The sociologist James S. Coleman once pointed out that policymakers often use research on education to legitimize, not to guide, their policies. Or, as one author wrote about an ideological opponent, “He uses statistics as a drunken man uses lampposts–for support rather than illumination.”

While policymakers aren’t the only ones who conveniently ignore uncomfortable facts and statistics, it is dangerous when they craft public policy on the basis of cherry-picked research. Their uninformed mandates can force conformity on millions of people in one fell swoop. Reviewing the history of education reveals that policies that are adapted during a crisis, real or imagined, sometimes remain with us for decades, if not centuries.

President Clinton has been leaning on “recent research” to support his plan to reduce class sizes. The truth is that the research Mr. Clinton hasn’t cited should make serious researchers suspicious that he merely sought out research to match his conclusions. University of Rochester economist Eric Hanushek examined 277 separate published studies on the effect of teacher-pupil ratios and class-size averages on student achievement. Only 15 percent suggested that there is a “statistically significant” improvement in achievement, 72 percent found no effect at all, and 13 percent found that reducing class size had a negative effect on achievement.

If the president doesn’t trust that research, he can consult the U.S. Department of Education. Although American students lag behind other students in international testing, American classrooms have an average class size of 23 students, incredibly few compared with the averages of 49 in South Korea, 44 in Taiwan, and 36 in Japan. Washington has an average class size below the national average, yet ranks near the bottom in academic achievement. We shouldn’t forget that average class size in American schools dropped from 30 in 1961 to 23 in 1998, without any improvement in standardized-test scores.

Just as bad money drives out good, the president’s plan to hand out money for reduced class sizes will drive out alternative approaches. Far too many teachers are ignorant of the subjects they’re teaching and are an educational liability, no matter how small their classes. As the head of one private school recently said, “We believe that a poor teacher can’t even teach five students–and a good teacher can teach a hundred.” About one-third of public school teachers lack majors or minors in the subjects they teach. The more advanced the subject, the greater the percentage of unqualified teachers. Will inner-city schools, where half the teachers lack a major or minor in their subjects, be helped by more of the same kinds of teachers produced by education schools?

While educators have long pooh-poohed the idea that one must be a master of one’s subject in order to teach it, the Thomas B. Fordham Foundation study, “Better Teachers, Better Schools,” found that students whose teachers have college degrees in math or have been specifically certified to teach math score significantly higher on standardized tests than students whose teachers did not specialize.

Mr. Clinton, of course, says he is focusing on teacher quality, too, but the Fordham study points out that approaches focusing on inputs (courses taken, requirements met, time spent, activities engaged in) rather than outputs (student achievement) will be counterproductive. Because not even the leader of the free world can mandate knowledge, he is seeking to decree the appearance of competence by requiring teaching certificates.

Pronouncements coming from the political leadership of federal departments are notorious for serving the interests of the party in the White House. The Education Department itself has flip-flopped on the issue since Mr. Clinton decided he wanted class sizes reduced. In 1988, the department concluded that reducing class sizes would be expensive and probably “a waste of money and effort.”

Mr. Clinton’s use of the department to promote research favorable to his proposals shows why the federal government shouldn’t be involved in producing education, especially given the government’s large role in producing research. This newspaper recently noted that federal money pays for 60 percent to 75 percent of all research on education conducted nationwide. (“Researchers Call for OERI Reforms at Hearing ‚” June 23, 1999.) How likely is it that federal research would show that Mr. Clinton’s class-size initiative is a waste of money and effort, especially with the Education Department’s office of educational research and improvement coming up for reauthorization soon?

The political battle over reducing class size is sure to continue; how great a role research will actually play is a different question, and one that should worry researchers. There is a lot of solid research warning us that reducing class sizes will just be education’s latest fad embraced by policymakers.