Document Type


Publication Date



school choice, school movement, cream skimming, student selectivity, survival model


Private school choice programs often are accused of failing to serve disadvantaged students. Critics claim that participating private schools “skim the cream off the top” by admitting only the best students and “push out” students who are the most difficult to teach. This study tests these student selection hypotheses in the context of the Louisiana Scholarship Program (LSP). We find LSP applicants are less advantaged than their public school peers regarding their family socioeconomic status and initial test scores. No consistent evidence indicates that the LSP private schools are “skimming the cream” or “pushing out” students based on their family social status or initial test scores. However, students with disabilities are less likely than students without disabilities to use a voucher initially. Students who were placed in LSP private schools that were farther from their homes or that serve a larger minority population are more likely to leave their LSP schools than LSP students placed in schools closer to their homes or that serve smaller minority populations. LSP students with better educational resources in their residential public school district are more likely to leave the LSP than students with worse educational resources. Finally, the LSP students still using vouchers after three years are more likely to have a low family income, more likely to be African American, and more likely to be female than the population of non-applicants to the program.