Making Evaluations Work for Students and Teachers

School districts across the country are considering important changes to how teachers are evaluated, a change that is going to lead to better teaching and learning. That's critical. U.S. students are doing mediocre at best on international tests. We can and must do better.

There is not a single school factor that has more of an effect on student learning than teacher quality. It's more important than shrinking class sizes or building state-of-the-art science labs. Sure, those are nice. But having a highly effective teacher is essential.

So how do we get there? When I was chancellor of District of Columbia Public Schools from 2007 until late last year, we put a robust teacher evaluation system in place. However, before we got started, we came up with clear expectations. For example, teachers had to show they could deliver content in an understandable way and differentiate their instruction to reach a diverse group of kids.

Once we raised awareness around what good teaching looked like, we tackled evaluations. It was a complete turnaround for the district. Like most, it had a weak system in which teachers were reviewed inconsistently and infrequently, leaving them without the feedback all professionals want and need.

A recent Aspen Institute report stated that the problem was crystallized when you looked at student achievement data showing less than half of students were proficient on district reading and math tests, while 95 percent of teachers either met or exceeded expectations. It doesn't take a numbers whiz to realize those figures don't add up. Teachers were getting passing grades even as their students were failing.

By the end of the first year using the new evaluation system, about 16 percent of teachers got the very highest rating, meaning they exceeded expectations. That was a drop from about 45 percent the year before, under the old system. Those top performers can now earn up to $25,000 in bonuses, and they can get raises if they receive the highest rating again in a consecutive year. At the other end of the spectrum, the new system makes it easier to remove the most ineffective teachers, something that was extremely difficult to do before.

For educators who teach subjects and grades tested under federal law, half of a teacher's evaluation is based on student achievement data. Critics say we shouldn't judge teachers based on how their students do on tests, but that doesn't make sense. It would be irresponsible to ignore student growth when we have the ability to measure it. Good teachers know this. They don't want to ignore the evidence either.

The rest of a teacher's evaluation is largely based on classroom observations. Teachers are evaluated five times a year by their peers and principals. A lot of input went into creating the new system. We held more than 150 focus group and feedback sessions and met with more than 1,500 teachers, principals and other school employees. We listened closely to what they had to say. To paraphrase a line from President Obama, we didn't do it to them. We did it with them.

The Aspen report described DCPS as a "frontrunner in redefining teaching standards and evaluation, long the Achilles' heel of public education." It also said other districts should learn from our experience and improve on it. I welcome that and hope very much that they succeed. Kids and teachers are depending on it.