This NYT article seems incomplete, but I like it for one simple reason: people talk all the time about charters cherry picking or cream skimming kids, but never seem to ask the right question . . . this article does. There's a ton of evidence that most charters do not take the highest scoring students (see, for example, this chapter from this new book) and those data are used as evidence that charters don't skim.
Case closed, right? Do charters skim? No, they don't.
But that's the wrong question. The issue shouldn't be whether charters take the highest scoring students, it should be whether they enroll the best-behaved and/or most motivated students (and then nudge out those who are unruly and/or unmotivated).
In other words, we should be asking if charters enroll kids who are better students instead of asking if they're enrolling students who previously earned higher scores. Why? Once you get a class or school full of motivated, attentive, and polite students it's a heck of a lot easier to teach them. And a heck of a lot easier to see large gains in test scores.
I have yet to see any rigorous analysis of the extent to which charters do, in fact, enroll or retain better students. Instead, I read a lot of anecdotes like the one from the NYT article I linked to above. Were I to hazard a guess, it would be that there's at least one charter out there that enrolls/retains substantially better students than the surrounding schools.
Even if I'm right, whether or not that's a good thing or a bad thing is a whole separate discussion. But let's start that discussion by asking the right questions.