Would School Choice Segregate Well-Off Students?

This post originally appeared on Jay P. Greene’s Blog.

The confirmation of Betsy DeVos as the nation’s Secretary of Education is shining a national spotlight on educational choice. It has also drawn attention from school choice skeptics and opponents and a flurry of criticisms about choice with it.

A recent report by Halley Potter of the Century Foundation claims that educational choice increases ethnic segregation. Never mind that it misinterpreted a study on Louisiana by Anna Egalite, Jonathan Mills, and Patrick Wolf (you can find Egalite’s rejoinder here).

But ethnic segregation is not the only kind of segregation about which concerns are raised. Opponents also argue that choice policies will lead to “creaming,” in which well-off students disproportionately choose to participate in choice programs, leaving public schools worse off.

These claims are making a prediction about which students and families will respond more to the offer of an ESA or voucher. Economists use the term elasticity to describe this responsiveness. In the context of school choice, for a given change in the price of private schooling (which is what ESAs and vouchers essentially do), a higher elasticity means that a larger number of students will respond by enrolling in or leaving a given school.

The analytic challenges involved in estimating elasticities of demand for private schooling are substantial because it is difficult for researchers to obtain causal estimates. Fortunately, a team of researchers conducted a study which speaks to the issue of school choice’s probably effects on this kind of segregation, and its analysis produced some interesting findings.

Susan Dynarski of the University of Michigan, Higgy winner Jonathan Gruber of the Massachusetts Institute of Technology, and Danielle Li of Harvard estimate the price elasticity of demand for private schooling. This team of researchers could observe differences in responsiveness to price among families with different backgrounds.

In a finding with huge relevance to the school choice debate, they conclude that families “with lower levels of parental education are about over four times as price elastic than other families.” In the words of the researchers:

The results indicate that vouchers would tend to increase the share of private school students who come from families with relatively low levels of parental education.

Moreover:

These results suggest that vouchers would increase the representation of low- and middle-income families at private schools.

Other model specifications “indicate that families with the highest predicted probability of private school attendance are the least sensitive to price” (p. 29).

The authors conclude:

These results suggest that a voucher program would disproportionately induce into private schools those who, along observable dimensions such as race, ethnicity, income and parental education, are dissimilar from those who currently attend private school. This is in marked contrast to the assumption made in previous studies… that the new students that vouchers would induce into private school would look demographically similar to current private school students.

…Overall, it is those families who (along observable dimensions) are least like the current population of private school customers that are most sensitive to price, suggesting that vouchers would substantially alter the socioeconomic composition of private schools.

While this study provides one useful data point for policy makers who are considering introducing or expanding educational choice in their states, policy makers should also consider information generated by studies that have already measured the impact of educational choice on segregation. The most rigorous studies available examined Louisiana’s voucher program, where researchers found that the program reduced segregation. Other studies found that school choice programs move students into less segregated schools in D.C. and Cleveland; results in Milwaukee either find no difference or suggest a positive effect. When one weighs the overall evidence about the impact of private school choice on segregation, a picture develops where findings from empirical research on school choice programs bolster the predictions suggested by Dynarski, Gruber, and Li’s findings.

These studies suggest that empowering parents to choose would change private schools. And empirical research on private school choice’s effects on segregation are largely positive. For those who value diversity and empowering parents, increasing educational options is a good thing.