School Vouchers Are Not a Cure For Segregation: Part I

Here are links to all five parts of the series:

Part I

Part II

Part III : New Orleans

Part IV: Milwaukee

PartV: Washington DC and Conclusion

* * *

How white are America's private schools? Really white:

Students in the nation’s private schools are disproportionately — and in some states overwhelmingly — white.

While that’s not entirely surprising, a new analysis from the Southern Education Foundation quantifies the continued segregation of white students in private schools, particularly in the South, where private-school enrollment jumped in the 1950s and 1960s as white families sought to avoid attending integrated public schools.

Should public tax dollars, through the funding of private school vouchers, but used to subsidize such a segregated system? The author of this report doesn't think so:

The author of the analysis also puts forth a provocative argument: Because of this historical pattern, private schools that take public money (via vouchers and voucher-like programs) should not be able to select the students they admit. Instead, those schools should have to admit anyone who applies, just like public schools do, said Steve Suitts, who wrote the study as a senior fellow at the Southern Education Foundation.

“The public-school system is built on the bedrock notion that we want each child to have a chance for a good education,” said Suitts, now an adjunct professor at Emory University. “And if private schools do not wish to advance that national purpose, then they ought not receive public funding.”

Let's make something clear before we continue: the notion that all public schools "admit anyone who applies" is awfully strained when we consider that our country's public, district schools are highly segregated themselves. Schools segregation is largely a result of residential segregation, which is a result of years of both official and unofficial policies.

So no one should point a finger at segregated private schools and simultaneously pretend there aren't public schools that engage in the same de facto practices. But at least we can all agree that segregation in the public and private school sectors is something we need to address. I mean, no one serious would ever claim that expanding private school enrollments would actually help ameliorate school segregation...

Would they?

There is no denying history and the motives of some parents and politicians 50 years ago, who feared desegregation and were racially motivated to send their children to private schools. However, the history of 50 years ago doesn’t align with the reality of today. Private school choice programs now exist in 25 states and Washington, D.C. Through vouchers, tax credit scholarships, and Education Savings Account programs, nearly 400,000 children are accessing a private school of their parents' choice. Today’s system is color blind and largely benefits minority families. [emphasis mine]

That is Pitbull's good buddy Kevin Chavous -- always reliably reformy, and always ready to champion market-style "choice" in education. Of course, Chavous is coming to us via The 74, an outlet that never met a charter school expansion or school voucher system it couldn't love.

How does Chavous rebut Suitt's claim that private schools are largely segregated -- that the system is, instead, "color blind"? After lauding voucher programs in Alabama, Florida, and Louisiana, Chavous (as all reformy types apparently must) states outright that Suitt just doesn't care about children as much as folks who support vouchers:

These are just three examples of where Suitts gets it wrong. These are also three examples of programs designed to help low-income families – most of whom are minority. This isn’t an anomaly, this is the reality of the opportunity provided to minority students through private school choice. If Suitts had his way, poor black families would be left in Birmingham, AL to attend underperforming schools in a school district that’s 95 percent African American and reports a graduation rate 20 percent lower than the statewide average. The same would be the case in Louisiana, Florida, Georgia, North Carolina and South Carolina, but private school choice offers these families hope and the children an opportunity to attend a quality private school that would have otherwise been out of reach.

Like so many of the writers who appear in The 74 (and, for that matter, Education Post), Chavous has the uncanny ability to fling indignation towards a guy who is pointing out the facts while simultaneously ignoring the main thrust of his argument. Suitt is saying private schools, particularly in the South, are highly segregated, and supporting them with taxpayer dollars is against the interests of society if we want to move toward desegregation.

No one would argue with Chavous that it is acceptable to continue to cluster African American children in segregated, underperforming (and under-resourced) schools. The question is whether, given what we know currently about American's private schools, expanding a system of "choice" through vouchers would lead to more segregation or less.

Let's dig through the data and see if we can shed some light on this.

One reason Chavous doesn't dispute any of the data in Suitt's report is that it's quite straightforward: basic descriptive statistics about average racial compositions of schools from the NCES Private School Survey. You really can't argue with Suitt's facts -- but I do think they are well worth reexamining.

Rather than replicate Suitt's work, let me show it in a different light. We'll start with Alabama as an example (click on any graph to expand).

This histogram divides the total number of students enrolled in Alabama's private schools into ten "bins" that run across the horizontal axis. The first bin, which is the first gold column all the way to the left, represents schools where between 0 and 10 percent of the students are black. Two-thirds of all Alabama private schools students attend a school where less than 10 percent of the students are black.

I've added a red line to this and all of the graphs that follow: it represents the proportion of the total state population of children (under age 18) of the race represented by the graph (data from the American Community Survey, 5-Year Estimates (2010-2014) of the US Census Bureau). In other words, 30 percent of the children in Alabama are black. Notice how few students are found at the line? If Alabama's private schools were representative of the state as a whole, we'd see the most students enrolled in schools that were 30 percent black; instead, we see very few students are in such schools.

Here's how the same histogram looks for Alabama's white* students:

59 percent of Alabama's children are white, yet only a tiny fraction of the state's private schools students attend schools with a percentage of white students anywhere close to this average. Notice also that there are a number of private schools that enroll a very small proportion of white students (less than 10 percent). Which begs a question: what if all the black students using Chavous's vaunted vouchers are going to private schools just as segregated as the public schools they are leaving?

I've posted histograms for the other states Chavous mentions below. But let me stop here and point out that this analysis still may not be adequate for our discussion. A state is a large geographic area, full of many different regions with their own demographic characteristics and patterns. Averaging these descriptives across an entire state might give us a distorted view of the segregative trends that may or may not take place within a specific region.

We'll get to that in Part II. Until then, the other statewide histograms are below. Note the trend: very few southern private schools have student populations that mirror their state's overall racial profiles.

More to come; stand by...

Florida: 21.2 % black, 44.4% white.**

Georgia: 34% black, 46.2% white.**

Louisiana: 37.6% black, 52.1% white.**

North Carolina: 23.6% black, 54.2% white.**

South Carolina: 31.5% black, 55.1% white.**

* Throughout this series, "white" when referring to US Census data means white, non-Hispanic. NCES data reports race differently than the US Census; these are the most comparable measures. 

** Again: all percentages from the US Census American Community Survey 5-year Estimates (2010-2014), children under 18 years in households.