The extensive research by John Hattie shows that reduced class sizes has a small but positive effect on student achievement but relative to other factors, it is very inefficient:
One explanation to reduced class sizes not having a larger effect is that teachers, given smaller classes, rarely change their practice to optimize for the new situation.
True. If I remember Hattie correctly, he argues something akin to that reducing class sizes is mostly a political move (it's certainly not research based) — and after spending huge amounts of money on that — there is no political incentive to spend at least as much on retraining teachers.
Also, if some of the money had been spent on professional development instead of reducing class sizes in the first place, his research shows that the effect of that on student achievement had been much, much higher.
http://visible-learning.org/hattie-ranking-influences-effect...
One explanation to reduced class sizes not having a larger effect is that teachers, given smaller classes, rarely change their practice to optimize for the new situation.