The class size tradeoff

Lowering class size is an education intervention popular among both parents and teachers. When given the chance, voters have eagerly endorsed low class sizes in states from Florida to California.

Still, like other interventions, lowering class sizes carries a cost. Resources allocated to readying more classrooms or hiring more teachers cannot simultaneously be spent on new technology, professional development, or raising teacher salaries. What do we know about the tradeoffs?

Research done by Andreas Schleicher of OECD using PISA scores is suggestive. Schleicher looks at K-12 school spending in OECD countries in terms of salary cost per student. For primary school students, the US spends several hundred dollars more per student than our international peers. For "lower secondary" students--approximately middle school--we spend essentially the average amount; and for high school students we actually spend less than average. (We do spend more on non-salary costs such as capital expenditures.)

But what's interesting is what contributes to that salary cost. Relative to GDP and to other college graduates, US teachers' earnings are only about average. We spend relatively less than other nations on professional development and time for teachers to collaborate. We save money by asking teachers to spend much more of their school day teaching than in other, higher-performing countries.

So if our higher per-pupil costs aren't going to high salaries, great professional development, or allowing teachers non-instructional time to work together perfecting their lessons, where do they go? To lower class sizes.

High-performing countries like Korea and Japan tend to tilt their spending toward teacher salaries that are high relative to other professions requiring similar education; toward supporting teachers with time and coaching; and toward lengthening the school year. They "pay" for these luxuries with large class sizes. We have chosen the opposite tack, squeezing everywhere else in service of keeping class sizes small. Schleicher's research suggests that for us, the tradeoff has not paid off in terms of student performance.

How have we become so convinced that lower class size means quality? Everybody "knows" it; average class size even figures into college rankings. The most convincing evidence came from the STAR study in Tennessee, which showed that in the earliest grades, particularly among poor students, lowering class sizes below 17 in a class led to lasting benefits in student learning.

The results did not extend to later grades, class sizes in the twenties, or middle class students. Moreover, California's huge experiment in lowering class size in the early grades could not be shown to provide any benefit, perhaps because meeting it required hiring just about anybody with a pulse to teach in hastily constructed portable classrooms.

Perhaps, then, quality of teaching, supported by competitive salaries, high quality ongoing training, and time for teachers to work together as a team is more important for most students' progress than class size. As we try to do better with less in K-12 education, will teachers, administrators, and parents be willing to consider letting class sizes rise closer to the international average? Or will we stick with our gut feeling that smaller is better without regard to the opportunity costs involved?