Choice serves families at opposite ends of the social-class spectrum in America: about 15 percent of low- and high-income parents opt to exit their neighborhood school for a public or private school of choice (NCES 1995). More than 20 states have now approved charter school legislation. Publicly supported voucher programs operate in Milwaukee and, beginning last month, in Cleveland. The private sector supports voucher or scholarship programs in 20 cities nationwide (Moe 1996).
But rising enthusiasm over school choice has far outpaced careful attempts to assess the concrete effects of choice experiments: charters, magnets, and voucher programs. Harvard University recently sponsored a three-year-long seminar series attended by researchers who are evaluating choice programs. Are these experiments working? Which children benefit, and which ones get overlooked? Can we make choice fairer for all students? The empirical results of these studies appear in Who Chooses, Who Loses? Culture, Institutions, and the Unequal Effects of School Choice (Fuller and Elmore 1996).
Three Questions
Our project examined three questions that empirical evidence can help answer—with an eye on how to improve the design of choice programs.
1. Which families exit their neighborhood school and exercise choice? Choice programs are enormously popular, especially in low-income and working-class communities, where parental dissatisfaction with local schools is the greatest. Yet even within these neighborhoods, better educated parents and those already involved in their children's schooling participate more frequently in choice programs. This is not surprising. What's notable is that this self-selection and sorting, based on social class and family practices, occurs even among low-income and working-class families.
In Milwaukee, even though three-fourths of all participating students come from female-headed households, maternal education levels vary, as does the extent to which mothers supervise their children's homework. More than half of the parents who participated in the choice program had attended college, compared with just 30 percent of those who did not participate.
How children are selected by choice schools can exacerbate the inequality between participants and nonparticipants. In the San Antonio voucher program, for instance, administrators use the child's prior academic performance in determining access to a limited number of slots in the bicultural choice schools. The reading scores of students admitted to these schools are twice as high as those of children who do not apply to choice schools. This creaming-off of the strongest students serves to concentrate high-achieving students in new choice schools and leaves behind lower achievers in neighborhood schools.
Choice does little to desegregate schools or to break down ethnic enclaves of families. It can even reinforce racial separation. In Milwaukee, several of the nonsectarian community schools now supported by vouchers are ethnocentric in their identity and their curriculum. Nearly 100 percent of students enrolled in Milwaukee's Urban Day and Harambee School are black, and the student population at Bruce-Guadalupe Community School is nearly all Hispanic. This unanticipated resegregation effect is troubling.
2. Do a thousand organizational flowers bloom under liberalized school markets ? Pro-choice financing structures, under some conditions, do stimulate the creation of innovative schools. We know that charter schools, magnets, and voucher-supported community schools differ from typical neighborhood schools. Yet we have no evidence that choice schools boost children's achievement. We simply do not know enough about what is happening in the classroom. In short, there is no reason to assume that different is necessarily better. As we see in the marketplace every day, products and companies fail. We can expect the same with schools. The difference is that it is much easier to write off a lost investment than a lost educational opportunity for children.
One curiosity is that choice schools, despite their rebellion against school bureaucracy, are usually founded on publicly financed organizations. In Milwaukee, for example, the 15 community schools that now accept voucher students grew out of pre-existing youth centers, preschools, or nonprofit community action organizations. California's 93 charter schools reflect a related pattern. The most common force behind these schools is a small cluster of public school teachers or an activist principal who has grown dissatisfied with bureaucracy.
3. Does school choice improve student performance or boost parents' commitment to their child's development ? If parent satisfaction were a reliable indicator, the answer would be an emphatic yes. A recent U.S. Department of Education study found steady growth in parental demand for magnets and waiting lists for these schools in many urban districts. More textured data gathered from the Milwaukee voucher program show that parental satisfaction does rise after one's child enters a choice school, and this is reflected in parents' pressure to expand the program (Witte 1996).
What's troubling is the lack of any compelling evidence that entry into a choice school actually results in measurable achievement gains. In Milwaukee, reading gains during the first year were somewhat higher among choice students than among a control group of similar low-income students. But in the second and third years, results were quite inconsistent, especially in mathematics achievement. One weakness of the Milwaukee data is that they are aggregated for all choice students. We need to look at individual schools and classrooms to assess the success of actual school management and classroom practices. In contrast, the first-ever longitudinal study of achievement (over three years) in magnet schools, completed earlier this year at the University of Wisconsin, shows significant gains in high school compared with students attending neighborhood schools (Witte 1996).
In San Antonio, choice participants did display markedly higher learning curves during the first year, relative to their counterparts who remained in neighborhood schools. These encouraging gains appear to be significant even after taking into account children's family background and prior achievement. It is difficult, however, to distinguish improvements resulting from better schooling from prior effects that result from the selection of those students with the best record of prior achievement.
We have no direct evidence on what happens to the performance of children left behind in neighborhood schools. We do know that the parents of these youngsters are less educated and less involved in their children's schooling. Earlier research has identified the problem of negative "peer effects" on classroom performance: When low performers are concentrated in the same classroom, they tend to suppress one another's achievement (for example, Bryk and Raudenbush 1992).
Evidence emerging from the preschool arena—a vibrant mixed market of 48,000 small organizations serving 5 million children—offers important lessons for the future of school choice. As with the Milwaukee program, most preschool subsidies go to low-income families, either directly through parental vouchers or indirectly as per-capita grants to nonprofit child-care organizations, including church-affiliated institutions. This strategy, which can be characterized as "choice with equity," appears to be largely successful. In the space of three decades, access to preschooling has become fairly equitable, reaching about half of all 3- to 5-year-olds from low-income and middle-class families, including programs based in public schools (Fuller et al. 1996). The related issue of whether preschool quality is distributed fairly is more difficult to answer.
Toward Greater Equity
What we've learned thus far can help us rethink how to better harness school choice in order to yield more consistent gains in children's learning and ensure greater equity. Findings to date point to the risks of rushing headlong down the path of unrestrained school choice programs. In general, these tend to reward parents who are already most committed to their youngsters' education, leaving behind children who receive the least help and encouragement from their families. Achievement is likely to rise for those in the choice schools and fall for those left behind in the neighborhood schools. No evidence yet demonstrates that choice sparks inventive changes inside classrooms.
Democratic government must certainly create incentives that reward parents who work hard to push their offspring up the ladder of economic and social mobility—without harming children whose parents are less committed or unaware of how to attend to their youngster's schooling. And political leaders continue to search for reforms that will truly touch the day-to-day management of schools and the vigor and rigor of classroom life. Toward these ends, can we modify the design and implementation of choice programs to lead to more impressive and equitable gains?
One positive step would be to encourage all parents within a town or city to actively consider their school options and express a discrete choice. In Cambridge, Massachusetts, for example, each year all families must list their top three schools. Public authorities then balance these individual choices against the state's interest in preventing overcrowding and in providing ethnic diversity within each school.
A simpler option is to randomly select from among families requesting choice schools, rather than use students' prior achievement or parents' initiative as criteria. Using prior school achievement to select "choice students" does just that: it skims off the most able and isolates kids who are receiving the least amount of support from their parents and teachers.
Better information about school options, employing media that are particularly effective in reaching different ethnic groups, may broaden interest in school choice to a wider set of parents. In Montgomery County, Maryland, one study found that Hispanic parents were half as likely to understand their magnet school options, compared to white parents.
In addition, choice schools should have clear plans for how they will alter classrooms and the teaching-learning process. Cosmetic factors have been enormously important in selling charters, small schools, magnets, and voucher schools. But their leaders should be accountable for explaining precisely how they will enliven the classroom process, as should public educators.
We must evaluate concrete effects on children and families. As the school choice movement has gained political adherents, tough questions about real effects are being asked less frequently. We have promised poor and working-class families lots of things since the 1960s. Choice is being advanced as another medicinal panacea for the ills linked to poverty. For a small investment, we can build additional evidence on which types of local programs yield what kinds of effects on parental commitment and children's achievement.
One fact is clear: The sprouting of choice initiatives in education and family policy offers a vibrant response to the increasingly pluralistic character of American society. Whether pro-choice policy fundamentally alters realities at the grass roots remains to be seen. Choice, ideologically, advances our Jeffersonian attraction to decentralized government as a way to advance rugged individualism. But Thomas Jefferson also viewed the common school as a balance wheel that would nurture shared values and a richer understanding of difference and human diversity. If the common school cannot provide this balance, we must find common ground through new forms of schooling.
Most fundamentally, real progress will occur only when two crucial, human-scale organizations, the classroom and the family, make gains. Effective education and the careful upbringing of children stem from subtle interactions among child, teacher, and parent. School choice proponents are just beginning to look into this core and ask how we can approach teaching and child development anew.
