Let’s pile on the evidence that across the board cuts in class size don’t help very much. Authors Cho, Glewwe, and Whitler looked at class size reductions in Minnesota. The results are quite unsurprising, although I won’t be surprised if politicians and the public continue to ignore the evidence. First, this is what the authors found:
…reducing class size increases mathematics and reading test scores in Minnesota. Yet these impacts are very small; a decrease of ten students would increase test scores by only 0.04–0.05 standard deviations (of the distribution of test scores). Thus class size reductions are unlikely to lead to sizeable increases in student learning.
Here’s bit about the science behind their work. If you want to know the effect of class size, you can’t just compare the difference in student achievement between small classes and larger ones. The problem is that there are probably other things different between the classes, not just the size. For example, ambitious parents might maneuver their kids into smaller classes. The kids do better because of the pushy parents, not the smaller class. Or savvy parents might move into school catchment areas with better schools–pushing up classes sizes.
What Cho, Glewwe, and Whitler did to get around this problem was to use random variations in births in different areas of Minnesota as the source of variation in class size. This let them be reasonably certain that the student accomplishment was not reflecting some unobserved factor.
Nice work. It’d be even nicer if someone listened.