Contrary to popular perception, the public schools do not accept every child. Those students with behaviors too disruptive or disabilities too severe may be outplaced by the public schools to private-sector providers. Moreover, some families prefer that their special-needs child be served in a private or nonpublic school where their unique learning needs may be better addressed.

The fact that public schools do not, and cannot, accommodate every child is not a condemnation of public education. Public schools rightly recognize that they cannot provide quality education to every child. By enlisting the cooperation of private and nonpublic schools, the public-education system ensures that all children will be served, although the public schools themselves do not necessarily deliver the services.

Where the public-education system should be criticized is in its selective use of the private sector to serve only those children the public schools will not educate. America’s public-education system should invite the participation of both public and private schools in educating all children. School choice policies, which allow students and their families to voluntarily enroll in the public or private school of their choice, would expand the educational opportunities of all students.

Critics of school choice argue that public schools would become the "dumping ground" for the most difficult students under a system of school choice. Only good parents would take the opportunity to exercise choice, the argument runs, and private schools would shun difficult-to-educate students. Under a properly designed school-choice system, that scenario is unlikely.

Many private and nonpublic schools exist to serve students with all variety of disabilities, at-risk behaviors, or other difficulties. Students from all socio-economic and family backgrounds are served by these schools, ranging from those children from abusive families who may care little for their children to families who are highly involved in their children’s upbringing and education.

Noteworthy is the wide variety of private-sector schools specializing in serving difficult-to-educate students. Private and nonpublic schools for at-risk students, for example, run the gamut from independent study to residential centers, embracing secular or nonsecular curricula, and emphasizing college preparatory to vocational education. Some schools catering to students with disabilities integrate them with their nondisabled peers, and more private schools will be likely to do so if the demand for such integration grows. This suggests that a large-scale school-choice program would result in a supply of schools as diverse as the students they are intended to serve. That private schools serve such a variety of special-needs students, and sometimes enroll the most difficult among them, lays bare the myth of the public-school dumping ground and the notion that school choice will leave only the worst students in the public schools.

From the standpoint of policy design, school-choice comes in different forms, including tuition tax credits, privately funded tuition-vouchers, and government-financed tuition vouchers. Several implications for government-funded vouchers are raised by the findings in this policy study. Government-funded vouchers, though they would introduce market dynamics into public education, would not create a free market in education. Like nonpublic schools, voucher-redeeming private schools may be a hybrid of the public and private sector. Nonpublic schools tend to be more regulated than their purely private counterparts. Over time, voucher-redeeming private schools might develop similar characteristics.

In several important ways, however, a contract arrangement between a public agency and a private provider (the nonpublic-school model) is different from a voucher. Vouchers are given to parents and it is parents, not public agencies, who decide where they should be used. (In general, parents are passive choosers among nonpublic schools because placement agencies are usually made by a government agency or child-study team.) Competition may be more robust in a voucher system than among nonpublic schools because parents may be more vigilant about demanding quality services for their own child than a government agency would. If the voucher is designed so that parents pay a share of the tuition costs, voucher schools may also be more aggressive in controlling costs. (In the public sector, it is generally thought that local agencies will make a greater effort to minimize costs for, say special education, if they have a financial stake in the placement decision. Parents may behave similarly if they are required to pay even a small portion of private-school costs.)

The nonpublic-school arrangement may indicate some of the ways in which a voucher system might operate, but it also has shortcomings as a predictive model. Future research efforts may wish to explore how a school-choice system should be designed and financed to meet the needs of difficult-to-educate students, especially students with disabilities who may require vouchers with higher dollar amounts compared to the voucher amount for regular students.