One of the arguments made against the rehabilitative effect of prisons is that prison teachers criminals to be better criminals by putting them into a situation where they can freely network and disseminate ideas. Long and short: it is often argued that prisons don't rehabilitate because they get criminals used to being around high concentrations of criminals.
The same can sometimes - certainly not always - be said of special education. When one puts a student with behavior or social problems in either a self-contained class of nothing but special education students, or an inclusion class with many special education students, often, that child will become worse because he now has others to play off of.
In particular, my high school has a "social skills" class. At first, we all thought this was a great idea, but now, it has become jokingly referred to as the "antisocial skills class." Two of the students on my case-load (who I think are decent kids around the right people) have become worse because they have begun hanging out with some of the worse kids in the class, and have developed those kids' attitudes. One, who I was helping to quit smoking, has now been overheard about smoking cigarettes and weed after school.
Another example is a student on my caseload who is very smart but has social problems interacting with peers. This student is quite mature when dealing with "good" students and quickly changes into an "in your face" student when dealing with "bad" kids. In Spanish, the student does fine (and even avoids conflict), but in "special ed classes" like Social Skills (or a Government class with many other students with learning and other disabilities), this student becomes much more difficult and obnoxious, even instigating verbal altercations.
The problem is that, as much as we special educators try and deny or excuse it away, "inclusion" classes - classes with higher numbers of special education students - are generally the classes with the highest proportion of behavior problems, underachievement, and distraction. It seems to make less and less to put learning disabled students in classes where less learning occurs - "inclusion" classes. And there seems to be no wisdom at all in putting an ADHD student in the class with the most non-academic distraction.
So, the dilemma is that while we all want to see heterogeneous groupings of students, we all know that higher percentages of special ed students in a particular class means the less productive that class is likely to be. And while we want to "spread the special education students" out so that we can avoid high concentrations in relatively few classes, that would bring up a "manpower" problem because special educators cannot float around to too many classes (which is why putting all the special education students in relatively few classes makes more "manpower" sense).
I suppose the situation of whether to put learning disabled students in a "regular" or "inclusion" class should be made on an individual basis. The question I have begun asking myself is whether x student would benefit more from the individualization and assistance that an "inclusion" class (usually with a general and special educator) could provide is outweighed by the student's need for a distraction-reduced and more focused environment that could be provided in the "regular" class.
As a result, I have placed more than a few students in "regular" classes that, for all intents and purposes, should have been in inclusion classes simply because I feared that the student would be undercut by the inclusion environment. If the goal of special education at the high school level is to make students more independent, I think that, done cautiously, will be a good decision.