The Myth of the Evil Teacher Union, Part V

5. Teachers unions prevent effective merit pay systems from being put in place.

This one is half true, which makes it completely false. Teachers unions have actively worked to prevent merit pay systems from being put in place. That’s because merit pay systems, at least all the ones I’ve ever read about, will not be effective at improving teacher performance or student performance. Worse than that, I don’t honestly believe they are designed to do so.



As pointed out above, merit pay systems complicate the dynamic between teachers and students. Unless students have a reason to try hard on standardized tests, merit pay sytems offer them a vehicle to punish their teachers without consequences. We saw this dynamic play itself out when there was no perverse incentive for the students at all; for years our state tests were used only to measure teachers and schools, but had no bearing on a student's grades, gradation, or acceptance to colleges. Consequently, students just didn't care. This year has been the first year when passing the tests has become a graduation requirement. Suddenly (suprise!) student effort on the tests has increased dramatically. To tie teacher pay to student performance might undermine this effort in a way: Students who have "met" could graduate without trying to "exceed", thereby hurting specific teachers they don't like. Also, students who do like their teachers and want to please them would face even more guilt if they failed. I prefer to have my students doing their best for their own benefit, not trying to hurt me or feeling bad if they take money out of my pocket.

[Lest I forget to mention it, let’s not forget what was pointed out in the book Freakonomics: Merit pay systems encourage teachers to cheat.]

There are different models for merit pay, but most rely on test scores. Ignoring the fact that No Child Left Behind leaves testing up to states, which has encouraged states to create easier and easier tests each year in order to show fictional improvement, the tests are still shoddy measures of teacher performance. In most states (Oregon included) students are not tested against their own previous performance, but at specific grade levels, with their scores compared to previous students taking similar tests, or against other students in the state taking the same test. In other words, as we norm the test we hold the students up against other kids. That’s not a means to test a student. It’s a means to test a school. Even then, it’s a poor measure because the school has no control over who comes through the door. We are not factories trying to make machines out of identical widgets. Regardless of the condition our “widgets” come in, we are not allowed to turn them away. Nor are our students like customers, who can be enticed by better products or lower prices. The idea that schools should be run like businesses doesn’t apply, which is why charter schools keep under-performing on state tests when compared to public schools. They are private businesses, which inclines them to want to import business models onto education. And that causes them to fail.

Despite this knowledge or the limits of their utility (and tests are NOT all bad. Despite what some critics might tell you, they do have a place) merit pay systems use test scores, generally as the be all and end all. They either grade teachers based on the improvement of different students within the district, year after year, or based on over-all passing rates in the state. From there, they do one of two things. Either they identify the “best” teachers and pay them more, or identify the “best” schools and pay all their teachers a bonus.

I’m not sure which model is worse. The first, in my opinion, is a blatant union busting move. To use a sports analogy, it’s locker-room poison. If teachers were rewarded based on the over-all passing rates of their students, some teachers would get a bonus every year, while the teachers (like those who teach our special ed populations) would never get one. How long before that would cause dissension within the ranks? On the other hand, if we measured by student improvement, high performers would be missed. The tests only identify students as meeting standards or exceeding them. I’d love to take all the credit for my kids’ successes, but in my honors classes the kids have been exceeding every year since they were in elementary school. There is no room for improvement, according to the state tests. So I would never get a bonus check based on my honors classes, while those working with what we call the “bubble kids”, those right on the cusp, would see the most dramatic gains every year. And again, those working with our developmentally disabled children would be penalized every year. I also work with second language learners. They often show the most dramatic improvement, but because they are so many years behind their peers in English language acquisition, they almost never meet the state standards. How would we account for that?

I can’t see a way of making an individual merit pay system fair without completely un-doing the system of specialization we have in place, which is one of the things our schools do really well. We should be reinforcing the fact that some teachers have special training and ability when it comes to teaching special populations, like special ed. students, second language learners, and gifted students. To give every teacher an equal chance at a yearly merit bonus, we’d have to give every teacher and equal distribution of students. That would rob the kids of the teachers best bale to serve their specialized needs.

The model of rewarding entire schools rather than individual students seems to be an appealing work-around for this problem. Its advantage is that it encourages teachers to collaborate to bring test scores up school wide. That’s good. But it can’t account for the fact that schools serve different populations, so schools that are behind will have less incentive with which to acquire and retain good teachers, thus falling further behind. In that system whole schools would face the challenge faced by individual teachers in the first model: Schools with disproportionately advantaged students (i.e. schools in wealthy areas) would either succeed every year in meeting benchmarks, or fail every year to show dramatic growth. Schools serving disproportionately challenged populations (high poverty, high non-English speaking populations, etc.) would also be winners or losers depending on whether we measure meeting state standards or measure improvement. In both cases, the success would be largely out of teacher’s control, thus eliminating the chief aim of merit pay, which is to motivate teachers.

At its heart, the problem of merit pay systems is that they are based on an incorrect assumption; that students are failing in school because their teachers are unmotivated. If teachers were primarily motivated by pay …THEY WOULDN’T BE TEACHERS! We don’t do this to get rich. We would like fair pay. I think that any merit pay system that doesn’t recognize this will read like a slap in the face, because that’s exactly what it is. If a pay scheme were to be devised which attempted to deal with all the variables mentioned above, it should still be preceded by a general pay increase. That would show teachers that the scheme isn’t based, first and foremost, on an insulting presupposition, but really is a means to reward the best performers among a group of already respected professionals.

Tomorrow, Myth #6: The Teachers Unions are in bed with the Democratic Party.