This is (sort of) inconvenient for the Department of Justice: The Louisiana State school-voucher program it recently filed a suit about, alleging that the program is hampering integration efforts, is actually promoting integration and reducing segregation, according to a new study. The work, by the University of Arkansas’s School Choice Demonstration Project, adopts a slightly different definition of segregation than the DOJ does and looks at the program’s overall effects; it’s probably not legally relevant. But their finding actually does a much better job than the DOJ did at assessing whether the program is a good thing for desegregation efforts overall: It tries to measure the net impact of the program on integration and segregation, while the DOJ merely identified 34 schools where they say the problem got worse and sued on those grounds, without measuring whether vouchers improved the situation in other places with desegregation orders.
Because of various data limitations, the study’s authors had to take a subset of the students actually participating in the program (about 500 from about 5,000), which provides vouchers for low-income students who are in failing schools to pay tuition at private schools. They find that the vast majority (83 percent) of students transferred reduced the level of segregation at the schools they leave, and the impact on the schools they arrive at is, overall, negligible. And schools in districts under the DOJ’s desegregation orders that lose students via the program and the ones that gain them do better, meaning that the private schools where students go to use the vouchers are actually becoming, of their own free will, more integrated because of the program.
The study’s conclusions are in line with the broader literature, which finds that segregation is either unaffected or reduced by school-choice programs. But the overall effects simply aren’t that great: The DOJ’s lawsuit, for instance, identifies 34 schools where the student transfers “impeded the desegregation process.” That’s out of a program with more than 5,000 students participating (it’s not clear how many schools they come from). And the two cases they cite, which are presumably some of the worst, see six black students and five white students leave them, respectively, and that supposedly “reinforces” the racial identity of each school – despite this amounting to just a one- or two-percentage-point shift in racial breakdown.
So while the definitions used by the Arkansas study were slightly different and it’s certainly possible this study has some sampling problems, the program could well be moving schools under desegregation orders, on average, toward the DOJ’s definition of integration. We don’t know, because all the federal lawsuit aims to do is show that the program disrupted “desegregation efforts” in a few places, rather than looking at its net effects.
The lawsuit demands that the program be halted until the state can demonstrate that the program is being run in such a way as not to interrupt desegregation efforts anywhere — which may literally mean preventing black or white students from getting vouchers in order to avoid fiddling with the racial balance at the failing schools they want to escape.
Jonathan Strong had a piece for NRO at the beginning of September examining the feds’ motivation for blocking the program, given that it has an unclear impact (until this study) on racial integration efforts overall — some see it as a politicized attack on school choice. That isn’t the only line of attack on Jindal’s comprehensive, dramatic education reforms (which I wrote about last spring): Courts ruled that Jindal had to find funding somewhere other than the existing public-school budget, which obviated one school district’s objection (sustained by a court) that the program sapped the funding for specific desegregation programs, which has federal protection. Again, all this to oppose a program which the best evidence now suggests encourages desegregation on average.
As I said, the study used a slightly different definition of segregation/integration than the DOJ does, but probably a better one: It defined a school’s level of integration as the degree to which its racial composition varies from the surrounding area — which makes sense, if the idea is that you don’t want people to be able to segregate themselves into schools less homogenous than the larger community they actually live in (specifically, the area they used is what’s called a Core Based Statisticall Area, which they argue is ”the geographical area from which a school could reasonably be expected to draw students in the absence of legal or political boundaries” — the largest is about 220,000 people). The DOJ instead just looks at whether schools hew to the racial breakdown of the school district in which they reside, on the grounds that certain schools within school districts are identified as “black” or “white” schools. This can allow de facto geographic segregation and in fact insists on preserving it, because a school district could be, say, 90 percent black, and the desegregation order will restrict any changes that will move schools in the district away from being anything but 90 percent black (obviously this would not be a problem for everyone but it is for the stated goals of the DOJ’s desegregation efforts).