The Knowledge is Power Program is one of our current recommended charities, but I think that Sarah Mosle’s critique in Slate is very much worth keeping in mind.

Mosle writes:

While KIPP does have outreach efforts to broaden its applicant pool, only the most determined parents are likely to respond to … sign KIPP’s demanding contract. This dedication suggests a higher value on education within these families, and thus kids better able or willing to learn. And the weakest students, not surprisingly, get disproportionately winnowed. In KIPP’s schools in the San Francisco Bay Area, for example, the worst-performing kids have dropped out (or been expelled) in greater numbers in the higher grades; the result has been to inflate the schools’ grade-to-grade improvement.

We agree that the superior performance of KIPP’s students can’t be taken fully at face value, because they may not be a truly representative set of disadvantage students. Our analysis concludes that KIPP most likely is making a difference for the students that it serves, despite these concerns (and Mosle thinks so as well).

However, just because KIPP is making a difference for the students it serves doesn’t mean its model can be fully generalized to close the achievement gap. For one thing, it isn’t clear how many teachers can be found that are at the caliber KIPP aims for. For another, KIPP appears to be aimed at a particular kind of student. I think Mosle’s closing concern is right on target:

But since the biggest debate about KIPP, on both the ideological left and right, is whether or not its methods can work for all disadvantaged children (instead of just a handful of self-selecting families), why wouldn’t it—and its financial, ideological, and media backers—have a strong interest in answering this question once and for all by taking on an entire urban area or even, for that matter, a single neighborhood as, say, Geoffrey Canada has tried to do in Harlem with his Harlem’s Children’s Zone?

There’s something perversely evasive about KIPP’s opening up just one school in Dallas, one school in Albany, N.Y., one school in Oakland, Calif., one school in Charlotte, N.C., one school in Nashville, Tenn., and so on—as if the program recognizes that its best chance at success is to be the exception rather than the rule in any city where it operates.

I believe anyone pointing to KIPP as “the path to closing the achievement gap” is being far too optimistic, although KIPP is a promising way to improve outcomes for the individuals it serves.

Additional GiveWell materials related to KIPP:

• The summary of our review of KIPP is available here with a link to our full-length review.
• Our overview of programs aiming to increase equality of opportunity is available here.
• We’ve blogged about KIPP here and here.