Not yet finished with the article, which is equally dismissive of past democratic and republican efforts to improve schools over the past decades. So far though, I think this critique of RTTT gets at the heart of the problem with Duncan’s innovative–but in the end still ineffective–attempt to implement federal reforms in a way that does not produce the perverse incentives of NCLB (for a moment, please put aside whether you agree or disagree with his specific agenda and think more about the policy design):
“The result was a competition that did more to reward grant-writing prowess and allegiance to the fads of the moment than meaningful, structural changes. Race to the Top turned into a bonanza for consultants, with state plans offering hundreds of pages of indecipherable jargon and unenforceable, poorly articulated (but nevertheless expensive) reform schemes. Inevitably, Race to the Top disappointed. Every one of the first- and second-round winners has failed to follow through on key provisions of its plan — and yet all continue to collect their federal awards.”
Very curious to see if his alternative is as coherent as his critique.
Okay. Here are his solutions: federal government needs to adopt two primarily roles: data collector and union buster. Regarding the former, he stresses–I believe rightly so–that we need a shift away from measuring inputs and towards measuring productivity (he is thinking more about labor/college success, I believe, than test scores). This intrigues me. Part of what sucked about NCLB is that its incentives were so stark and easy to measure that schools, acting rationally, did everything they could to boost those statistics. However, what if schools are measured in ways that are much harder to directly adhere to: high schools could be measured by percent who graduate from a four year college (or have some sustaining career). This get’s murkier in elementary and middle school, but I think the idea is interesting. I believe the best thing KIPP has done is start trying to measure and teach character, which was in response to trying to boost a more amorphous statistic: college graduation rate (can’t teach to that test). Perhaps designing long-term incentives (or even just gathering that data) for schools to prepare students for later success could have a degree of meaningful accountability without the level of perverse outcomes. Also, in general, I think the more short-term the incentives, the more likely districts/principals will try to boost their statistics in superficial ways.
While problematically vague, my other concern is that Hess’ philosophy does rest on the assumption that school choice will be ubiquitous. Hess admits choice is far from a panacea…but these two modest roles for government (data gatherer and union buster) hardly seem to be total game-changers in allowing school choice to flourish. I mean, if Sweden can’t pull off a well-regulated system of school choice (with very reasonable unions, incredibly great data and you know…homogeneity and income equality), how in the world could we?