Thursday, August 1, 2013

Tony Bennett: Songs in the Key of C, No Wait, A

Here's my two cents about the Tony Bennett grade changing story that Tom LoBianco broke. This was hard to write--with blogging there can be so much pressure to be timely that conflicts with my wanting to read and carefully consider everything on the subject. I'll do my best here.

My initial concern was that the liberal media would treat this merely as a "bad actor" case, you know: Look how corrupt this GOP guy is! Republicans are corrupt! I mean, look at the headline of the original story: "GOP Donor's School Grade Changed." I don't share most Republican ways of thinking, but corruption is political party-blind. In turn, I was concerned that conservatives would treat this as a "political hit job" conspiracy on the part of the liberal media or "opponents of reform." Like this.

Looking at all of the coverage--the journalistic coverage, the analyses, and the defenses (the list in the link doesn't include the Fordham Institute's Mike Petriili's defense, Rick Hess's of AEI interview with Bennett, or this article about the context of Bennett's decision)--there's a real disconnect. Many skeptics of current education reforms such as test-based accountability are saying this is corruption and by design. Many accountability hawks say there's a reasonable explanation for this and that there's little to nothing wrong with what Bennett did. If you read the defenses and the interview, they are earnest.

Now, don't get me wrong, the whole campaign donation business is shady and certainly the charter sector is ripe for crony capitalism and hucksterism--just look at what goes with the charter sector in Florida (speak of the devil) and the White Hat charter school company in Ohio. But I have no evidence that the donation influenced Bennett's thinking or that he's particularly corrupt. What he does seem to suffer from is a deep certainty that he's right about his education reform policies and that the statistics must be on his side, even if that means having them fixed. I've heard it said about Michelle Rhee when she was in DC that it was almost like she wasn't telling mis-truths when she did. She was so convinced of her own rightness that she couldn't hear herself saying one thing one day and a different thing the next. It was all the truth to her. I'm afraid that Tony Bennett seems to be suffering from this malady, as well.

How it worked in Bennett's office seems to be how it works in the work places of reformers. "Choice" and charters as a model are always better. It's okay if we lose a few neighborhood or comprehensive schools because those are probably failing or close enough to failing anyway. How many times have we heard that it's okay to sacrifice a few good teachers here and a few decent schools there in the service of "objective" evaluations system? Systems that will largely weed out the bad and identify the good. So, you lose a few good teachers. So you close some decent neighborhood schools. Oh well. No use crying over a little spilled milk. In the face of schools like this one being labeled failing, and teachers like this one getting a fire-ably low evaluation, how many times have the proponents of such systems said, Well, the evaluations are not perfect but they're better than what we had before.  (Um, who has demanded perfection?)

And so two things happened. First, a double standard: a well, it's not perfect, sorry was not issued in the case of Christel House Charter (and does anyone else find it unseemly that the school is named after it's director and maybe biggest donor? Is that what a public school should be?) Then, the formula was fixed so the appropriate grade would come through. The math was done in such a way that it didn't maintain the integrity of the formula. Sherman Dorn's analysis shows why this was unethical while Anne Hyslop explains how the math doesn't add up.

Surprisingly, most of my thoughts mirror the thoughts given at this Fordham forum published today, which is not to say I agree with all of the thoughts expressed--I especially think the word "flap" in the title is a pretty glaring understatement. But otherwise it is the most frank, humble, and thoughtful  discussion of the limits of school grading I think I've ever heard from accountability hawks. I found myself nodding in agreement. This particular process was unfair and showed favoritism. These processes need to be more science than art. We need an objective measure that everyone adheres to and this can provide that. These grading systems aren't ready for prime time. Schools shouldn't be boiled down to a single grade. Please read the whole thing.

But I was still left wanting.

1) Where in this conversation is the statistician to discuss the validity of the school grading process? First of all, I'm pretty sure these metrics--school grading metrics as well as teacher evaluation metrics--are being used in ways they were never intended to be used. I admit when I'm in huff, I refer to them as "junk science," but as Matt DiCarlo points out here, they're not junk science even if they're just being used in junky ways. If these metrics were being used as thermometers that would be one thing, but they're being used as a hammers. Second of all,  I'm not at all convinced that there is such a metric system out there somewhere waiting to be formulated that would ever work well as a school grading or teacher evaluation system. And I'm not going to take Fordham's or New America's or Tony Bennett's word for it that there is, any more than I would take my own word for it that there isn't. I want to hear from an expert. If anything, the stats people in the Bennett e-mails show that Indiana's school grading process is not valid. Having to "run different options" to arrive at a desired outcome shows that. What went on in Indiana seems to me at best an exercise in statistical gymnastics and at worst, one in statistical fraud. I have yet to be convinced by anyone with any statistics expertise that these systems are valid. In general, the people with expertise in statistics that I trust the most have been lukewarm at best on their efficacy.

2) Not ready for prime time?! Shouldn't be used for high stakes decisions?! Maybe this isn't a science yet?! There shouldn't be stakes attached to these?! I agree but isn't it a little late to be saying this? What rock have you people been under? All of these metrics ARE prime time. They've been prime time! Remember when Fordham crowned Indiana as its Education Reform Idol? Was the school grading plan not one of the criteria for judging the winner? These metrics have done been high-stakes. That ship has sailed. They've been used to evaluate teachers and schools, to fire teachers, to unfairly label schools, close down schools, to allow for burdensome federal interventions and harsh state takeovers. Lots of people have been saying these metrics are not ready for prime time, if they're appropriate at all. Why wasn't this thoughtful discussion had before spending so much money, before making high-stakes decisions, before making such a mess. Where was the humility and thoughtfulness then?

3) Where was the stakeholder from Indiana in the forum? A superintendent maybe. A parent or two. Why not ask them what they think of school grading in Indiana? There was talk of transparency. Isn't this an accountability measure that's being discussed? Isn't the premise behind these school grading systems to help parents and the public? Aren't such systems produced on the public dime? Isn't someone like Tony Bennett a public servant, accountable to the public? Okay, so no parent on the panel--maybe no one was available. Look at the public's "accountability moment" then. Didn't the public speak loud and clear when they failed to re-elect Tony Bennett? As Kombiz Lavasany tweeted: " Two Republicans lost in Indiana last year. One was Richard Mourdock for his rape comments. The other was this guy [Bennett]." The school grading plan was the center of his policy platform. And he lost. What do you think the public, who ostensibly this metric was to benefit, then thought of that policy?

This is exactly what happens when you rush into big, "disruptive" changes without thinking about them or fully understanding what you're doing. You break things that weren't already broken and you make messes. And it's what has happened over and over again with education reform in this country. It's time to call a moratorium on school grading and teacher evaluation metrics and maybe on some other stuff, too.


  1. Well stated. I take exception to this...

    "How many times have we heard that it's okay to sacrifice a few good teachers here and a few decent schools there in the service of "objective" evaluations system?"

    ...because I've never heard it stated. Maybe I don't listen carefully or engage deeply or often enough. But it seems that the accountability hawks' opinions are even more cocksure than you suggest. Breathtakingly so. Maybe unbelievably so, because these people aren't dumb.

    1. @Bill Thanks for reading and commenting. Well, mostly you hear that with the "it's not perfect but. . ." lines.

      Frankly, no matter the system, there's always going to be some for some poorly run schools and some teachers who shouldn't be in the classroom--that's unavoidable. The problem with modern education reform as I see it is that while there is some good (I say "good" as an ESL teacher), it's doing more harm than good over all and it's destroyed a lot of the good that was already there. And it seems to be in many cases replacing the bad with something equally bad if not worse, denigrating intellectualism in favor of entrepreneurialsm, and in many cases undermining democratic principles and eroding democratic institutions.

      That cost in damage to our society of some poorly run schools and some bad teachers is high to be sure, but the cost in damage to our society of modern ed reform is astronomical.

    2. Bill,

      VAM advocates are saying that - albeit with the words that they borrowed from a California federal case that they see as precedent. They assert that false positives can do more damage than false negatives. hard to believe but its the latest spin.
      fyi This makes a case similar to Rachel's

  2. The malady you're talking about Rhee having is psychopathy. She has all the major symptoms of a psychopath.

    1. You know, you're not the first person to (seriously) suggest that there is some sort of pathology there. I couldn't help but be reminded of Rhee when I read the story of Rachel Yould.


Note: Only a member of this blog may post a comment.