In mathematics many of us believe the inverse, that is, one bad teacher before high school can leave a student in poor shape for a long time, in a way that good or reasonable teachers cannot easily undo.

Bad here, though, doesn't mean some percentile rank, it means uncomfortable with content.

I know what will really advance the profession--getting the argument out there that good teachers don't matter!

I'll try my hand at reading the math, but now you mention it, your analysis makes sense. I'm constantly having students come back and thanking me, but I've never had an alumni come back and thank me for raising test scores. We discuss what they learned in our class, what they remember, and how it was relevant. Not surprisingly, my former students retain a way of looking at the world, of looking at school work, an interest and motivation etc. Its always fun to hear about the specifics they retained and why, but I don't doubt that standardizing testing improvements would fade.

I also know that some of the things we learned in class are discussed afterwards on the street corner as homeboys are hanging out. I wish they had left the gang, but of course I'm incredibly proud of that compliment.

But getting back to the teacher quality argument, under what scenario would you find inner city neighborhood schools where students would get four or five warm bodies in a row? Has anyone ever heard of a high poverty neighborhood school where you could get four competent teachers in a row? In my experience, year in and year out, regular students find that about 1/2 of their classes are functional. In many cases, those classes have a good teacher, but given the circumstances, they aren't close to effective.

But I've got news for the left and the right wing true believers. If they could coerce districts into coercing effective teachers to transfer from effective schools, we would still end up the same. Those effective teachers would bail out immediately, unless we did something about chronically chaotic schools. Its the dispect shown to teachers, and the inability of the system to let teachers teach, that has driven so many great young teachers out of high poverty school.

Its such a shame, because if you can do it and maintain an emotional equilibrium, teaching in high poverty schools is the greatest job in the world. Show some respect for teachers and we could provide four or five effective teachers in a row. Until we can address systemic realities, though, the best we can do it just get rid of the most ineffective teachers even if we can't find more that a Long Term Sub to fill in.

John #1 - If the post was unclear, I apologize - but the finding of this paper is not that good teachers don't matter. Even if students don't take forward the entire gain provided by a highly effective teacher, they are still better off having an above rather than a below average teacher. It's simply that the size of these effects on test scores may be smaller than we previously thought - and getting a better read on these effect sizes is important when we're making claims about closing gaps. And as John T. points out above, teachers "matter" in lots of other ways as well.

John T. - You make an important point - that the contexts in which teachers teach enable and constrain good teachers. Plugging in exceptional individuals will never be enough - something that a lot of the teacher quality debate misses.

Jonathan - I like the way you flip this question around.

I've always been surprised that people are so quick to label teachers as "good" or "bad" without considering their teaching environment. As John T. pointed out, a good teacher in a bad teaching situation may not create a stellar outcome. Also, teachers who excel in one type of school are not necessarily going to excel in all types of schools.

For instance, I found that in inner city schools, a highly academic teacher, who would be perfectly suited for teaching AP Physics in the 'burbs, may not have as much success with the kids as a teacher with no graduate degree but with great classroom management skills. A PhD isn't going to do much good if the kids won't sit still for the lesson!

As eduwonkette noted, serious socio-economic problems bedevil students in low income schools -- problems that even the "best" teachers aren't going to be able to fix by themselves.

It's tough to say precisely what a good teacher is, particularly when your sole measure is test scores. A social studies teacher of my acquaintance tells me in his previous school, about 30% of his students passed the Regents exam, and that he had the highest passing rate in the building. In our school, his passing rate is closer to 90%.

He swears his methodology hasn't changed at all.