In his State of the Union message on January 27, 1998, President Bill Clinton embraced the idea of reducing class size to 18 for students in grades one through three in the nation’s public schools.
While the President’s proposal has strong political appeal, it represents one of the most expensive education reforms possible, and its benefits are far from certain. The President’s timing may be off as well: many taxpayers are already chafing at the high cost of public education, and other, less expensive and more cost-effective, reforms are available.
Still, it was Republican James S. Gilmore’s promise to reduce class sizes that helped him become governor of Virginia last November. In California, Pete Wilson’s efforts to reduce class size have won acclaim from parents. Last summer, 18 state legislatures were considering or had already passed bills to reduce class sizes, according to Stanford University education professor Michael Kirst.
Assessing the Costs and Benefits
Although the President’s plan calls for spending $12 billion over seven years to hire 100,000 new teachers, the federal money would pay only part of the new teachers’ salaries for a limited time, leaving states and local school districts to pick up the final bill. In California, it cost an extra $800 per pupil to reduce class sizes. U.S. News and World Report estimates that to reduce the average class size by 10 students nationally would cost more than $85 billion in teachers’ salaries and benefits alone.
“Cutting classes is expensive,” said Kirst in a New York Times article last August, citing the cost of portable classrooms at $50,000 each, plus an already tight labor market for teachers. In pursuing lower class sizes, he said, politicians “are committing their state to one of the most expensive educational reforms possible and neglecting lower cost, effective alternatives.”
Although parents like the idea of smaller classes, and the National Education Association has been seeking an “optimum” class size of 15 since 1986, research support for the benefits of smaller classes is limited.
In the study most frequently cited in favor of lower class sizes, Tennessee first, second, and third graders taught in classes of 13 to 17 achieved higher reading and math scores than those taught in classes of 22 to 25.
The results from the Tennessee experiment simply confirm the findings of a research study performed 20 years ago by Gene Glass and Mary Lee Smith, who concluded from a “meta-analysis” of 14 other research studies that the relationship between class size and student achievement was best described as an “effects curve” with different effects at different class size ranges. The characteristics of the “effects curve are as follows:
- In the range of 20 to 40 students, student achievement is largely insensitive to changes in class size; if 40 students are taught together, they learn only 5 percent less than would a group of 20;
- In the range of 15 students and below, student achievement is dependent on class size; as class size is reduced below 15, student achievement levels increase, though not by a large amount.
“When we looked at the Tennessee data, it fell right on the curve,” said Glass, now Associate Dean for Research in the College of Education at Arizona State University at Tempe.
Glass and Smith concluded that not until class sizes are lowered to 15 students or fewer will class size reduction alone reliably improve student performance. And only when class sizes fall to around 10 students or fewer do achievement gains become appreciable. The President’s proposal to reduce class size to 18 could be expected to have only a marginal effect on achievement, at a very high cost.
Do Lower Class Sizes Really Matter?
“The issue is not a student achievement issue, it’s a teacher workload issue,” said Glass about the proposed reduction in class size to 18. “The achievement gain for students wouldn’t be anything significant, but the workload reduction for teachers would be appreciable.”
Critics of the President’s proposal also note that class sizes in public schools have been falling for decades with no apparent gain in student achievement. By contrast, large classes in parochial schools appear to be no deterrent to academic success.
“Urban parochial schools have proved that smaller classes may not be necessary,” noted Kirst. “They tend to have large classes, but their students have higher achievement scores than their counterparts at many public schools.”
Other ways of improving student achievement have been suggested as alternatives to the President’s proposal. Improving the instructional competence of teachers, for example, would lighten their workload by helping them perform more effectively. Students would benefit from expanding summer school programs; hiring aides and special instructors to assist with weak students; and adding tutoring programs. All would be less costly and hold more promise for success than the President’s plan.
George A. Clowes is managing editor of School Reform News. His email address is [email protected].
George Clowes is former Heartland senior fellow addressing education policy. He served as founding managing editor of School Reform News between November 1996 and January 2005.
