Dallas' Teachers Unions Are Ready for Combat over Merit Pay

The real issue the teachers unions have with Dallas school Superintendent Mike Miles is not Mike Miles. It's merit pay. They don't like it.

Angela Davis, Region 4D president of the Texas State Teachers Association (NEA) in Dallas, told me: "TSTA strongly opposes merit pay for teachers."

Rena Honea, head of the other big union, Alliance-AFT, wouldn't even talk to me about merit pay.

When I wrote an article for our news blog two weeks ago saying the unions were using attacks on Miles as a shield for their real agenda, defeat of merit pay, a typical response from a reader was this: "Alleged 'merit pay' has absolutely no track record of improving outcomes anywhere it's been implemented."

The national anti-merit-pay camp has more arrows in its quiver. Nationally known education historian and author Diane Ravitch paints the school reform issue as a close cousin of tax policy and fiscal regulation, all about who's on which side.

"It is breathtaking to see how closely aligned are the agendas of conservative governors and the Obama administration when it comes to education," Ravitch writes. "... but corporate reformers think they know best."

But what is it? What would merit pay for teachers be in Dallas? For one thing, it would be based only partially on student test scores. Various versions of a merit pay system under discussion here — already presented to the Dallas school board in briefings over the last several months — would use student achievement scores for less than half of the equation for measuring merit.

When "No Child Left Behind" was new in Washington under President George W. Bush, people were going out scooping up raw student test scores without any allowance for where those students had started out the school year. A teacher who inherited a roomful of future rocket scientists looked golden — all high scores for her kids — while a teacher with a roomful of kids who started the year seriously below grade level or otherwise challenged looked bad with all low scores. Obviously that wasn't fair.

In the briefings to the Dallas board, Miles and his staff have explained new measures designed to take into account where each kid starts the year in order to see how far the teacher is able to move him. Teachers would be measured against other teachers with the same kind of kids they have.

The other thing the Miles team seems to acknowledge in its briefings to the board is that achievement tests are not a be-all and end-all. They cite repeated findings in mainstream research to show that other measurements are at least as important. Especially useful are frequent classroom observations by principals and other supervisors or teaching coaches.

A lot of this grows out of research by Harvard Graduate School of Education Professor Richard Elmore and others. They have found too much emphasis has been placed in the past on who teachers are — their character and personality traits — and not enough on what they do. The biggest bang for the buck in terms of student achievement, they have found, comes not from continually trying to find and recruit the perfect teacher but taking the teachers you've got and teaching them to teach.

The most effective place to teach teaching, researchers have found, is in the classroom with frequent observations and later coaching sessions away from students' eyes and ears. Those visits also provide principals with a window on which teachers are getting it and which ones simply won't or cannot adapt their methods. So another major element of the merit metric for teachers is what the district calls "performance," meaning how well a teacher learns and uses the lessons provided by a coach or supervisor.

If that seems to you like a pretty subjective call, the Miles team agrees, and therefore these observation-based performance scores for teachers are shared out and weighted among different kinds of evaluators. One would be a teacher's principal, of course, but others would be drop-ins who had no personal familiarity with a teacher before visiting the classroom.

The third element in the merit algorithm under consideration at DISD is the one that really caught me by surprise. It turns out that one pretty good way of identifying good teachers is by asking the kids. Who knew? Student surveys, written and carried out the right way, have been identified in research as quite valuable indicators of teacher quality.

Obviously you don't ask them, "Does she give you enough recess?" You have to sneak up on them a little better than that. But there are ways to get the kids to tell you what goes on.

All of these factors — student achievement on tests, performance evaluated by multiple diverse observers and student surveys — will be considered together and counterbalanced according to some weighting formula not yet determined. But that issue, the weighting, is where the most important element comes into the picture — the teachers themselves.