Congress is again debating the proper role of the federal government in educating the nation's children, but on one area liberals and conservatives agree: The federal government should promote quality research on the effectiveness of various education reforms. The No Child Left Behind Act of 2001 (H.R. 1) would take a step in that direction. To evaluate how parental choice programs affect student achievement, a provision has been added to the bill to fund large-scale school-choice experiments around the country. This is sound policymaking. Too often in the past, Washington simply has continued to fund education programs without regard for results. As calls for parental choice grow, solid research is needed to enable lawmakers to target resources to the most effective programs and policies so that, as President Bush explains, "no child is left behind."

Evaluating the Effects of Choice
Experts from differing ideological backgrounds have called on the federal government to study school choice to see if it boosts achievement in children, particularly low-income students. A National Research Council (NRC) study commissioned by the Clinton Administration recommended in 1999 the establishment of a "large and ambitious" research experiment to determine if school choice programs improve student performance. The NRC's independent panel of 18 education experts included Dr. Helen Ladd, Professor of Public Policy Studies at Duke University and critic of school choice. Ladd noted that the scholars looked carefully at the existing research but concluded that they did not know very much. While school choice was an important public policy issue, a rich body of empirical data on the subject was lacking.

The NRC report, Making Money Matter: Financing America's Schools, points out that while "housing, welfare, and medical policies have frequently been the subject of this kind of experimental research," school choice has been immune from the same rigorous test of effectiveness. The NRC recommends that experimental projects continue for up to 10 years in order to determine if school choice "can bring about broad-based improvement in educational outcomes, especially for children in concentrated areas of disadvantage." To be scientifically valid, scholarships must be offered to a "significant fraction of students in several different school districts."

The House Education and Workforce Committee responded to the growing calls by including a
provision in H.R. 1, the reauthorization of the Elementary and Secondary Education Act, to spend $25 million each year through 2006 to research the effects of school choice. Under this plan, state and local education agencies and private non-profit groups seeking to establish pilot choice programs would submit an application to the Secretary of Education. To qualify, the projects would need to be of sufficient size to permit researchers to measure their impact on public and private schools in the surrounding area.

Only students who receive free or subsidized meals under the federal School Lunch Program would be eligible to participate in these projects, and 90 percent of the federal funds for the program would need to reach students in the form of direct grants. Participating schools must uphold civil rights laws and have non-discriminatory enrollment procedures. An independent third-party evaluator would measure student performance in the 3rd through 8th grades. Finally, recipient schools would report annually on academic achievement and parental satisfaction. Thus, the pilot programs would provide the kind of research data that the NRC sought.

Lingering Questions
Though recent analyses suggest that school choice improves student outcomes, many questions remain unanswered. A prominent study of privately funded scholarships conducted by the Harvard University Program on Education Policy and Governance found that
average test scores among African-American choice students improved by a statistically significant 6 percent relative to their public school peers. To avoid selection bias, the Harvard study used two random samples of students--children receiving scholarships and attending private schools, and public school students who unsuccessfully applied for scholarships. Although revealing, the study left unanswered a number of important research questions regarding grade-level variances, frequency of testing, and which students benefit most from choice.

Professor Patrick Wolfe of Georgetown University, a lead researcher on the Harvard study who is sympathetic to school choice, notes that while many initial findings were positive, the inconsistent research on which students improve when expanded choice is introduced was puzzling. Different studies also suggest that academic improvement varies according to ethnicity, but experts disagree on what causes these differences. Martin Carnoy, Professor of Education and Economics at Stanford University and a critic of the Harvard study, thinks the limited gains of African-American students in choice programs are rooted in "a more structured private school environment" and smaller classes. The Harvard research certainly suggests that expanded choice benefits certain students. But the pool of research is limited. Further research trials are needed to determine the impact of expanded parental choice in education.

Meeting Critics' Concerns
The provision in H.R. 1 is designed to satisfy critics' concerns about financing and participation. For example, Senator Joseph Lieberman (D-CT) has argued that choice programs should involve only low-income children who are arguably trapped in schools that are not educating them, and not take money out of the public school budget. The provision in H.R. 1 requires any local entity petitioning the Department of Education for funding to "focus on students and families that lack the financial resources to take advantage of available educational options." Participation would be voluntary, and funds for the demonstration programs and research would be set aside and independent of other programs to ensure that no money is diverted to these pilot projects from Title I or any other federal education program.

Conclusion
The House provision to fund school-choice pilot programs presents Congress with a constructive way to increase the amount of information available on the effects of school choice. Too often, the lack of accountability in how education tax dollars are spent makes children vulnerable to politically popular policy fads that fail to boost achievement. Education policy should be firmly grounded in scientific research, and the results should enable federal, state, and local officials to make informed decisions on proposals such as parental choice, while enabling the public to hold their elected representatives and schools accountable for results.

Thomas Dawson is a Fellow in Educational Affairs at The Heritage Foundation.