If that's true, Kate, then why are you guys pushing so hard to pay teachers based on (highly inaccurate) measures of student performance?
I still contend that one of the largest problems of the corporate reform movement is that it is being run by people who don't have enough real-life teaching and administration experience. This is a great example. Walsh admits SES "swamps" school effects (of which "teacher quality" is still a subset). Yet she supports using student test scores as a basis for awarding merit pay.
But every elementary teacher knows classes are different. Yes, principals will makes class lists that attempt to distribute academic talent, but they also take into account special needs, disruptive behavior, parents (yes, folks, they do), friendships, and a whole host of other factors. I've seen principals who've unintentionally given a teacher a rough group try to make up for it by giving them an "easier" group the next year. I've seen principals who identify a teacher in a grade level as the one who can get "tough" kids back in line; their Value-Added Modeling scores would never be as high as their peers, but that doesn't mean they aren't doing the job well.
If we're hellbent on using test scores and VAM to rate teachers, all of that has to be thrown away. The only fair way to proceed would be to instead sort the students demographically and then evenly distribute them among classrooms. Let's leave aside that sorting the kids would be impractical, invasive, and probably illegal ("This kid goes in the Asian/male/lower-middle-class/single-parent/heterosexual(we think)/non-college-educated-parents/ESL pile..."). How would you distribute them? By school? Across a district? Across a region?
"Hey, your school has too many white kids; I'll trade you for three free-lunch students and a Pacific-Islander to be named later..."
Are we prepared to go through this just so we can have a teacher evaluation system based on invalid and unreliable measures that is supposed to root out an unspecified number of "bad" teachers? It's insane to even consider it - if you know something about how schools work.
These people clearly do not; further evidence is found in a footnote to Walsh's piece:
So higher SES parents "choose" "better" teachers. Does Ballou think those teachers who teach high-SES kids would be just as good at teaching low-SES kids?
I'm not even referring to the obvious circular logic here ("good" students make "good" teachers - duh): I'm asking, does a teacher who knows how to reach affluent, suburban, white kids necessarily know how to teach poor, urban, minority kids? Could it possibly be that some teachers teach particular groups of kids better than others?
The answer is obvious, and it negates Ballou's argument. But that's hardly surprising; this stuff is being worked out in the rarified air of Washington think tanks, far from the odor of Crayolas and mystery meat. These people have not reached out to those of us down in the trenches; they consider us the problem, and have convinced themselves it's their duty to "fix" us.
If they'd just stop and listen, maybe they'd see that they are making mistakes that are going to haunt public education for years to come.