Dear Steve, Part Three. An Open Letter to the Education Reform Movement

I’ve made it to the last part of my letter to you, Steve. Maybe you’re just a bit relieved. It’s been a long letter after all. But I think you and other education reformers may be acutely interested in this part of my letter as I’m going to present an idea that may have political legs. It’s an idea that you ought to put your energy behind, and that teacher unions ought to aggressively support.

The issue has to do with merit pay. Here I need to say right up front that I’m not necessarily a supporter of the “bonus for performance” approach you mention in your Op-Ed. Frankly, it feels to me as a teacher that when you (and others) suggest this kind of bonus system, what you really want is to impose the culture of the private corporation on the culture of teaching. You appear to believe in your bones that if only teachers had the right ambition, the right values, the right drive, they would welcome such a bonus system. Maybe, like Dr. Strangelove in the old film, you can teach people to learn to love, if not the bomb, then at least the bonus and the particular ethical and normative universe of private sector corporate culture.

No, you’re not really saying this, you may argue. But think about what I hear when you state in your Op-Ed that “talented people” are currently being “deterred” from teaching careers by the stigma of a tenure system under which “nonperforming” people get to stick around. What I hear is that you think my colleagues and I are not, in fact, already talented. Worse, I hear that you think we need a program of moral uplift and discipline imposed upon us. The bonus is the method in the system you want to impose. Those who do not accept the method are apostates.

Still, as I said before, I do want a system of merit pay, and very significant merit pay in the multiple tens of thousands of dollars, to exist for the very best teachers who choose the very hardest assignments. If you hear me out I think you may see the merit, as it were, of focusing merit pay systems only on the few, and leaving most teachers, and teacher compensation systems, well enough alone.

I already shared this idea in a piece for KQED Public Radio’s Perspectives program in October 2012. Here I merely want to review and slightly expand upon the idea KQED allowed me to transmit to their listeners some five years ago.

My idea derives from three principles. One is fairness. Students needed to be treated fairly, and under the current system, the students in greatest socioeconomic need are the very students most in need of equal access to the best teachers. I’m not just saying this. The data overwhelmingly support my case regarding lack of equal access to the best teachers. I’ll point to a great study by Goldhaber, Lavery, and Theobold here. This study found that no matter what measure of teacher quality you look at, whether it be years on the job, value-added score, or licensure exam score, disadvantaged students get less qualified teachers in their classrooms, no matter how you define disadvantage. Most of this “teacher quality gap” comes from teacher “sorting across districts and schools,” but it also comes from “sorting across classrooms in schools.”

Here I am clueing you into a dark secret, Steve. This is the place you and other education reformers can legitimately criticize teachers. This is our weak spot. It is a weak spot because a good portion of the “teacher quality gap” comes from teacher choices. We know this in our hearts to be true.

This is not to say that there aren’t amazing teachers currently teaching low-income populations, but it is to say that we clearly need a lot more of them. It is, to be honest, about the fact that teachers with choices, like me, tend to choose to make their already difficult jobs easier by teaching populations of students who, if not universally easy to teach, are not the very hardest.

So in the interest of fairness to students, to receive the multiple tens of thousands of dollars in merit pay I have in mind, to enter the role of what I have previously called Tier 3 teaching, teachers would need to agree to be moved to the spots in their districts, or perhaps in their metro areas, where the highest need exists. Or, if they already teach in high needs schools, then they would need to agree to be moved to high needs classrooms. Period.

Teachers also, of course, need to be treated fairly. It should be obvious that any system of merit pay based on student test scores needs to focus exclusively on score improvements, not raw scores. But fairness goes deeper than this, which gets me to my second principle. Aside from valuing fairness, as a math teacher, I also value respect for mathematics. And I know enough about that newest major branch of the tree of mathematics we call statistics to say that it’s hard to prove that a data value is statistically significant.

We underestimate how terribly hard this is to accept. As human beings, our brains are hard-wired to search for patterns and hierarchies lurking behind every bush. After all, seeing patterns in bush movements may have saved our distant ancestors from getting eaten by predators hiding in said bushes. But statistics as discipline forces us to accept on principle the notion that a lot, really most, of the phenomena out there in the world may be random. This truly is hard. Most of us want to believe that things happen for a reason. Statistics absolutely requires us, as a matter of mathematics, to assume that they don’t unless we can prove otherwise.

To turn back to less abstract topics, if education policymakers were to really accept statistical logic, they would quit trying to focus on how teachers in the wide center of the bell curve, that is to say, most teachers, affect student test score improvements. They would drop their adherence to the belief that all teachers can and should conform to measurable systems of discipline based on test scores and money bonuses. Instead, they would adopt the mathematical view that most teacher-generated effects on student test scores will be no different than the effects that might be generated by a teacher selected at random from the sea of reasonably well-trained and credentialed workers.

On the other hand, teacher unions and others opposed to looking at student test scores also need to get with their mathematics. They need to accept the fact that some teachers will generate student test score improvements (and diminishments) that are almost certainly non-random. These are the teachers out toward the tails of the proverbial bell curve, in not just one year but most years.

This gets me to my third principle. This principle is that it pays to focus on the positive. Yes, it’s true that teachers who, year after year, struggle more than other teachers to help their students learn, need to be subject to systems of retraining and discipline. Yes, there do need to be better ways to move the very weakest teachers out of the system, and yes teacher unions can sometimes be obstructionist in this regard.

But education reformers like you Steve consistently make a vast political mistake. You focus relentlessly on the problem of “bad” teachers. Listen to yourself, Steve. In your Op-Ed, you state that “we can’t fire our way to a better teaching force.” But aside from the impression left by this statement that you would if you could, please simply notice that the bulk of your policy prescriptions are stuck on a good teacher/bad teacher dichotomy. You seem to focus on tenure, talent, and merit pay using a mental model not of a statistical bell curve but of a pie chart. One chunk of this pie chart is the good teacher chunk. The other is the bad teacher chunk. Insofar as we can grow the size of the good teacher chunk, the world will be a better place and we will somehow get better at the social project of what you describe in your Op-Ed as “reducing inequality.”

I’m going to take a risk here of saying a couple of things that I hope you don’t find too patronizing. But you have to remember that you opened yourself up to this kind of criticism by shaming me on my first day of school.

First off, I want to say that if the pie chart is one’s implicit mental model, then it’s easy to see how one might be sorely tempted to focus strongly, even obsessively, on growing the good part of the pie by getting rid of bad teachers on the bad part. You and other reformers need to drop the pie chart worldview and think more mathematically in terms of statistical significance. 

I also want you to reflect seriously on that corporate and business-oriented worldview I mentioned earlier. If you adopt a worldview where virtuous people are rewarded for merit and merit and virtue are one and the same, then it makes sense why one might be obsessively focused on weeding out bad teachers. But the thing is, Steve, virtue, and merit may not be as linked, or as measurable, as you think. Some teachers may be incredible for a few years. Then their priorities might change a bit, or they might just hit a dry spell. They might have a baby or a sick relative. They might actually be human, in other words. 

I really want to tell you that both the pie chart mental model and the corporatist, neo-libertarian mental model are wrong. This is not just to shame you back. Instead, I truly want you and other education reformers like you to get past the barriers that seem to keep you from directing your efforts toward the positive. What if you took all that energy focused on railing against teacher unions and teacher tenure, and turned it toward developing a system of radical rewards for the very top teachers out on that positive, right-hand side of the bell curve?

I can review what I am proposing very quickly and very simply. Any merit pay system needs to focus mostly on the positive and needs to be grounded in principles of fairness and respect for mathematics. And yes, there should be a merit pay system in teaching. But no, it is not the merit pay system most education reformers imagine. Merit pay should only go to the very highest performing cohort of teachers, and even then only if they agree to go to the highest need schools. But if they do agree, and as long as they can continue to generate significant student test score improvements, then society should agree to pay them at a level commensurate with, say, doctors or lawyers or senior engineers at the big firms near where you live in Silicon Valley.

I want to leave this long essay with an interesting financial problem. The problem is this. It is almost certainly the case that very high-performing teachers do not, in general, become as amazing as they are in isolation. They become amazing in large part through systems and work cultures and relationships that encourage good teaching. If exceptional Tier 3 teachers can create these systems at the neediest schools, in the neediest classrooms, then school districts and the state governments that fund them may find themselves in the position of paying out a whole lot of very high salaries to teachers in schools who are truly turning things around for the neediest children in our society.

Then again, Steve, you did say you wanted better teachers.