More By Author
November 02, 2009
Leave it to Rick Hess to find the current lightening rod issue. The other day it was an interview with KIPP CEO Richard Barth, who was discussing the recent study of the network's success in getting kids through college: 33% of KIPP students who had completed eighth grade ten or more years ago (this was the early days) finished college within six years.? Rick's Q&A is worth the read to hear Barth talk about the challenges of tracking KIPP kids through college ?(something that the Christian Brothers (see here) have been doing for a while); about the lessons KIPP has learned (better expand to K?12); and, especially, about transparency (why would you sponsor a study that could make you look bad?). ?As Barth notes, in answer to the transparency question, one of the good things about funding studies like this is to remind their teachers ?how difficult this is?.? [T]his is the mountain we're climbing.? (No miracles here.)
But the issue that caught my eye was that of whether KIPP has anything to crow about.? In an early question to Barth, Rick says, ?some critics have asked? whether KIPP's long-term college graduation rates are ?really four times the comparable cohort, given that KIPP students have chosen to attend.? ?The key phrase here, of course, is ?chosen to attend.? And Hess is asking the question that has dogged charters from the beginning.? Do they succeed because they do a better job educating kids or because they select better, more motivated kids? (Does Harvard make the man or does the man make Harvard?)
One of the more succinct statements about the ?problem? was a letter to the editor of the Wall Street Journal last month.? One Daniel Cooper was responding to a Joel Klein op-ed piece (Scenes from the New York Education Wars). Cooper, a parent, wasn't persuaded by Klein's random lottery argument to counter the skimming charge because ?it is a random lottery among the self-selected.?? Charter students have, as Rick says, ?chosen? to attend.? And, continues Cooper:
Once chosen by the random lottery, there are further conditions for families to satisfy. When my child was chosen to be offered a seat in a Harlem charter school, the admissions packet we received included a commitment card to be returned by a specified due date, a set of 20 forms to be completed, and an announcement indicating that a parent and the child must attend a mandatory 5-hour enrollment meeting on a specified Saturday in May as a condition for accepting the seat. These requirements are not onerous enough to deter a family that is committed to a child's education, but almost certainly they would screen out a family that is not so committed?. [C]harter schools benefit from filtering mechanisms unavailable to nearby district schools?.
True? ?Barth's reply to the ?chosen to attend? question is interesting:
Again, we welcome these tough questions. What the Mathematica research is showing is, in the case of academic readiness, our fifth graders are coming in really, really behind. They are coming in farther behind the students in districts in which these schools are located. Over time, our research is showing that our schools can make a big difference. And we're incredibly proud of our outcomes.
Good answer, except that he doesn't really address the ?chosen to attend? problem. ?He's defensive.? My answer would be this:? Part of what KIPP ? and every other charter ? does is unleash pent-up demand. It doesn't steal motivated parents; it creates them.? How?? Charters market ?themselves as better schools, for one thing.? Hype?? You bet.? Carnival barkers?? This way to the Education!? Previously unmotivated parents suddenly get motivated; no one has ever vied for their attentions, surely not the neighborhood schools that think that most of them are unmotivated. ?Traditional schools would do well to emulate charters, work to earn parents respect and ?chosen to attend? status by creating good schools.? This way they might find out they have better parents and kids than they thought.
--Peter Meyer, Bernard Lee Schwartz Policy Fellow