School choice in Milwaukee - school-voucher plan

Public Interest, Fall, 1996 by Paul E. Peterson, Jay P. Greene, Chad Noyes

So, despite hardships, the Milwaukee experiment still provided evidence that school choice can lead to more effective schools. Parents were pleased, retention rates increased, test scores rose. As a result, Milwaukee's choice program has become an institutionalized component of the city's educational system. The program has even prospered to the point that the Wisconsin state legislature felt confident enough in the spring of 1995 to strip the program of many of its restrictions. Although the expanded version still must survive constitutional scrutiny by the Wisconsin judiciary, the very fact that choice continues to move toward the center of the state's educational agenda suggests that the Milwaukee experiment might still "wag the dog."

The politics of evaluation

But what about the Witte evaluation? Did it not find that the choice students learn no more than public-school students? Can parental satisfaction and anecdotal evidence substitute for the knowledge generated by high-powered academic research? Given the politically biased origins of the Witte evaluation, the answer is yes.

The legislature gave the responsibility for overseeing the evaluation to Superintendent Grover, who picked his evaluator on September 9, 1990, just days after the first choice students began going to school. A choice critic like Grover had good reason to move expeditiously. The early years are times when experimental programs are most fragile, most prone to error. Nothing can pose as great a danger to an innovative governmental program as an immediate, apparently rigorous evaluation of its effects. But Grover's eagerness to move quickly jeopardized his ability to obtain a high-quality evaluation. Reliable evaluations are complicated undertakings in which a number of major research organizations have specialized, including such well-known entities as Abt Associates, Mathematica, NORC, RAND, and the Manpower Development Research Corporation. If Grover had followed standard practice, he would have asked for bids from these research organizations to undertake the evaluation. Instead, he gave the assignment to John Witte, a political-science professor with minimal experience with large-scale evaluations.

Moreover, Witte was hardly disinterested. In fact, he had written a paper, well received in teacher-union circles, which criticized studies finding private schools outperforming public schools. Witte would later become a public opponent of the expansion of the Milwaukee choice program in 1995.

Witte quite frankly reports that choice-school parents were much happier with their new school than with their prior public school. Seventy-five percent of choice parents gave their child's school a grade of either "A" or "B," 10 percentage points higher than the grades given public schools. Choice parents expressed substantially greater satisfaction than public-school parents with every aspect of their child's education: the amount their child learned, the teacher's performance, the program of instruction, the discipline in the school, the opportunities for parental involvement, the textbooks, and the location of the school.