It doesn't take an economist to tell you that just one or two really disruptive kids in a class can absorb a disproportionate share of the teacher's time and make things harder for everyone. But two economists, Scott Carrell and Mark Hoekstra, have studied the question of just how much effect does a disruptive kid have on the outcomes of the other students in a class. Here's the paper, here's a less technical version of it, and here's a Freakonomics post about it. (From last summer, although the paper is dated this month -- I guess there was a pre-publication version circulating.)
They were able to link school records with court records for domestic violence cases, and looked at both the children of the parents in the DV cases, and their classmates. And they did find worse school outcomes -- poorer grades and more disciplinary infractions -- among the children who had classmates from these troubled families than among their peers in other classes and even than the same children in other years.
This is elegant research, but the findings shouldn't shock anyone. Catholic schools pay their teachers worse than public school teachers (even when they're not nuns) and have bigger classes, but their not-so-secret advantage over the public schools is that they can kick the troublemakers out. The Great Expectations School isn't great literature, but it's a brutally honest report by a rookie teacher of how the classroom management challenge just kicked his ass and made his teaching skills pretty much irrelevant.
This also once again makes me wonder how much of the success of places like KIPP is due to selection. I don't think they can expel kids any more easily than the regular public schools, but making parents go through even modest hurdles to enroll their kids probably winnows out a lot of the most troubled kids. And is that ok? If the lifeboat is sinking and they can rescue some of the kids, but not all, isn't that better than letting them all drown?
But what are you going to do with those kids? Carrell and Hoekstra's methodology reminds us that these "troublemaker" kids aren't bad seeds -- they're kids with pretty messed up home lives, who are probably used to violence as a way to solve problems, even if they're not being hit themselves. If anyone deserves help, they do. But it's probably too much to ask of overwhelmed teachers that they be the ones to provide this help, while also trying to teach 25 other kids to read.