The Mission of the Manhattan Institute is
to develop and disseminate new ideas that
foster greater economic choice and
individual responsibility.

Civic Report
No. 38 June 2003


Vouchers for Special Education Students: An Evaluation of Florida’s McKay Scholarship Program

Jay P. Greene, Ph.D.
Senior Fellow, Manhattan Institute for Policy Research
Greg Forster, Ph.D.
Senior Research Associate, Manhattan Institute for Policy Research

Executive Summary

The McKay Scholarship Program for Students with Disabilities makes a school voucher available to any special education student in Florida public schools. This program is the second largest school voucher program in the country, and with approximately 375,000 eligible special education students it is likely to become the largest soon. Currently, 9,202 students use McKay vouchers.

This study is the first empirical evaluation of the McKay program’s performance. Based on two telephone surveys—one of parents currently using a McKay voucher and the other of parents who previously used a voucher but no longer do—this study shows that parents are much more satisfied with their experiences in private McKay schools than they were with their experiences in the public schools. This is true both for currently participating parents and for parents who have left the program.

Highlights of this study include:

  • 92.7% of current McKay participants are satisfied or very satisfied with their McKay schools; only 32.7% were similarly satisfied with their public schools;
  • Those participants also saw class size drop dramatically, from an average of 25.1 students per class in public schools to 12.8 students per class in McKay schools;
  • Participating students were victimized far less by other students because of their disabilities in McKay schools. In public schools, 46.8% were bothered often and 24.7% were physically assaulted, while in McKay schools 5.3% were bothered often and 6.0% were assaulted;
  • McKay schools also outperformed public schools on our measurement of accountability for services provided. Only 30.2% of current participants say they received all services required under federal law from their public school, while 86.0% report their McKay school has provided all the services they promised to provide;
  • Behavior problems have also dropped in McKay schools. 40.3% of current participants said their special education children exhibited behavior problems in the public school, but only 18.8% report such behavior in McKay schools;
  • Former McKay participants provide similar responses. 62.3% were satisfied with their McKay school, while only 45.2% were satisfied with their old public school. Their class sizes also dropped from an average of 21.8 students to 12.7 students. Former participants also reported that their McKay schools performed better than their public schools on almost every other measure;
  • This superior performance by McKay schools was largely provided for the same or only slightly more money per pupil than is spent in public schools. Even though the McKay program allows participants to choose schools that charge tuition above the amount of the voucher, 71.7% of current participants and 75.8% of former participants report paying either nothing at all or less than $1,000 per year above the voucher;
  • Perhaps the strongest evidence regarding the McKay program’s performance is that over 90% of parents who have left the program believe it should continue to be available to those who wish to use it.

The results of these surveys indicate that participants in the McKay program are being significantly better served by McKay schools at no additional cost to the taxpayer and no or little additional cost to their families.

*********************************************

About the Authors

Jay P. Greene
is a Senior Fellow at the Manhattan Institute for Policy Research where he conducts research and writes about education policy. He has conducted evaluations of school choice and accountability programs in Florida, Charlotte, Milwaukee, Cleveland, and San Antonio. He also recently published a report and a number of articles on the role of funding incentives in special education enrollment increases.

His research was cited four times in the Supreme Court’s opinions in the landmark Zelman v. Simmons-Harris case on school vouchers. His articles have appeared in policy journals, such as The Public Interest, City Journal, and Education Next, in academic journals, such as The Georgetown Public Policy Review, Education and Urban Society, and The British Journal of Political Science, as well as in major newspapers, such as the Wall Street Journal and Christian Science Monitor.

Greene has been a professor of government at the University of Texas at Austin and the University of Houston. He received his B.A. in history from Tufts University in 1988 and his Ph.D. from the Government Department at Harvard University in 1995. He lives with his wife and three children in Weston, Florida.

Greg Forster is a Senior Research Associate at the Manhattan Institute’s Education Research Office. He is the co-author of several education studies and op-ed articles. He received a Ph.D. with distinction in Political Science from Yale University in May 2002, and his B.A. from the University of Virginia, where he double-majored in Political and Social Thought and Rhetoric and Communications Studies, in 1995.

Acknowledgements

We would like to thank the Florida Department of Education for making parent contact information available. We would also like to thank the people at the survey research firm, Taylor Nelson Sofres Intersearch, for their assistance in conducting the polls. And finally, we would like to thank the parents of special education students in Florida who took the time to discuss their children’s experiences.

*********************************************

Introduction

The McKay Scholarship Program for Students with Disabilities provides a school voucher to any parent with a special education student in a Florida public school who is dissatisfied with that school’s performance. This study, an analysis of a telephone survey of both current and former participants in the McKay program, is the first to evaluate the program’s performance. It finds that the schools students attend through the McKay program outperform the public schools they previously attended both in terms of parental satisfaction and on a variety of objective measurements.

The average satisfaction level for current McKay participants is 92.7% for their McKay schools, as opposed to 32.7% for the public schools they previously attended. Their average class size dropped from 25.1 students per class in public school to 12.8 students per class in McKay schools. Only 30.2% report that their public schools provided all required services, while 86.0% report that their McKay schools provide all the services they promise to provide. And students are far less likely to be bothered or assaulted by other students because of their disabilities in McKay schools than in public schools—46.8% were bothered often and 24.7% were assaulted in public school, compared to 5.3% bothered often and 6.0% assaulted at McKay schools. Current participants also saw a drop in students reporting behavior problems at school, from 40.3% in public school to 18.8% in McKay schools.

It is not surprising that current participants would be better served by their McKay schools than they had been by their public schools, because otherwise they would presumably leave the program. But even families that have left the program report that they were better served on virtually every measure by their McKay schools than by the public schools their children previously attended. Their average satisfaction level is 45.2% for public schools and 62.3% for McKay schools. They saw average class size drop from 21.8 students to 12.7 students. Public schools provided all required services for 36.3% of former participants, while McKay schools provided all the services they promised to provide for 49.3%. Former participants also had fewer problems with other students at McKay schools; 41.4% were bothered often and 26.0% were assaulted in public school, compared to 20.0% bothered often and 10.2% assaulted at McKay schools.

Despite the assertions of some of the program’s critics, in almost all cases there were no significant differences in program outcomes based on the participant’s race, income, or severity of disability. Perhaps the strongest evidence regarding the McKay program’s performance is that over 90% of parents who have left the program believe that it should continue to be available for those who wish to use it.

Background

In 1999, as part of a larger package of reforms aimed at extending educational choice to students who weren’t receiving an adequate education in Florida public schools, the Florida legislature created the McKay Scholarship Program for Students with Disabilities. Named for then-Florida Senate President John McKay, himself the father of a child with disabilities, the McKay program provides a school voucher to special education students. A student must have been enrolled in a Florida public school for one year to be eligible to participate. The voucher is good at any private school, religious or non-religious, that meets certain minimal requirements, such as financial soundness and compliance with nondiscrimination regulations. The amount of each student’s voucher is equal to the total cost of educating that child in public school. After a one-year demonstration period, the program was fully implemented during the 2000-01 school year.

Although the McKay program is different from most other school choice programs in that it specifically serves special education students, it must still be considered one of the nation’s most important school choice programs. Since all Florida public-school special education students are eligible to participate, the McKay program has the largest pool of eligible students of any school choice program in the United States—during the 2001-02 school year there were 374,834 special education students in Florida public schools. Currently, 9,202 students are using McKay vouchers.

Advocates of the McKay program argue that it brings the benefits of school choice to special education students, who may be more likely than other students to lack good educational opportunities in the public school system (for example, see Salisbury 2003). When parents are unable to afford a private school education, advocates suggest, they are stuck with whatever education their local public school system provides. This may well be an adequate or even excellent education, but when it is not, parents have little recourse—teachers answer not to parents but to the public school bureaucracy, which is large and not easily moved. More affluent and sophisticated parents of special education students may hire lawyers to compel public schools to comply with federal laws protecting the disabled and provide their children with a better education. But advocates contend that the expense and adversarial nature of litigation deter even many sophisticated parents from pursuing it, and even those who pursue litigation and prevail can only force their schools to live up to a federally-mandated minimum level of service. According to the advocates’ argument, the McKay program allows parents who are satisfied with the education their children are getting in public school to leave them there, while extending to dissatisfied parents the choice to put their students into private schools where their children might get better services without having to resort to the trouble and expense of litigation.

Furthermore, the program’s advocates argue that when public schools know most parents have no alternative, those schools have no incentive to provide a better education to students. Educating a student with disabilities can be particularly challenging, so the absence of a strong incentive to do a better job can be decisive. By giving parents a choice of schools, the McKay program makes schools ultimately accountable to parents—if parents are dissatisfied with the public system, they can leave it. Advocates of school choice argue that this produces a strong incentive for public schools to provide a better education.

However, many common criticisms of school choice programs are as relevant to the McKay program as to any other. While there is no concern that the program might “cream off” the best students from public school, since it serves special education students, other criticisms do apply. One of the most important of these is concern over the level of public accountability for education received in private schools. The McKay program contains no requirements related to education outcomes—as long as a student’s parents choose to leave him in private school, that student will remain there regardless of whether he is actually learning. Critics point out that the adequacy of private services is a matter of particular importance for special education students, who require specialized—sometimes very highly specialized—education programs, including special equipment and teachers with special training.

Critics also express concern over equitable provision of services to students of all income levels. Schools participating in the McKay program are permitted to charge whatever tuition they see fit, and for special education students, whose educational needs can be expensive, private school tuition may be high. If tuition is higher than the amount of the McKay scholarship, parents must cover the difference either by finding other funds (such as other scholarships) or by paying it themselves. This leaves open the possibility that parents of low income may be unable to afford a private school even with the scholarship, while more wealthy parents who could have sent their children to private school even without the McKay program still collect the scholarship.

The President’s Commission on Special Education has recommended the McKay program as a model for reforming special education, and several states are considering adopting some kind of voucher program for disabled students. Finding out whether the program is working satisfactorily is therefore a matter of growing importance not only for Florida policymakers, but for education policy nationwide.

Previous Research

The McKay program is the nation’s first and only school choice program for special education, and despite the growing importance of the debate over school choice in special education reform, no previous empirical studies of it have been undertaken. There have been a few reports by advocacy groups, such as People for the American Way and the CATO Institute, analyzing the theoretical virtues or defects of the McKay program, but they contain no more than anecdotal evidence bearing on this issue. However, there is a significant amount of research on school choice programs serving other student populations.

School vouchers are by far the most contentious issue in education reform, and have been the subject of much study. Two of the most widely-studied questions are whether students who attend private schools through voucher programs achieve better academic outcomes than similarly situated students who remain in public school, and whether public schools respond to the challenge of vouchers by improving themselves in order to better compete with private schools for students. The hypothesis put forward by school choice advocates is that vouchers improve outcomes for participating students because they allow parents to choose the best available schools, while they improve education for non-participating public school students by providing public schools with an incentive to improve. On both questions, research suggests that school choice has beneficial effects.

The effect of vouchers on the academic performance of participating students has been the subject of five “random assignment” studies. These studies are possible when voucher programs randomly choose which parents who apply to participate will be invited to do so. This process produces two populations that are virtually identical to one another in every respect except one: whether or not they received a voucher. Random assignment studies are the most reliable of all social science studies. Since the only difference between the two groups is that they receive different treatment, if the groups produce different outcomes then we can have a high degree of confidence that these differences result from the difference in treatment between the two groups rather than to any preexisting differences in the group members themselves. Random assignment studies are rare in public policy; only one other major education policy issue has been studied through random assignment methods: class sizes (see Krueger and Whitmore 2001). The existence of five random assignment studies on vouchers represents an unusually large amount of high-quality academic research.

Virtually all of this research has found positive results for students participating in voucher programs. A Manhattan Institute study in Charlotte found that students attending private schools through a voucher program showed a 5.9 percentile point gain in math scores and a 6.5 percentile point gain in reading scores after one year when compared to students still in public school (see Greene 2001a). In Milwaukee, one study found that such students gained 11 percentile points in math and 6 percentile points in reading after four years (see Greene, Peterson, and Du 1999) while another study found an 8 percentile point gain in math after four years (see Rouse 1998) over students still in public school. A non-random assignment study in Cleveland found that students who moved from public to private schools via a school choice program improved their math scores by 15 percentile points and their reading scores by 7 percentile points (see Peterson, Howell, and Greene 1999). But a recent report by researchers at Indiana University found no significant academic benefits for student participating in the Cleveland voucher program (see Metcalf et. al. 2003).

Recent random assignment studies conducted by Howell and Peterson in New York City, Washington, D.C., and Dayton found that black students attending private schools though school voucher programs scored significantly better on a combined math and reading test than their peers in public schools. In New York, they found a 9.2 percentile point gain in the test score, in Washington they found a gain of 13 percentile points, and in Dayton they found a gain of 6.5 percentile points (see Howell and Peterson, 2002). These studies found no statistically significant gains for students of other races. Howell and Peterson explain this as a result of black students being more likely than other students to be stuck in lower-quality public schools, leaving more room for improvement through vouchers. Another possible explanation is that the much lower number of nonblack students in the subject group obscured the results for those students. A re-analysis of the New York data by Krueger and Zhu, however, finds that the academic benefits fall short of statistical significance if race is measured differently and if prior test scores are not controlled statistically (see Krueger and Zhu 2003).

These studies also find that parents participating in school voucher programs are significantly more satisfied with their children’s educations than parents of students in public schools. In the Cleveland study, about half of parents in a voucher program were satisfied with the academic program, safety, discipline, and teaching of moral values at their children’s schools, while between a quarter and a third of public-school parents were satisfied with those aspects of their children’s schools; voucher parents were significantly more satisfied than public-school parents on several other factors as well, and were not significantly less satisfied on any factors (see Peterson, Howell, and Greene 1999). The Charlotte study found that on a variety of survey questions measuring parental satisfaction, voucher parents and public-school parents reported high satisfaction with their children’s schools at rates of about one-half and one-quarter, respectively (see Greene 2001a). Howell and Peterson’s three studies also found that voucher parents were significantly more satisfied than public-school parents; in Dayton, they found that half of voucher parents were “very satisfied” with their children’s academic program, compared to 19% of public-school parents (see Howell and Peterson, 2002).

Other studies address the question of whether public schools improve in response to the challenge of school choice programs. Such programs are not large enough and have not existed long enough for studies of their effects on public schools to be definitive, but the studies that have been done so far suggest that school choice creates positive incentives for public schools to improve performance.

A Manhattan Institute study in Milwaukee found that a school where only half of students were eligible for vouchers could expect a 5 point drop in 4th grade test scores over the study period, while a school where all students were eligible for vouchers could expect a 10 point gain in test scores over the same period, and that proximity to charter schools improved public schools’ performance on 10th grade test scores; the study also found that a school district in San Antonio with a voucher program performed equal to or better than 85% of Texas school districts on tests (see Greene and Forster 2002). Also in Milwaukee, Hoxby found that schools most exposed to school choice increased math scores by 7.1 percentile points, compared with an increase of 3.7 percentile points in schools not exposed to school choice; the study found similar gains in science and language scores (see Hoxby 2001). In Florida, where students become eligible for vouchers if their school receives two “F” grades within four years on a statewide test, a Manhattan Institute study found that schools with one “F”—schools facing the prospect of competition from vouchers if they did not improve—made exceptional test score gains, far outdistancing schools in all other categories (see Greene 2001b). Hoxby has also found improvement in public schools exposed to competition from charter schools and from other public school districts (see Hoxby 2001 and Hoxby 1998). A study of century-old voucher programs in Maine and Vermont found that if a town one mile away from a given school started offering vouchers, that school could expect a 3.4 point increase in test scores, a gain of 12% over existing scores (see Hammons 2001).

Method

One major obstacle to evaluating the success of the McKay program is the impossibility of a random-assignment research design. Without randomly assigned treatment and control groups it is very difficult to identify a population against whom the results for McKay students should be compared. Another obstacle is the difficulty of directly measuring academic outcomes for disabled students. Although testing of some kind is usually possible, interpreting the results is difficult given the absence of appropriate comparison groups and given that most tests are not well designed to capture the special skills and deficits of disabled students. Also, there are non-academic education outcomes that are particularly important for disabled students—victimization by other students, for example—that a study of test scores wouldn’t reflect. Given these constraints, the most effective way to measure the performance of the program is to survey parents who are or have been McKay participants and ask about their experiences. This gives us a reasonable measurement of how well the program is serving its participants.

It is important to include former participants in this kind of survey. After all, parents who are dissatisfied with their experiences in McKay schools are going to be likely to drop out of the program, and any survey that includes only current participants will not include such parents. Of course, if we find that the program provides satisfactory services for the over 9,000 students currently participating, that is certainly a significant finding. But we can get an even more meaningful picture of the program’s performance if we also consider the experiences of those who have dropped out of the program.

We began by obtaining contact information from the Florida Department of Education for families participating in the McKay program in the 2001-02 and 2002-03 school years. By comparing the two lists, we developed a list of families that participated in the program in 2001-02 but were not participating in 2002-03. We then selected two random samples for our telephone poll: a sample of families from the list of 2002-03 participants, and a sample of families from the list of those that participated in 2001-02 but not in 2002-03. The list of current participants contained 9,202 students, while the list of former participants contained 1,050 students; our survey samples contained 600 and 215 students, respectively.

Our survey asked the parent of the participating student (where possible, the student’s mother) a set of questions about the family’s experience in public school before it participated in the McKay program, followed by a set of questions about its experience in the McKay school. Only students who have attended public school for a year are eligible to participate in the program, so every participating family has had both a public school experience and a McKay school experience. The survey also collected demographic information to supplement the information provided to us by the Florida Department of Education. The survey and results are included in the appendix.

Once we had the survey results, we broke them down into results for current participants and former participants. We report the overall results for each of these two groups, along with results for several subdivisions within these groups. One of these subdivisions was between students with mild disabilities and students with more severe disabilities. In the state of Florida every child is classified not only by type of disability (blindness, deafness, learning disability, etc.) but also by severity of disability (Matrix Levels 1-5). Matrix Level 1 includes students whose disabilities are mildest, while Matrix Level 5 includes those whose disabilities are most severe. Since roughly half of all students are in Matrix Level 1, we reported separate results for students in Matrix Level 1 and students in Matrix Levels 2-5. We also reported separate results for household incomes above and below $30,000 per year, and for white and nonwhite students.

Some questions concerned the student’s Individual Education Plan (IEP). Public schools must develop an IEP for each special education student in consultation with the student’s parents. If parents believe their student’s IEP is inadequate, they can sue the school to get it changed. The IEP process is designed to give parents a voice in the education of disabled students. However, there is widespread concern about the adequacy of the IEP process. One issue of particular importance to the McKay program is that private schools are not required to provide an IEP; some do and some don’t. For this reason, we asked parents about their satisfaction with the process for developing an IEP at their public school, but we did not ask a corresponding satisfaction question about their McKay school. We also included a set of questions that asked parents about their experiences with and without IEPs in their public and McKay schools.

Our original intention was that parents would also be asked to rate their overall satisfaction with both the public school and the McKay school on a scale of one to ten. However, due to an error in carrying out the survey, not all parents were asked to rate the McKay school. To substitute for this comparison, we calculated an average satisfaction level for each type of school by taking the average percentage of parents who were either “very satisfied” or “satisfied” on each of the satisfaction measurements for the public and McKay schools.

Results

Parental Satisfaction

Results for the parental satisfaction measurements are presented in Tables 1-19 and Figures 1-9 (all tables and figures are contained in Appendix A). Parents were asked whether they were very satisfied, satisfied, dissatisfied, or very dissatisfied with each school on a set of measurements: individual attention, quality of services addressing the child’s disability, the child’s academic progress, class size, school facilities and equipment, teachers, communication with school staff, and school responsiveness to their needs. They were also asked for their level of satisfaction with the process for developing their child’s IEP at the public school. The tables and figures report the percentage of parents who were “very satisfied” or “satisfied” with each aspect of their schools.

As the figures show, for all of the satisfaction measurements except one, both current and former participants report being more satisfied with McKay schools than with public schools. In the case of school facilities and equipment, current participants are more satisfied with McKay schools than with public schools, while former participants are satisfied with both schools at about the same level (the difference is not statistically significant). In all other cases, McKay schools provided significantly more satisfactory results than public schools for both current and former participants.

The tables show results broken down into categories. Among the McKay schools, in virtually all cases there were no significant differences in satisfaction levels for students with mild or major disabilities, families with incomes above or below $30,000, and white or nonwhite students. This holds true for both current and former participants. Among current participants, the only two exceptions were that whites were moderately more satisfied than nonwhites with services addressing the child’s disability and with teacher quality. But in both cases nonwhites still reported very high satisfaction rates: 85.3% and 88.4%, respectively, as opposed to 29.5% and 48.2% in public school. Among former participants, the only significant difference was that parents of students with mild disabilities were more satisfied with class size. It is worth noting, however, that the sample of former participants is smaller than the sample of current participants, so it is more difficult for results to achieve statistical significance in this sample.

Class Size

Class size was a major concern for the parents in this survey. When asked to give the most important reason they decided to participate in the McKay program, 19.7% of current participants and 20.0% of former participants named class sizes, more than any other single reason, and a further 15.2% of current participants and 15.8% of former participants said they wanted more individual attention for their children, which amounts to much the same thing. By comparison, less than 1% of all respondents said they joined primarily to get better moral, civic, or religious instruction. (The question was open-ended; parents were not given a choice of possible answers but were simply asked to name the most important reason they participated.) So in one form or another, class size was by far the most frequently cited reason for choosing to take a McKay voucher.

Results for class sizes are presented in Tables 20-23 and Figures 10-11. As Figures 10 and 11 make clear, both current and former participants got what they said they wanted in terms of class sizes. While 63.2% of current participants and 47.4% of former participants reported 25 or more students in the public-school class in which their children spent the most time, only 6.0% in each group reported the same for their McKay-school classes. And while current and former participants averaged 25.1 and 21.8 students respectively in their primary classes in public school, they averaged 12.8 and 12.7 students in their McKay schools. As noted below (see p. 9-10), McKay schools receive approximately the same amount of money as public schools, yet the McKay schools provide special education students with classes roughly half as large. At McKay schools, there were no significant differences in average class sizes by race or income; current participants do show a significant difference by severity of disability, which is expected given that students with more severe disabilities are much more likely to need smaller classes.

Because many parents are concerned with the grouping of disabled students, our survey also asked how many disabled students there were in the student’s primary class. This gave us a measurement of the extent to which participating students were assigned to heavily-disabled classrooms or were “mainstreamed” into classrooms with few disabled students. Among current participants there was a significant difference between public-school and McKay-school classes: their public-school classes averaged 50.0% disabled students, while their McKay-school classes averaged 66.0% disabled students. By contrast, among former participants there was no statistically significant difference: their classes averaged 67.8% disabled students in public school and 70.8% disabled students in the McKay school. It appears that the focus on mainstreaming emphasized by disability advocacy groups is not shared by these parents, as they are chosing classes with higher concentrations of disabled students. They may give less weight to the importance of mainstreaming than advocacy groups do.

Services Provided

Public schools are required by IDEA, the federal law governing special education, to provide at least a certain minimum level of service to disabled students. Our survey asked parents whether their public schools provided all the services they were required to provide. Private schools are not required to provide any defined minimum level of service, so we asked parents whether their McKay schools provided all the services they had promised to provide. The difference arises from the different models of accountability under which the two systems operate: public schools are held to a process-compliance model in which a school is only required to fulfill a defined set of specific obligations, while the McKay program follows a parental-choice model in which schools are held accountable to parents who have the ability to withdraw their students from a school. The important thing for the public school model is that schools live up to their defined legal requirements; the important thing for private schools is that they deliver on the promises they make to parents.

Critics of the McKay program often argue that formal legal requirements like those that operate in public schools are the only effective means to ensure services are provided to special education students. A report on the program co-authored by People For the American Way and the Disability Rights and Education Defense Fund points out that “once parents use a voucher to transfer their children to private schools, they have effectively opted out of the legal rights and educational services guaranteed under IDEA.” The report concludes that the program “is not accountable to parents” (PFAW/DREDF 2003). The implied premise is that only a legal process-compliance model can hold schools accountable; by contrast, the premise of the McKay program is that the ability to control which school a child attends gives parents a more powerful way to ensure that schools provide adequate services.

Results for questions regarding accountability for services provided are presented in Tables 24-25 and Figure 12. Figure 12 shows that McKay schools did better than public schools on these services-provided measurements for both current and former participants; 30.2% of current participants and 36.3% of former participants feel their public schools provided all the services they were required to provide, while 86.0% of current participants and 49.3% of former participants feel their McKay schools provided all the services they promised to provide. Table 25 shows that there are no significant differences in McKay schools’ provision of promised services among participants in different race, income, or severity of disability groups.

Whenever parents reported a shortfall in services provided, our survey also asked how serious the shortfall was. In all cases, in both public and McKay schools, parents overwhelmingly described the shortfall as “somewhat serious” or “very serious” rather than “not too serious” or “not at all serious.” At McKay schools, there were no significant differences by race, income, or severity of disability among either current or former participants.

These results suggest that for many disabled students, IDEA’s formal legal guarantee of a particular level of services in public schools is not effectively ensuring that disabled students are actually provided the appropriate level of services. By contrast, current participants overwhelmingly report that their McKay schools are delivering the services they promise to deliver, and among former participants McKay schools perform significantly better on the services-provided measure than public schools. The ability to withdraw their students from public schools and place them in private schools has effectively empowered parents to ensure a level of services for their children that IDEA has often failed to ensure.

Individual Education Plans

As part of the transition from the public-school model of accountability through formal legal processes to the McKay model of accountability through parental choice, parents have the option of choosing a school that does not use an IEP process. Of course, just because a school does not have IEPs doesn’t mean that it has no method at all for determining what services are right for each child, or that the school doesn’t allow parents to participate in guiding their children’s educations. But whether a McKay school provides IEPs or not, the underlying system of accountability has changed. Rather than the threat of going to court—where schools have an advantage due to the government’s vastly superior resources—McKay schools are motivated by parents’ power to take their children to another school.

Critics of the program argue that only IEPs backed by the threat of lawsuits can motivate schools. The PFAW/DREDF report points out that “under McKay, private schools are not bound by IDEA or by the terms of a child’s IEP” and concludes that the program “sacrifices critical rights of parents and children” (PFAW/DREDF 2003). When the U.S. House considered encouraging states to adopt programs similar to McKay, Rep. George Miller (D-Ca.) argued that “this would create a wide-open voucher program with no accountability” (Goldstein 2003).

Results for IEP-related questions are presented in Tables 26-28. Parents were asked whether they came into conflict with their public schools over their students’ IEPs. Table 26 shows that 53.8% of current participants and 42.3% of former participants had such a conflict. Parents were then asked whether they had an IEP at their McKay school. Among current participants, 48.8% of students have an IEP; among former participants, 36.7% of participants had an IEP when they were in their McKay schools. There were no significant differences in whether students had IEPs in their McKay school by race, income, or severity of disability.

Those who had IEPs at their McKay schools were asked whether they had ever come into conflict with the school over it. Table 27 shows that 9.6% of current participants and 8.9% of former participants came into conflict with their McKay schools over their IEPs. Those who did not have IEPs at their McKay schools were asked whether this had ever been a problem. Table 28 shows that 19.9% of current participants and 42.6% of former participants reported that not having an IEP at their McKay school had been a problem. In both tables there are no significant differences by race, income, or severity of disability.

Those who reported any conflict over their IEPs, or who reported that not having an IEP was a problem, were asked how serious the conflict or problem was. As in the Services Provided section, parents overwhelmingly characterized their conflicts or problems as serious. At McKay schools, there were no significant differences by race, income, or severity of disability among current or former participants.

These results suggest that the IEP process works better when backed up by parents who can withdraw their children and place them in another school rather than by the threat of a lawsuit. Both current and former participants reported much lower rates of conflict with their schools over their IEPs in McKay schools than in public schools. Where parents chose not to have IEPs, current participants were much less likely to report that this was a problem than that they had come into conflict with their public schools over their IEPs. Former participants who chose not to have IEPs in their McKay schools found that this was a problem at almost exactly the same rate as they came into conflict with their public schools over their IEPs.

Behavior Problems

Schools serving disabled students face important non-academic challenges, and one of the biggest is helping them overcome behavior problems. Many disabled students have difficulty in learning the limits of acceptable behavior. One service a school provides to disabled students is educating them in a way that helps teach those limits. We asked parents whether their children had behavior problems at school in order to find out whether students in McKay schools are more likely, less likely, or equally likely to suffer from behavior problems than they did when they were in public school.

The results for behavior problems at school are presented in Tables 29-30 and Figure 13. Current participants reported that their children had significantly fewer behavior problems at school in their McKay schools than in public schools; 40.3% reported behavior problems in public school, compared to 18.8% in their McKay schools. For former participants, the rate of behavior problems in McKay schools as compared to public schools was also lower, but the difference was not statistically significant; 48.4% of former participants had behavior problems in public school, compared to 37.3% in McKay schools. As noted above, it is more difficult for former-participant results to achieve statistical significance because of the smaller sample.

In McKay schools there were no significant differences in rates of behavior problems by race or income. However, students with major disabilities were more likely than students with mild disabilities to have behavior problems. This is to be expected, since students with more severe disabilities will be more likely to have trouble overcoming behavior problems.

Parents who reported any behavior problems were asked how serious they were. In public school, 65.7% of current participants and 65.4% of former participants reported that the problems were “somewhat serious” or “very serious” rather than “not too serious” or “not at all serious.” In McKay schools, 32.7% of current participants and 59.3% of former participants reported that their problems were “somewhat serious” or “very serious,” and there were no significant differences by race, income, or severity of disability.

Problems with Other Students

Another non-academic outcome that is especially important for disabled students is the extent to which they are victimized by other students. Just as the management of disabled students’ behavior is part of the services a school provides, another part is the extent to which the school is able to manage the behavior of other students toward disabled students. Obviously no school can exercise complete control in this area, but such things as student grouping policies and discipline policies can have a significant impact. We asked parents how often their children were bothered by other students because of their disabilities at public and McKay schools, and whether their children had ever been physically assaulted at either school because of their disabilities.

The results for problems with other students are presented in Tables 31-34 and Figures 14-15. As the figures show, in both cases parents reported significantly fewer problems with other students at McKay schools than at public schools. Among current participants, 46.8% were bothered “very often” or “often” because of their disabilities at public school, compared to 5.3% at McKay schools; 24.7% were ever physically assaulted because of their disabilities at public school, compared to 6.0% at McKay schools. Among former participants, 41.4% were bothered “very often” or “often” because of their disabilities at public school, compared to 20.0% at McKay schools; 26.0% were ever physically assaulted because of their disabilities at public school, compared to 10.2% at McKay schools.

Among both current and former participants, there were no statistically significant differences by race or income for being bothered often by other students at McKay schools. For physical assaults at McKay schools, current participants had no significant differences by income and former participants had no significant differences by income or race, but among current participant nonwhites were more likely to be assaulted than whites. But nonwhite current participants still reported low rates of physical assault in McKay schools; only 8.9% of them were assaulted in McKay schools, compared to 27.2% in public schools.

Students with major disabilities were more likely to be bothered often or assaulted at McKay schools than students with mild disabilities, which is to be expected since they are more vulnerable. But these most-vulnerable students were still much less likely to be bothered or assaulted at McKay schools than at public schools: 50.0% of current participants with major disabilities were bothered often and 26.5% were assaulted at public schools, compared to 7.7% bothered often and 9.2% assaulted at McKay schools; 43.1% of former participants were bothered often and 31.0% assaulted at public schools, compared to 25.0% bothered often and 11.2% assaulted at McKay schools.

It is not difficult to understand the importance of this type of outcome for parents of special education students. Disabled children are more likely to be the targets of, and less likely to be able to defend themselves against, bullying and teasing by other children. One likely explanation for McKay schools’ superior performance here is that private schools are much more free to set their own discipline policies than public schools, and thus are able to take a stronger hand in protecting the most vulnerable children from victimization. Also, the significantly smaller classes at McKay schools probably make it easier to effectively implement those discipline policies.

Tuition and Fees Above the Scholarship

The McKay scholarship provides a voucher equal in value to the amount that each student’s public school would have spent on that student. Participating families are permitted to use this voucher at schools that charge more in tuition and fees than the scholarship will fully cover. These participants must pay the additional tuition and fees themselves. Many critics claim that only wealthy families will be able to afford good special education services in private schools. One New York Times news article asserted, without supporting evidence, that “in general, private special education services cost slightly more than twice as much as the same services in a public school setting” (Schemo 2002). The PFAW/DREDF report asserts that because parents are free to pay more than the amount of the scholarship, “McKay tends to subsidize middle- and upper-income parents, who can better afford to pay additional amounts to cover private school tuition and other fees” (PFAW/DREDF 2003). In his New York Times column, Richard Rothstein asserted that only schools with “no special services” for the disabled would take the scholarship as full payment, and “only better-off families can afford schools offering special education” (Rothstein 2002).

Results for participants paying tuition and fees above the amount covered by the McKay scholarship are presented in Table 35 and Figure 16. We found that majorities of both current and former participants pay nothing at all above the scholarship—53.8% of current participants and 58.1% of former participants. Among current participants, 71.7% are paying nothing or less than $1,000 above the scholarship, while 27.3% are paying $1,000 or more above the scholarship. Among former participants, 75.8% were paying nothing or less than $1,000 above the scholarship, while 21.9% were paying $1,000 or more above the scholarship.

One surprising result of our survey was that among current participants, students with major disabilities are significantly more likely than students with mild disabilities to pay nothing above the scholarship, and also less likely to pay $1,000 or more above the scholarship if they paid anything. This is probably because students with major disabilities receive a larger voucher, as public schools would have spent more on educating them. Less surprisingly, current participants were more likely to pay anything above the scholarship, and to pay $1,000 or more above the scholarship, if they were in the white or high-income groups. Former participants produced no statistically significant differences by race, income, or severity of disability.

These results show that the superior services parents are getting from McKay schools are mostly being obtained either at no cost to parents or at a relatively modest cost. Concerns that only those who pay large sums of money above the scholarship will receive adequate services appear to be unfounded. Contrary to the assertions of the Times’ news article and Rothstein’s column, most McKay participants are getting measurably better services from private schools for either the same amount of money that public schools spend or for only a moderately higher amount.

Demographics

Like many voucher programs, the McKay program suffers from the perception that it is simply a subsidy for wealthy white students who would be in private schools anyway. For example, Rothstein asserts that the program “has permitted vouchers to become only a subsidy for the relatively affluent” (Rothstein 2002). Demographic information on the participants in our survey is presented in Tables 36-43. As the tables show, McKay participants actually reflect a diverse cross-section of families in terms of race, income, parent’s education, and family status.

The PFAW/ DREDF report asserts that “many private schools that claim to welcome students with disabilities frequently pick and choose, denying admission to students with more severe or specific kinds of disabilities” (PFAW/DREDF 2003). But Table 36 shows that in terms of severity of disabilities, McKay students reflect the general population of disabled students in Florida. Although the state does not keep exact statewide numbers on matrix levels, the Florida Department of Education estimates that 60% of students are in Matrix Level 1 (the mildest category), 25% are in Matrix Level 2, 10% are in Matrix Level 3, and 5-6% are in Matrix Levels 4 and 5 (see Florida DOE 2002). No doubt many schools do “pick and choose,” but that’s a two-way street; some schools can’t accommodate severe disabilities and take only mild cases, but other schools specialize in students with severe disabilities. In any case, it appears from our survey that students with more severe disabilities are not underrepresented in the McKay program.

Difficulty Finding an Acceptable School

Closely related to concerns about demographics are concerns that participants might not be able to find an acceptable school. The PFAW/ DREDF report worries that “choices under McKay depend largely on where a family happens to reside,” because there may not be enough participating private schools in rural areas (PFAW/DREDF 2003). Rothstein also expresses concerns about the difficulty parents may have in finding private schools that provide special education services (see Rothstein 2002).

Results for difficulty finding an acceptable school are presented in Table 44. Among current participants, 27.5% reported that finding an acceptable school for their children was “very difficult” or “somewhat difficult,” as opposed to “not too difficult” or “not at all difficult.” There were no significant differences by race, income, or severity of disability. Among former participants, 47.9% reported that finding an acceptable school was “very difficult” or “somewhat difficult.” There were no significant differences by race or income, but participants with major disabilities were somewhat more likely than participants with mild disabilities to report that finding an acceptable school was difficult.

These results indicate that most participating families did not have significant difficulty in finding an acceptable school. It is true that former participants reported a higher rate of difficulty than current participants, but recall that former participants make up only 10.2% of all families that participated in 2001-02 and 2002-03. If we weight the current and former participant categories to represent their respective shares of the total participant population, we find that 29.8% of all participants from those two school years had difficulty finding an acceptable school. This means that seven out of ten participants found it “not too difficult” or “not at all difficult” to find an acceptable school. However, some parents did have difficulty finding an acceptable school, and this seems to be the main reason parents leave the program. When former participants were asked the main reason they left the program, the most common response, making up 16.3% of former participants, was that they had not found a private school with adequate services.

It is also important to look at the overall McKay participant population when considering the difficulty that students with major disabilities had in finding an appropriate school. When our survey samples are weighted to represent their shares of the total participant population, among all McKay participants in 2001-02 and 2002-03 we find that 29.3% of students with mild disabilities and 30.5% of students with major disabilities had difficulty finding an acceptable school. There is no statistically significant difference between these figures, indicating that in the overall participant population students with more severe disabilities were no more likely that students with mild disabilities to have difficulty in finding an acceptable school.

Transportation

Results for transportation are presented in Tables 45-46 and Figure 17. Parents were less likely to find transportation to public school burdensome than transportation to McKay schools. Among current participants, 18.5% found transportation to public school burdensome, and 34.5% found transportation to McKay schools burdensome. Among former participants, 16.7% found transportation to public schools burdensome, and 58.1% found transportation to McKay schools burdensome. There were no significant differences in reported transportation problems for families that chose McKay schools when those families are broken down by race, income, or severity of disability. Those who reported that transportation to either school was burdensome overwhelmingly reported that it was “very burdensome” or “somewhat burdensome” rather than “not too burdensome” or “not at all burdensome.”

As in the previous section, when interpreting these results it is important to remember that former participants represent only 10.2% of the total participating population for 2001-02 and 2002-03. If we weight our current and former participant samples to represent their respective shares of the total population, we find that 37.1% of all participants found transportation to the McKay school burdensome. This means that almost two thirds of participants did not find it burdensome. For those that did, this was sometimes a reason for leaving the program; 9.3% of former participants say the main reason they left the program was transportation to the McKay school.

Whether the McKay Program Should Continue

Ultimately, the most important question facing policymakers concerning the McKay program is whether students are better off with the program in existence than without it. Naturally, we would expect current participants to want the program to continue, since they have already expressed their support for it by their decision to stay in the program. However, it would be interesting to learn whether former participants think the program is something students in general would be better off without, or just something that for some reason wasn’t right for them.

Results for whether the McKay program should continue to be available or should be discontinued are presented in Table 47. Of former participants, 90.7% responded that the program should continue to be available. (Current participants were not asked this question.) There were no significant differences by race, income, or severity of disability. This must be considered a very strong show of support for the McKay program, considering that these are the parents who have chosen not to participate any longer. It appears that whatever their reasons for leaving, they don’t feel that other students would be worse off for having the program available.

Conclusion

The results of the survey indicate that, contrary to unsubstantiated claims made by the program’s critics, participants in the McKay program are being well served. Virtually all measurements showed higher levels of satisfaction, better provision of services, and better student environments at McKay schools than at public schools, and in almost all cases members of different racial, income, and disability groups showed no significant differences in their experiences. This is true even for families that have dropped out of the program, who also overwhelmingly endorse the program’s continued availability for other students. It is rare that any education reform program, much less one targeted at a population that is especially difficult to serve, shows results that are as broadly and strongly positive as those found in this survey for the McKay program.

 


Center for Civic Innovation.

EMAIL THIS | PRINTER FRIENDLY

CR 38 PDF (334 kb)

WHAT THE PRESS SAID

SUMMARY:
This new study, by Manhattan Institute Senior Fellow Jay P. Greene, Ph.D. and Greg Forster, Ph.D., is the first empirical examination of the Florida McKay Scholarship Program for Students with Disabilities, the nation’s second largest voucher program and the first and only program to offer school vouchers to all special education students in public schools. The study, through telephone surveys of parents of both current and former voucher recipients, shows that participants in the McKay program are being significantly better served by McKay schools compared to the child’s previous public school—regardless of the race, income, or disability level of the voucher recipient.

TABLE OF CONTENTS:

Executive Summary

About the Authors

Acknowledgments

Introduction

Background

Previous Research

Method

Results

Parental Satisfaction

Class Size

Services Provided

Individual Education Plans

Behavior Problems

Problems with Other Students

Tuition and Fees Above the Scholarship

Demographics

Difficulty Finding an Acceptable School

Transportation

Whether the McKay Program Should Continue

Conclusion

References

Appendix A: Tables and Figures

Appendix B

Appendix C: Survey Methodology

 


Home | About MI | Scholars | Publications | Books | Links | Contact MI
City Journal | CAU | CCI | CEPE | CLP | CMP | CRD | ECNY
Thank you for visiting us.
To receive a General Information Packet, please email support@manhattan-institute.org
and include your name and address in your e-mail message.
Copyright © 2009 Manhattan Institute for Policy Research, Inc. All rights reserved.
52 Vanderbilt Avenue, New York, N.Y. 10017
phone (212) 599-7000 / fax (212) 599-3494