This is a new step in a heated discussion in many places. You’ll hear quite often policymakers state that class size has little of no influence on learning. No influence is actually wrong, but e.g. in the meta-review by Hattie there are many more things policymakers can do that have a bigger influence (e.g. school size!). At the same time teachers (and parents) often find this counter-intuitive.

The thing is that examining the effect of class size isn’t that easy. Why are children in a smaller class? Is it because they are more wealthy? Is it because they are living in a more rural area? Is it because they have special needs? There can be many reasons. Researchers do try to eliminate as many of these possible influences as possible, but than there is a second problem. To measure the influence of small class sizes, you need a longitudinal approach, a costly kind of research.

To my knowledge there are not that many longitudinal studies, and this new report by Diane Whitmore Schanzenbach for the National Education Policy Center  at the School of Education, University of Colorado Boulder, relies heavily on two of those studies. Both the STAR-study as a recent Swedish study show that there are indeed benefits of reducing class size (the Swedish study states that there is a clear return on investment).

The conclusions of this report:

• Class size is an important determinant of student outcomes, and one that can be directly determined by policy. All else being equal, increasing class sizes will harm student outcomes.
• The evidence suggests that increasing class size will harm not only children’s test scores in the short run, but also their long-run human capital formation. Money saved today by increasing class sizes will result in more substantial social and educational costs in the future.
• The payoff from class-size reduction is greater for low-income and minority children, while any increases in class size will likely be most harmful to these populations.
• Policymakers should carefully weigh the efficacy of class-size policy against other potential uses of funds. While lower class size has a demonstrable cost, it may prove the more cost-effective policy overall.

Will this close the debate once and for all? No, not at all, I’m afraid. Critics may (and will) argue that the report omits a lot of other research showing that there is little impact, although the author does have a point in stating that most of that research is actually correlational rather than looking for causal impact. The most relevant element in the report to me is an insight that I had overlooked when reviewing the evidence myself for our book on myths in education:

“Most of the high-quality evidence on class-size reduction is based on studies of the early grades. The available high-quality evidence on the impact of class size on outcomes in older grades is more limited, and more research in this area is needed.”

Although the Swedish report does mention some insights on this topic, there is surely a need for more research on the impact in older grades indeed.

You can read the whole report here.