who score even higher.
This strikes us a serious defect. It’s particularly worrying for the cause of integration, because the very schools that are likely to get caught in this net are those where affluent (and generally high-achieving) whites attend school with less affluent black and Hispanic students—schools like Malcolm X Arts and Academics Magnet Elementary in Berkeley, California. Malcolm X has an almost equal mix of white and black students—which is not an accident, as the school is part of Berkeley’s “controlled choice” integration program—but posts a huge black-white achievement gap of 44 percentage points. That sounds terrible, but dig into the data and you’ll find that black students at Malcolm X are performing 14 percentage points above the state average for all black students. The reason the gap is so wide is that an impressive 98 percent of white students score at the proficient level (versus 54 percent for blacks). Certainly Malcolm X has room for improvement, but it is far from a “challenge school.” It might even be a model.
This doesn’t mean that the Administration’s proposal should be abandoned completely. But if Congress wants to move ahead with the idea, it should be careful with how it sets the parameters. One way is to define “large gaps” as large enough so you don’t inadvertently capture good schools like Malcolm X. For instance, intervening in the five percent of schools with the biggest black-white achievement gaps would, in California, pinpoint thirty-six schools, only two of which have above-average black achievement. (For the Hispanic-white achievement gap, you’d identify 137 schools, eight of which have above-average Hispanic achievement.) That’s not perfect, but it’s a lot better. You could also decide that only schools with large achievement gaps and low black or Hispanic performance would be identified as “challenge schools.” Surely there are other ways to address this issue too.
The “achievement gap” rhetoric has always been problematic because, if taken seriously, it assumes that high white performance is something to be avoided; nobody believes that to be the case. We’d be much better off talking about raising the achievement of all students—and putting in place policies that encourage exactly that. But if we must legislate gap-closing, let’s make sure we do it the right way: by encouraging higher performance among black and Hispanic students, not by applauding lackluster achievement among whites.