The Trouble With Desegregation

A couple of weeks ago, on MLK Day, Dana Goldstein noted that American schools are more segregated today than they were 43 years ago when Dr. King was killed. She applauded Arne’s Duncan’s support for the widely-praised and now-threatened integration program in Wake County, NC, but then said “The problem is that Arne Duncan’s words of support for the Wake County integration plan have never been backed up by Obama administration policy.” Ezra Klein followed up, noting that “the education reform movement more generally” doesn’t seem to be making desegregation a top priority.

Both observations are accurate. But the lack of attention to education-based desegregation strategies doesn’t stem from a lack of commitment to the goal, or from disbelief in research supporting desegregation’s positive effects. Rather, there are some basic structural and logistical barriers in place that nobody has figured out how to overcome.

The Wake County program, which caps the percent of poor students in a given school, is justly praised. Unfortunately, the new Tea Party-backed school board chair believes that ”If we had a school that was, like, 80 percent high-poverty, the public would see the challenges, the need to make it successful,”  This, of course, is why the Detroit Public School system is a world-renowned model of educational greatness.

But Wake County is unusual. It has two necessary conditions for district-based desegregation: size and diversity. You can’t manipulate the racial composition of your high schools if you only have one high school, or if all the students in your district are the same race. Wake County has size and diversity in spades. In fact, it’s the largest majority-white school district in America and only barely so: 52% of students are white. So Wake has a lot of schools to work with and an almost perfect aggregate racial balance. Montgomery County, Maryland, the other positive example Goldstein cites, is very similar–also one of the 20 biggest schools districts in the country, also incredibly diverse.

Most students don’t attend school districts like this. The average elementary school has about 470 students and the average secondary school has about 810. So let’s say, at an absolute minimum, for a school district to even think about Wake-style desegregation, it needs two high schools and eight elementary and middle schools, or 5,380 students. (Wake County has 163 schools and more than 140,000 students.) To be generous, we’ll round down to 5,000. The district also needs diversity. If the goal is integrating white and non-white students, let’s say we need at least 25% of the students to be white.

Most black students–54%–don’t attend school districts that are this large and diverse. For Hispanic students the proportion is even larger: 63%. Tighten the assumptions to, say, 10,000 students and 35% white and the numbers drop even further: 72% and 80%, respectively. Successful desegregation initiatives in districts smaller than this are few and far between. Indeed, are there any?

Most black and Hispanic student attend school districts that are either large and not diverse (Los Angeles, Chicago, Miami-Dade, Houston, Dallas, PG County, and Memphis are all less than 10% white while New York City and Philadelphia are under 15%) or not large. With the media attention given to big-city districts and the huge county-wide districts surrounding Washington, DC and other large east coast metropolitan areas, it’s easy to forget that of the nation’s 14,000 school districts, less than 10% enroll as many as 5,000 students.

That leads the conversation to desegregation policies that cross districts. After all, many high-minority districts sit inside larger metropolitan areas that are much more diverse. The fact that Wake County (atypically) encompasses all of such an area creates the conditions necessary for its desegregation program. But this also creates a new set of challenges.

First, it often means moving students non-trivial distances from their homes to schools and back again every day. That carries a significant non-educational time and money cost: resources spent on buses and trains aren’t being spent on better teachers, and that assumes you have buses and trains to begin with. Second, as Erin Dillon found a couple of years ago after making some commonsense assumptions about school expandability and using GIS software to analyze school enrollment, transportation network, and district boundary data for some of the biggest states, “only a limited number of students in a limited number of locations are likely to benefit from interdistrict choice.”

More broadly, a “school district” is basically just a geographically distinct area in which people pay a certain tax rate and make various educational decisions via locally-elected representatives. If students start crossing district lines in large numbers, the whole concept of “school district” pretty much falls apart, because “district line” doesn’t mean much other than “boundary students can’t legally cross.” That’s why Kelly Williams-Bolar is going to prison. Nobody is arguing she didn’t break the law. They’re arguing-and I agree–that a just society wouldn’t put parents in the position to choose between good schools and jail. But those lines are there right now, and they matter. What if, for example, a large number of minority students enroll in a neighboring district where their parents have no electoral representation? On some level, political accountability, resources, and attendance have to align.

So it’s not that Arne Duncan and education reformers don’t care about desegregation. The problem is that the legal and logistic barriers to creating anything like a comprehensive, effective national policy around the issue are very high. The best-known examples don’t travel all that well, and it’s not a coincidence that they are generally local policies. Where students go to school is an intensely personal decision for parents, one they often organize their whole lives around. I’m a pretty hardcore local control skeptic but this is an area where shifting decision-making up to state and federal governments strikes me as unusually hard to pull off.

So those who see desegregation as a key education strategy (there are, as Goldstein notes, distinct housing-focused strategies as well) need to articulate a more detailed, plausible way forward for national policymakers to adopt–one that likely involves going after the underlying district-based governance structure that has defined K-12 education pretty much since it began. I haven’t seen such a plan yet, and until it’s formulated I expect desegregation will continue to be a minor element of education reform.