By Joshua M. Cowen, David J. Fleming, John F. Witte, and Patrick J. Wolf
Conclusion
This discussion informs the literature on school vouchers by providing evidence from a program that has matured from a small pilot program to a widely available option for which most students are eligible. Most of the evidence for the efficacy of school vouchers comes from relatively small, one-time interventions. The data in this study offer the perspective of a group of students for whom the status quo is not the public sector they leave when offered a voucher as part of a small pilot initiative, but rather the private sector they leave after having made use of a large-scale and fully developed public voucher program.
In the voucher literature, the all-important question is that of selection—who, given the chance, chooses private apart from public education—and the results we present here provide new evidence that selection is not a single, static determinant of school sector but a fluid and ongoing process that develops over time. This evidence builds on earlier scholarship, most notably in Howell’s (2004) research on a randomized voucher trial in New York City. Such trials are, however, primarily concerned with obtaining internally valid estimates of voucher effects on student achievement, and necessarily sacrifice the benefits of external validity, generalizability, and scale. The data here are a representative sample of all students from a school voucher program that has been “scaled up” to a full-fledged program offering private education to students at a public cost of tens of millions of public dollars per year. In that sense, we are able to present evidence that even after such a large-scale program has been implemented for almost two decades, there does not develop a set of separate educational sectors where, for students, “never the ‘twain shall meet.” To the contrary, this evidence suggests that for some students at least, private education is a transitory condition, not necessarily a long-term alternative to public school.
Through this perspective, we see prior achievement and race as the primary student-level characteristics associated with attrition from the private sector. We have also shown that parents concerned that the needs of their child are not met by their private school are more likely to move to the public sector. Although several voucher studies have determined that on average, parental satisfaction increases as a result of participation in a voucher program, the results here indicate—at least in our data—that such satisfaction is not universal among voucher users. Indeed, the results of a larger array of survey data reported elsewhere (Witte, et al., 2008) suggest relatively similar levels of parental satisfaction in both sectors in Milwaukee. It is possible that changes in satisfaction are affected by winning (or, at least, not losing) a voucher lottery, as the largest impacts regarding the satisfaction of voucher parents come from the initial year of voucher participation and satisfaction levels degrade somewhat as families experience more of the reality of private schooling itself (e.g., Howell et al., 2006, pp. 172-177).16
Among school-level attributes, the most important appear to be the racial makeup of the school and the extent to which a student’s school serves primarily voucher or nonvoucher private school students. Students in schools with higher percentages of Hispanic students are more likely to leave, as are those with higher rates of African Americans, although the latter relationship is insignificant after accounting for the share of voucher users (a disproportionately African American group compared to typical Milwaukee students). For most observations in our sample, the greater the proportion of students in a school using a voucher, the more likely a student is to leave for the public school.
Private schools serving primarily or entirely voucher populations are by definition dependent on voucher revenue to remain open. These schools, less established with respect to their peer private schools that are not dependent on public funds (Kisida et al., 2011), are precisely those we would expect to serve a higher risk and ultimately more mobile student population. They are also the majority of participating schools in the MPCP. Practically speaking, the voucher program is means-tested and directed toward the poorer students even within a relatively poor urban area. A higher proportion of voucher students indicates a greater share of low socioeconomic students. One implication is that we should expect high rates of turnover to be among the features of widely available voucher programs if the majority of participating schools are depending on the program to exist. Whatever the explanation, these results suggest that future voucher work should consider not simply a single, monolithic private sector, but the ways in which different private schools may provide varying educational opportunities to students using public money to attend them.
Given the environment that we have described, sustained participation in a large-scale voucher program is directly related to student attributes, including academic success within it. One troubling implication of these results is that voucher programs are as prone to precisely the same patterns of mobility that so concern observers of traditional urban schools. If poorly performing and minority students are likely to leave voucher schools, just as they are likely to move frequently within the public sector, even a promising voucher initiative may have limited potential to improve education for students most in need of it (Wolf, 2010). This is especially problematic if, as some school choice advocates suggest, students need time in voucher schools before they realize positive academic impacts. What we may be observing in Milwaukee is a high-choice context in which private education is available to most students, with the public schools serving not only students who are unwilling or unable to take advantage of that option, but also those who attempted to do so but were unsuccessful. This suggests that there are some at-risk students for whom a large-scale voucher program may represent an opportunity, but others for whom a voucher is not a solution to persistent educational difficulty.
Source: Am Educ Res J 2012;49 231-256
http://aer.sagepub.com/cgi/content/abstract/49/2/231
No comments:
Post a Comment