Cincinnati has one of the most rigorous teacher evaluation systems in the country, entailing four largely unannounced classroom observances by trained evaluators that occur throughout an academic year. What have you learned in your study of that system?

We found two things that are particularly relevant as states and districts across the nation are grappling with how to implement their own teacher evaluation systems. First, we found that, indeed, what teachers are doing in the classroom as scored by evaluators is highly related to a teacher’s ability to promote student achievement growth as measured by test scores. This is powerful evidence that good classroom observation systems can discern, at relatively fine degrees, classroom practices that contribute to student learning.

Then we asked a second question: Does the very act of going through this year-long evaluation system in Cincinnati make you a better teacher? While we continue to explore the research and statistical details, our results suggest that the answer is yes, and the effect sizes at this stage of our research are large enough to be quite consequential. Furthermore, we found that not only does a teacher’s effectiveness increase in the year in which they are undergoing evaluation, but the effects of going through the evaluation cycle are even larger in the years after the evaluation.

We might expect teachers to be better in the year in which they are undergoing this process, if for no other reason than they bring their “A-game,” so to speak. But seeing these effects in the postevaluation years suggests that there is an important human capital component to going through a rigorous and robust evaluation system.

Why does this help teachers so much?

We don’t know the mechanisms — it could be the face-to-face and/or written feedback they receive from the evaluators, it could be the detailed language about what is, for example, “distinguished” versus “basic” teaching practice (terms the system uses to describe different levels of teaching). It could be that bringing to the table the specific classroom practices on which teachers are scored in TES changes the discourse around teaching among teachers and administrators in the district.

While we can’t identify the exact mechanisms, what the research does suggest is an evaluation system like Cincinnati’s can be a good professional development tool. This is important because rigorous research has thus far failed to identify effective professional development activities. Another reason we think the results from this study are important is because current discussions across the nation — and they are many and at high levels — focus on teacher evaluation primarily as ways of either identifying “bad” teachers and denying them tenure or removing them from the classroom, or in the case of merit pay schemes, identifying “good” teachers for pay bonuses. Our work suggests that the right kind of evaluation has the potential to help teachers become more effective and it’s the strongest evidence to date on that front. [The findings are detailed in a working paper with the National Bureau of Economic Research].

Using the same classroom observation data, your research being published this summer in the Journal of Human Resources and Education Next pinpoints a link between teaching practices and student achievement. What exactly did you find?

First, we established probably the strongest evidence to date that the classroom practices many teachers have long thought define “good” teaching do indeed result in greater student achievement growth. That is, a teacher’s overall score from the TES evaluation is highly predictive of that teacher’s ability to promote student achievement growth as measured by test scores. Our results indicate that getting better on any one of the eight practices that make up the overall score would result in a more effective teacher.

Second, we found that for promoting achievement growth in math, a teacher’s ability to develop and promote a positive classroom environment is relatively more important than a teacher’s instructional practices. That is, while improving your teaching “chops” will lead to higher math (and reading) achievement, managing student behavior, the physical classroom space, and creating an environment for learning is even more important when it comes to math achievement growth.

As a former math teacher myself, this was at first a bit troubling to me ... but then I read and agreed with an article by Douglas Lemov [“Building a Better Teacher,” New York Times Magazine, 2010], in which he said if you can’t get a fifth grader to stop what they’re doing and look you in the eye when you’re trying to teach something, it doesn’t matter what your teaching chops are or how well you know your content. At least for students in the third through the eighth grade (the focus of the study). If you can’t establish the right kind of classroom environment, then you have a much slimmer chance of promoting learning in math. My interpretation is that our results provide empirical support for what Lemov had found observationally.

So, having a good classroom environment promotes math scores. What about reading scores?

Again, remembering that getting better at any of the eight classroom practices being scored will improve math and reading, we found that a teacher’s ability to do a very specific thing can promote additional achievement in reading. Our results show that teachers who scored higher when it came to engaging their students with questions that promote discussion and further thinking were more effective at promoting reading achievement growth than teachers who were relatively more effective in their ability to teach standards and focused more on content-based teaching.

Is the Cincinnati system feasible for other school districts across the country?

I guess the answer depends on whether a district is focused on short-term costs or on the bigger picture. The bulk of the cost for a TES-like system is in personnel, as it can cost about $900,000 per year to take 12 or 13 of your best teachers out of the classroom to train and serve as evaluators. While that’s a lot of money, it is roughly about $4,000 per teacher evaluated each year, and for an elementary teacher with 20 students, this is about $200 per student. But our results show that the $4,000 estimated cost is a good investment because not only will that teacher’s students learn more in the year in which their teacher is being evaluated, that teacher’s students in the years after he or she is evaluated will learn more as well. And given the size of the achievement gains that we estimate, the expected lifetime earnings gains associated with students who learn more than they otherwise would have dwarf the immediate costs of TES evaluation.

What can we learn from this?

The design and rollout of teacher evaluation systems are going to be a part of education reform for the foreseeable future, in no small part as a result of Race to the Top-related incentives. Some systems are being developed at the state level, while in many other cases it is districts that are taking on the challenge. Our research suggests that when done right, teacher evaluation can both identify and incentivize classroom practices that promote real student learning and that in doing so, evaluation can also help develop more effective teachers. Those are important findings given the effort and dollars that will be invested in these systems in the coming years, but as we point out in both papers, the key qualifier here is that our findings are in a district that has given serious thought, attention, and resources in teacher evaluation. Absent that, all bets are off.