Earlier this month, Education Next hosted a forum on the research concerning private-school choice. One of the topics discussed was the question of who participates in these programs.
Do means-tested choice programs serve only the most advantaged among the disadvantaged, or do they disproportionately attract disadvantaged students even among the eligible populations?
Patrick Wolf explained that “private-school-choice programs disproportionately attract students from disadvantaged backgrounds,” noting that the choice participants are “considerably more likely to be low-income, lower-achieving, and African American, and much less likely to be white, as compared to the average public-school student in their area.”
By contrast, Douglas Harris claimed, “Even when limited to low-income populations, though, vouchers tend to serve a socioeconomically advantaged portion” of the eligible student population.
Of course, it’s possible that both of these claims are true, but this raises a question: do means-tested choice programs serve only the most advantaged among the disadvantaged, or do they disproportionately attract disadvantaged students even among the eligible populations?
At least 10 studies have examined the relative advantage of children that applied to a private-school-choice program relative to the population of eligible students in the same location. As shown in the second to last column in the table below, only two of the 10 studies supported the claim that students applying to the programs were, on average, more advantaged. Four of the remaining studies found that applicants were less advantaged than the eligible population.
For example, Florida State University’s 2017 study of the Florida Tax Credit Scholarship Program found that participants were four percentage points less likely to be white, one percentage point more likely to qualify for free lunch, and had prior math and reading scores that were two to four percentile points lower than eligible students that did not participate in the choice program. Each of these statistics suggest that the applicants were less advantaged than the eligible population overall.
Four of the 10 studies found mixed results. For example, a 2005 study of the nationwide Children’s Scholarship Fund program found that applicants were more likely to come from two parent households, have parents who were involved in their education, and have more educated mothers. However, the same study found that applicants were more likely to be minorities, had lower income levels, and had parents who were less satisfied with their assigned district schools.
Table 1: Private School Choice Program Selection (Applicants versus Eligible Students)
|Figlio, Hart, & Metzger (2009)||Florida Corporate Tax Credit Scholarship Program||Applicants are more likely to be minority, more economically disadvantaged, and lower prior math and reading test scores||Less advantaged than eligible population||Less advantaged than eligible population|
|Kisa et al. (2017)||Florida Tax Credit Scholarship Program||Participants have lower prior math and reading test scores, more likely to be minority, more likely to be free-lunch eligible, less likely to be from “A” or “B” schools, less likely to be English-language learners||Less advantaged than eligible population||Less advantaged than eligible population|
|Hart (2014)||Florida Tax Credit Scholarship Program||Applicants have lower prior test scores, more school violence, more out-of-school suspension, lower ratings from principals and teachers||Less advantaged than eligible population|
|Anderson & Wolf (2017)||D.C. Opportunity Scholarship Program||Applicants are more likely to be disadvantaged based on unobservable characteristics||Less advantaged than eligible population|
|Fleming et al. (2013)||Milwaukee Parental Choice Program||Applicants’ parents have lower income but have more years of education||Mixed||Mixed|
|Paul, Legan, & Metcalf (2007)||Cleveland Scholarship and Tutoring Program||No differences||Mixed||Mixed|
|Howell (2004)||New York City’s School Choice Scholarships Foundation Program||Applicants more likely to be on welfare, more likely to have mothers who were employed, more educated mothers||Mixed||Mixed|
|Campbell, West, & Peterson (2005); same data as Peterson, Campbell, & West (2002) and Howell & Peterson (2006)||Children’s Scholarship Fund||Applicants mothers have more years of education, more likely to be from two parent household, more parent involvement, but more likely to be minority race, lower income levels, parents are less satisfied with the public school’s academics, safety, location, discipline, values, and overall.||Mixed||Mixed|
|Martinez, Godwin, & Kremerer (1995)||San Antonio CEO Voucher||Applicants’ parents have more years of education, higher income, higher employment, more likely to be in a two-parent household||More advantaged than eligible population||More advantaged than eligible population|
|Chakrabarti (2013)||Milwaukee Parental Choice Program||Applicants’ parents have more years of education, more involved, and students have higher prior test scores||More advantaged than eligible population||More advantaged than eligible population|
Notes: Purple boxes indicate evidence of negative selection (i.e. applicants are less advantaged than eligible students). Orange boxes indicate evidence of positive selection (i.e. applicants are more advantaged than the eligible students). Yellow boxes indicate mixed results. White boxes indicate that the study does not specifically assess the question.
Of course, Harris didn’t claim specifically that studies showed applicants were more advantaged overall. He stated that applicants were “socioeconomically advantaged.” Socioeconomic status is most-often defined as a combined measure of education, income, and employment. As shown in the table’s last column, focusing solely on socioeconomic status reveals a similar - although less clear - picture. Two studies previously indicating negative selection drop from the review because they don’t explicitly examine a measure of socioeconomic status.
Only two of the eight remaining studies support the claim that vouchers serve a socioeconomically advantaged set of the eligible population. Two reveal the opposite conclusion, while four remain mixed.
The preponderance of the evidence does not indicate that the most advantaged families apply for school choice programs when given the option. On the contrary, the evidence seems to suggest that the families that are most in need of school choice - minorities, low-income households, and students with lower prior academic achievement - are more likely to apply.