Class-Size Reduction Brings Mixed Results

Published January 1, 2004

Two recent studies of student achievement for students enrolled in class-size reduction programs in Wisconsin and California offer mixed results and call into question the cost effectiveness of large-scale programs with mandatory class-size caps.

Class-size mandates are a growing national trend. At least 20 states have some form of class-size reduction legislation, and smaller classes remain a popular school reform with parents and teachers. In 2002, Florida voters passed a statewide class-size reduction law through the initiative process, and advocacy groups and parents recently lost a battle for mandatory class-size reduction in New York City.

Wisconsin Results

In “Class Size Reduction in Wisconsin: A Fresh Look at the Data,” Arizona State University education researchers at the Education Policy Research Unit (EPRU) re-evaluate the effectiveness of Wisconsin’s Student Achievement Guarantee in Education (SAGE) program. SAGE is a statewide effort to increase the academic achievement of children living in poverty with an intervention plan that includes reducing the student-teacher ratio in grades K-3 to 15:1. SAGE also involves a more rigorous curriculum, before- and after-school activities, and professional development.

The researchers compared the academic achievement of students in SAGE schools to that of students in non-SAGE comparison schools located in SAGE districts. They found SAGE students outperformed non-SAGE students, gaining 25-30 percent of a year’s additional growth by the end of first grade, but with no further gain in later grades. Overall, the researchers conclude the SAGE program increases students’ achievement, upholds gains through third grade, is most beneficial to African-American students, narrows the achievement gap between African-American and white students, and compensates for poor attendance. They did not attempt to distinguish class-size effects from the effects of other SAGE program elements.

California Results Differ

A new RAND study, by contrast, finds that California’s large-scale class-size reduction program–without the other components of the SAGE intervention–has had little effect on student achievement in the Golden State.

In 1996, California enacted SB 1777, which provided a substantial incentive for school districts to reduce their class sizes from an average of roughly 30 students per class to 20 or fewer. As a result of SB 1777, nearly $1 billion in education funds was provided to districts in 1996-97 to reduce class size in grades K-3. The program currently provides more than $1.7 billion a year to schools for class-size reduction.

RAND researchers examined the standardized test scores over five years for pupils in 2,892 schools across the state. Some children had spent only their second- and third-grade years in smaller classes. Others had been in small classes for the first, second, and third grades. All other factors being equal, the researchers found getting the extra year of small classes in first grade did not result in significant test-score gains.

Nevertheless, California’s class-size reduction program did have consequences–unintended ones. Qualified teachers in urban areas fled to higher-performing schools in the suburbs, where class-size reduction meant new teaching positions opened up. Urban schools were faced with huge shortages of classroom space and qualified teachers. As a result, many less-experienced teachers were hired. Since student achievement tends to be more strongly correlated with teacher quality than with small class size, many urban students were actually worse off after the class-size reduction program took effect.

Stanford’s Eric Hanushek provides some insight into the discrepancy in findings between the SAGE report and the RAND report. In The Class Size Debate (Economic Policy Institute, 2002, Lawrence Mishel and Richard Rothstein, eds.), he points out the issue is not that class-size reduction is never effective.

“Surely class-size reductions are beneficial in specific circumstances–for specific groups of students, subject matters, and teachers,” Hanushek writes. Instead, the issue is whether broad reductions in class size across all schools, subject matters, and grades is effective–especially considering the cost compared to other reforms.

Hanushek points to the history of class-size reduction in the United States as strong evidence of the failure of that reform to yield increases in student achievement. Between 1960 and 1995, average student-teacher ratios in U.S. schools fell by one-third. Yet student achievement trends on both the SAT and the National Assessment of Educational Progress (NAEP) showed a general decline in test scores and achievement even as class-size fell during that period.

“If past declines in class size have not had any discernible effect on student outcomes,” concludes Hanushek, “why should we believe that future declines would yield any different results?”

Expensive Reform

Class-size reduction also is a very expensive reform. In Florida, for example, lawmakers are struggling to find a way to pay for a constitutional provision that requires the state to give school districts enough money to lower class sizes over the next eight years. The Florida initiative is estimated to cost $27 billion over eight years–more than double what the federal government spends on Title I education for the entire nation every year.

Other types of school reform may be more effective at lower cost. The RAND study finds that many of California’s standards-based reforms may be more responsible than class-size reduction for any existing student achievement gains.

Meanwhile, many California schools are raising class sizes because of budget constraints. The Elk Grove School district, which Michael Winerip profiled in a positive New York Times portrayal of class-size reduction, is being forced by budget constraints to increase kindergarten class sizes to 33 students per teacher.

Lisa Snell is director of the education program for the Reason Foundation in Los Angeles. Her email address is [email protected].

For more information …

The paper by Phil Smith, Alex Molnar, and John Zahorik of Arizona State University’s Education Policy Research Unit, “Class Size Reduction in Wisconsin: A Fresh Look at the Data,” is available online at

The RAND study by Brian M. Stecher, Daniel F. McCaffrey, and Delia Bugliari, “The Relationship between Exposure to Class Size Reduction and Student Achievement in California,” has been published by the Education Policy Analysis Archives and is available online at

Eric Hanushek’s chapter, “Evidence, Politics, and the Class Size Debate,” from the book The Class Size Debate (Economic Policy Institute, 2002, Lawrence Mishel and Richard Rothstein, eds.) is available online at

Accounts of earlier SAGE results, including online links, are: “Smaller Classes No Remedy for Poor Teachers,” School Reform News, May 2002, and “Study: Student-Centered Learning Ineffective,” School Reform News, July 2001,