Charter Schools: Normalizing Segregation?

The debate about charter schools really boils down to one question: can "successful" charters be replicated?

If a "successful" charter is "successful" because it spends more money, or offers extensive tutoring and personalized learning, or incorporates wrap-around services, then it's replicable. If, however, the charter is "successful" because of peer effect - because it serves fewer kids in deep poverty, or fewer kids who don't speak English at home, or fewer kids with special educational needs, or fewer boys than girls - it will be impossible to scale up its "success" to include most children.

I can hardly claim to have done a comprehensive study of charter schools on this blog; I have, however, detected a pattern:

Chris Christie, like most politicians and pundits who cheer for charters, loves to drop by the "successful" ones and sing their praises. But these "successful" charters are often the ones that clearly owe their "success" to having a student population substantially different from their neighboring public schools. 

Christie isn't alone. Mayor Rahm Emanuel says the Nobel Charter network in Chicago "found the combination to the lock," but the school serves a different population than the public schools of the city. Jon Alter defends the KIPP network, but gets indignant when confronted with peer-reviewed academic research that shows KIPP engages in significant patterns of student attrition.

And then there's the Star-Ledger, predictably toeing the NJDOE line this morning: 

Newark’s charters particularly stood out. They showed some of the highest achievement gains in the country, almost twice that of their peers in traditional public schools — roughly the equivalent of an additional seven to nine months in school each year.

That’s a great indication, but not a slam-dunk. Because there’s also evidence of mediocrity here: Charters in cities such as Camden, Jersey City, Trenton and Paterson did not outperform traditional schools. That underscores the need to vigorously weed out bad charters.

We must be very careful whom we allow the privilege of running a school. State Education Commissioner Christopher Cerf deserves credit on that point: He’s emphasized a cautious expansion of charters, favoring those with a proven track record, such as Newark’s highly successful TEAM Schools — which also aggressively recruit the neediest students.

First of all, TEAM may try to recruit the "neediest" students - and they do a better job than many other "successful" Newark Charters - but they still have fewer poor and special-needs students than the publics, and they do engage in significant patterns of student attrition.

Second, the primary reason Cerf pulled back on charter expansion was because Christie's constituency in the suburbs stood up - loudly - and nixed the idea. Cerf was well behind the parents of this state when it comes to being cautious about charters.

Third, the question ultimately isn't whether we should "weed out bad charters"; it's whether "bad" charters are inevitable, because "good" charters can't be replicated. If "good" charters owe their success to segregation, we can only expand them so much.

This is a point that Tom Moran and the SL Editorial Board consistently refuse to address:

Even though this new study matches students of similar demographics, critics say, the overall populations of charter and district schools still differ. Charters may have fewer students in special education, for example, or the deepest pockets of poverty. That creates a different school environment.

But these researchers didn’t cherry-pick schools: They used all charters for which data were available from the state Department of Education. Some charters might not have been in existence long enough to have consecutive test scores, they said. But none was deliberately excluded.

Oy, this stuff makes me crazy. Can you see how the second paragraph has nothing to do with the first? Let me rephrase this:

Q: Do "successful" charters segregate?
A: No, because the study didn't cherry-pick the schools.

That's not an answer! It doesn't matter if the study "cherry-picked" the schools or not. All that matters is whether the schools serve the same student population as the neighboring public schools!

And p.13 of the CREDO study clearly shows charters do segregate: by race, by special need, and by fluency in English. Worse, because the study conflated different levels of poverty and special needs, the CREDO report undersold the amount of segregation by class and special education classification that occurs in charters - particularly in Newark's.

So the issue is not oversight, and it's not the study's methods, and it's not even the "willingness" of a charter to take the students who are the most difficult and expensive to educate. The issue is whether a "successful" charter can be successful when forced to educate the same student population that the public schools must take.

It's a small measure of success that the Star-Ledger at least acknowledges the question. The next step is to make sure they can't get away with evasive answers anymore.

Attrition? Where?