Duncan in Denial

There are many portions of Arne Duncan's educational policies that are... what's the word? Counter-intuitive? Not aligned with reality as experienced by most sentient beings? Baloney? There are days when I imagine that the energy Duncan expends just holding cognitive dissonance at bay must be enough to power a small country (like, say, Estonia).

But nowhere are Duncan's powers of denial more obvious than in his deep and abiding love for Value Added Measures. Arne loves him some VAM sauce, and it is a love that simply refuses to die. "You just don't know her the way I do," he cries, as the rest of us just shake our heads.

At this point, VAM is no spring chicken, and perhaps when it was fresh and young some affection for it could be justified. After all, lots of folks, including non-reformy folks, like the idea of recognizing and rewarding teachers for being excellent. But how would we identify these pillars of excellence? That was the puzzler for ages until VAM jumped up to say, "We can do it! With Science!!" We'll give some tests and then use super-sciency math to filter out every influence that's Not a Teacher and we'll know exactly how much learnin' that teacher poured into that kid.

The plan is simple and elegant. All it requires is two simple tools:

1) A standardized test that reliably and validly measures how much students know
2) A super-sciency math algorithm that will reliably and validly strip out all influences except that of the teacher.

Unfortunately, we don't have either.

We know we don't have either. We are particularly clear on the degree to which we do not have the second. Scan the list of reformster programs, and while you can find plenty of principled disagreement on most points, there is no part of the reformster education platform that has been so thoroughly, widely debunked as VAM-for-teacher-evaluation. The National Association of Secondary School Principals has taken a stand, and if you read their resolution, you'll find not just a philosophical argument, but a list of striking debunkers. The American Statistical Association has made its own statement in opposition. A peer-reviewed study paid for by the Gates Foundation itself, the grand-daddy of all reformster backers, declared in no uncertain terms that VAM tells us nothing about teacher quality. The blog Vamboozled (by Audrey Amrein-Beardsley) provides unplumbable depths of VAM-busting research and essays.

At this point, even the Flat Earth Society would be reluctant to endorse VAM as a measure of teacher effectiveness.

NO portion of his policy has been so thoroughly disproven, and yet no portion of his policy has earned more of Duncan's loyalty. He stopped saying "Common Core" out loud. He at least pretends to be cooling off on testing. Even he has to admit that some charters have issues. And data collection has become the love that dare not speak its name. But VAM still owns a place close to Arne's heart.

Witness the most recent doubling down on VAM, in which Duncan not only pledges his allegiance to the flagging monster, but announces his intention to extend its reach, taking the already invalid VAM ratings of individual teachers and taking a giant leap backwards to use them to evaluate the college that trained that teacher. Is there anybody else who can present this idea with a straight face? Read Anthony Cody here as he takes this proposal down, then note that you have over a month to register your disagreement wit the feds, and do it.

Why would someone who professes such love for data and critical thinking stay so attached to a policy that is supported by neither? Why does Duncan insist on such a mountain of denial?

Well, I can't pretend to see into his brain. But I can see that if Duncan were to admit that his beloved VAM is a useless tool, a snub-nosed screwdriver with a briar-encrusted handle, then all his other favorite programs would collapse as well.

Everywhere we turn in reformsterland, we keep coming back to teacher effectiveness. Every one of the policies and programs either begins or ends with measuring teacher effectiveness. Why do we give the Big Test? To measure teacher effectiveness. How do we rank and evaluate our schools? By looking at teacher effectiveness. How do we find the teachers that we are going to move around so that every classroom has a great teacher? With teacher effectiveness ratings. How do we institute merit pay and a career ladder? By looking at teacher effectiveness. How do we evaluate every single program instituted in any school? By checking to see how it affects teacher effectivesness. How do we prove that centralized planning (such as Common Core) is working? By looking at teacher effectiveness. How do we prove that corporate involvement at every stage is a Good Thing? By looking at teacher effectiveness.And by "teacher effectiveness," we always mean VAM (because we don't know any other way, at all).

If our measure of teacher effectiveness, our magic VAM sauce, is a sham and a delusion and a big bowl of nothing, then a critical piece of the entire reformy puzzle is missing. We have no proof that we need reform, and we have no method of proving that reform is working (we already have means of measuring reform's effects, but we don't like those because the answers are not the ones we want).

Duncan has to hold onto his belief in VAM because without it, the whole ugly sweater of reform starts to unravel even faster than it already is.

VAM is the compass by which reform steers. To admit that it is random and useless would be to admit that our political leaders have been piloting the ship of education blindly, cluelessly, haplessly, that they are steering us onto the rocks and that they have no idea how to get us anywhere else. Either that, or they would have to admit that they've known all along exactly where they were taking us, and the VAM compass has just been a big fat lie to keep the passengers quiet and calm. Either way, admitting VAM is a fraud would be inviting (further) mutiny, and Duncan can't do that any time soon.