Teachers are getting more media attention this season but most of it is pretty vitriolic. In my previous post I wrote that if I wasn’t a teacher, and only received my information about public education from the media and Arne Duncan, I’d probably think an epidemic of lazy and incompetent teachers are a main reason many American students are struggling. Many voices would seek to convince me that implementing performance pay, or merit pay, for teachers is essential to whipping our school system into shape.

The implication is that with the chance to earn a bonus for higher student test scores, would-be competent teachers would try harder and incompetent teachers would be identified and weeded out. That’s cynical and wrong on the first count and betrays a misplaced faith in the accuracy of state test data on the second.

Teacher evaluations do need to be overhauled; I don’t think anyone— unions included— would defend the current ad hoc system. However, business models imported from the private sector emphasizing year-to-year high-stakes test score changes just aren’t a constructive way to do it. Paying (or withholding payment) for year-to-year test scores is an unhealthy, priority-distorting enterprise.

Many have an entrenched interest (Michelle Rhee, Joel Klein, Presidents Obama and Bush) in pressing a reductive high-stakes testing regime that spits out easy-to-understand data. They want to trumpet gains on the graph this news cycle, this election cycle. They have the biggest microphones on their side. (See Waiting for Superman, No Child Left Behind, and the Obama Administation’s Race to the Top criteria as examples where “state test scores” and “student achievement” are conflated by design.)

Who can speak out— to a significant audience— for solutions that work in living classrooms? Teachers are too busy grinding it out in their classrooms to cultivate mainstream media clout. And it’s not like the media has displayed great interest in amplifying on-the-ground voices in education.

This is where Diane Ravitch comes in. Yesterday, the widely read Daily Beast put her essay titled “Stop Trashing Teachers!” as the top story on their top ten “Cheat Sheet: Must Reads From All Over” list.

 I’ve excerpted below some of Ravitch’s comments on merit pay. How I wish that every stakeholder in education would read this and nod in understanding. Life in a classroom is a messy, human endeavor that requires multiple forms of personal, on-the-ground observation and assessment. Merit pay is no silver bullet; on the contrary it may be a self-inflicted wound to our long term goal of a better-educated citizenry. I know if my ability to provide for my family was determined by my students’ test scores, my curriculum would be drastically altered (and not for the richer) and my stress would be through the ceiling.

Here is Diane Ravitch:

 …The claim that merit pay will improve student performance has been disproven again and again. Whenever businessmen decide to "reform" education, they insist on merit pay. But it doesn't work. The latest study, released only a week ago by the National Center on Performance Incentives, was the most rigorous evaluation of merit pay ever conducted. One group of teachers in Nashville was offered bonuses up to $15,000 if they raised students' math scores; another, the control group, was offered nothing. The average teacher pay is about $50,000, so this was a significant incentive to get higher scores. Over the three years of the study, both groups produced the same results. The economists, who were scrupulously nonpartisan, concluded that performance pay had no effect on student performance. It turns out that teachers were working as hard as they knew how, with or without the bonus.

The claim that teachers can be accurately evaluated by student test scores has been refuted again and again by scholars. The Economic Policy Institute released a statement by many of the nation's leading testing experts warning that this method was riddled with error and instability. A study released days ago by Sean Corcoran of New York University showed that a teacher who was ranked at the 43rd percentile, using student test scores, might actually be at the 15th percentile or the 71st percentile because the margin of error in this methodology is so large.

Tests that assess what students have learned are not intended to be, nor are they, measures of teacher quality. It is easier for teachers to get higher test scores if they teach advantaged students. If they teach children who are poor or children who are English language learners, or homeless children, or children with disabilities, they will not get big score gains. So, the result of this approach—judging teachers by the score gains of their students—will incentivize teachers to avoid students with the greatest needs. This is just plain stupid as a matter of policy.