Schoolhouse Rock

Cleaning House

A reader who spent the last five years teaching in Los Angeles and Chicago writes,  

You’re right that the next fight in education is all about the teachers. We need to pay them better. But there are so many ineffectual teachers out there that it is difficult if not impossible to weed them out.

What you have yet to mention is the difficulty in weeding out those teachers. You have talked about rewarding teachers who are “successful,” but on what grounds will that success be measured? Testing can only measure so much, and some subjects (the phys ed teacher for one) aren’t tested in any meaningful way anyway. If testing cannot or does not measure the success of a teacher, then that puts power into the hands of the principals, which can be incredibly dangerous.

I agree, it’s a difficult question. But some useful answers are emerging. The best research I’ve read about teacher quality is a paper from 2006 by Robert Gordon, Thomas Kane, and Douglas Staiger.  

They make five points:

1) We don’t know yet exactly how best to measure teachers, and the federal government should give R&D money to the states to test out some different methods.

2) That said, it seems that some measure that combines growth in test scores—how much a teacher raises his students’ scores over their scores the previous year—with more subjective measures like principal evaluations will actually give us a pretty accurate picture.  

3) How can we tell it’s accurate? Because teachers who do well in one year by these measures tend to do well the next year, too. In other words, in any given school, there are teachers who consistently do well and teachers who consistently do poorly, and it’s not all that hard to figure out who’s who. ( Though it is hard to predict ahead of time, at hiring .)

4) We should offer bonuses to teachers who perform well by those measures—in the top quartile, say—though only if they’re willing to teach in schools with a lot of low-income students, where they’re needed most.  

5) Even though bonuses are nice, the more effective way to use the evaluations is to fire the bottom quarter of teachers. This is less politically palatable. But these teachers are the ones, Gordon and his co-authors say, who do real damage to the students they teach, and they tend to do that consistently.

The authors have a neat firing idea—they make it the default. So a principal simply couldn’t continue to employ a second-year teacher who performs in the bottom quartile without sending a note home to parents explaining why this low-performing teacher deserves to keep his job.   

It’s not a solution we’re likely to hear on the campaign trail any time soon. Pro-layoff politicians don’t last long. But, politics aside, what these authors are proposing is a valuable and natural flipside to a bonus system. What good is paying the high performers more if the low performers are still hanging around, mishandling another classroom of kids every year?