Believe it or not, the research findings on this are not that strong. I should see if I can look them up. I know it sounds unbelievable and my instincts tell me otherwise, too. However, my DD got a better education in a class of 18 than in a class of 11.
ETA: OKay, so there is one large study (Project STAR) that showed some decent effects for grades K-3. Other than that, the effects are quite mixed. Florida implemented a class size limit at great expense. Thus far there have been no clear positive effects:
http://www.hks.harvard.edu/news-events/news/press-releases/pr-pepg-research-may10The STAR project seemed to involve classes I'd think of as genuinely small (13-17) - is it maybe the case that some of the other research that hasn't shown an effect has been looking at larger classes? That's what I remember from some of the research I've seen, although it isn't something I've looked into properly. I think it would be tricky, because clearly small classes are going to be
automatically better - at best, they might be better if the teacher takes advantage of the extra time/student available, and that takes time and effort to learn (if it's possible at all).
My own impression/hunch of how the small classes work at DS's school is that the most important thing is not really what happens in class, so much as the effect on the teachers' planning and general energy level. If it is easier to teach a smaller class, more of the teacher's energy is left for addressing the difficult problems we want teachers to address, like how best to help each child. DS's school also has a no-homework policy which surely also helps (less marking!) - but the thing I most appreciate there is that they do seem to have time and energy to be genuinely interested in thinking about how to help DS. I can't help but think that if each teacher was dealing with twice as many children, this would be less the case!
