School vouchers: some testing, please

Commonweal, Jan 15, 1999 by Edward McGlenn Gaffney, Jr.

In a 1932 case that allowed the state of Oklahoma to control the manufacturing of ice, Supreme Court Justice Louis Brandeis wrote in dissent: "The business of supplying ice is not only a necessity, like that of supplying food or clothing or shelter, but the legislature could also consider that it is one which lends itself peculiarly to monopoly." If the manufacturing of ice "lends itself peculiarly to monopoly," the monopolization of the transmission of ideas and values in education is all the more peculiar. In Pierce v. Society of Sisters (1925), the Court had ruled that the states could not literally monopolize education in the sense of prohibiting all competitors to its own schools, as Oregon - egged on by the Ku Klux Klan in requiring attendance at public schools - had tried to do. But in a series of cases since 1947 the Court has, with a few notable exceptions, ruled that public financial support for religious schools is prohibited by the First Amendment's religion clause. The net effect of these decisions has been a virtual monopoly of tax dollars for the schools that the government operates.

Commentators of all stripes agree that the Court's jurisprudence on this matter has been unsatisfactory. University of California law professor Jesse Choper describes the Court's decisions as "ad hoc judgments which are incapable of being reconciled on any principled basis." For example, New York may lend secular textbooks to children attending religious schools, but Pennsylvania and Ohio may not give maps to a religious school. Senator Daniel Patrick Moynihan (D-N.Y.) pondered what the Court would do with an atlas, a book of maps. At one point the Court almost gave up on its duty to instruct clearly on constitutional values, stating that their teaching "sacrifices clarity and predictability for flexibility." Rigid, absolute lines may not be desirable when the practices of thousands of school districts are at issue, but doctrinal chaos is not a good alternative. The most hopeful sign of clarity emerging from chaos on these matters was the articulation of a clear principle in the 1997 Agostini case, in which the Court allowed public school teachers to provide remedial education for poor children on the premises of religious schools. Justice Sandra Day O'Connor wrote that there is no impermissible financial incentive to advance religion when "aid is allocated on the basis of neutral, secular criteria that neither favor nor disfavor religion, and is made available to both religious and secular beneficiaries on a nondiscriminatory basis."

An editorial in these pages noted that recent rulings of the Court "suggest that benefits permitted to secular private enterprises should not be denied to similar religious enterprises" ("Good Exercise," July 17, 1998). The editorial was commenting on the Milwaukee Parental Choice Program. The purpose of this program was to give low-income parents an opportunity to have their children educated outside that city's embattled public schools. Poor people are empowered by this program to do exactly what people of means would do: get better value elsewhere. Last June the Wisconsin Supreme Court sustained the program against constitutional challenges. Although both parties urged the Supreme Court of the United States to review the case, the Court declined to do so last November. The debate over the constitutionality and policy wisdom of vouchers will assuredly go on apace. But this debate will now have to be informed more by facts than by predilections fueled by passions about the separation of church and state.

Why do facts matter in the constitutional calculus? Because one of the Court's three tests governing the permissibility of funding for religious education asks whether the primary and principal effect of a program is to advance religion. One can speculate - as lawyers and judges do when they "think like lawyers" - about the purposes of programs, but to inquire about their effects is to ask an empirical question that deserves an empirical answer. Regrettably, the litigation process has studiously avoided serious social science data about religious schools accumulated by researchers of national stature, such as the pioneers James Coleman and Andrew Greeley. Never will you find a reference in such legal decisions to Anthony Bryk, Peter Holland, and Valerie Lee, who demonstrated in Catholic Schools and the Common Good (1993) that Catholic schools are particularly effective (if I may use that nonlegal word) with low-income African-American and Latino students. Litigation on school choice has been not only unscientific in the sense that lawyers and judges routinely ignore relevant data. It has even been anti-scientific in the sense that the outcome in most cases has been an injunction prohibiting any state or local community from daring to attempt any experiment that might provide data relevant to the empirical "test" under which the cases were purportedly being decided.

In the Oklahoma ice case mentioned above, Justice Brandeis concluded his dissent: "To stay experimentation in things social and economic is a grave responsibility. Denial of the right to experiment may be fraught with serious consequences to the nation. It is one of the happy incidents of the federal system that a single courageous state may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country. This Court has the power to prevent an experiment.... But, in the exercise of this high power, we must be ever on our guard, lest we erect our prejudices into legal principles. If we would guide by the light of reason, we must let our minds be bold."