Connecticut’s Senate Bill 24 echoes a battle being fought around the country: Should teachers be subject to high stakes evaluation? Some of the issues in contention are matters of opinion and matters of politics. But some of the issues are matters of science. If we can clear up some of the science, maybe we can even bring those who differ on opinion and politics a little closer together.
The science first. Opponents of SB 24 have been arguing that because student background (like poverty, or the involvement of their parents) is so important, the role played by teachers is not so important, or that teachers cannot significantly address outside challenges. This is just wrong. And since the opponents are quoting research by one of us (Goldhaber) on this topic, you can take my word for the fact that they are misinterpreting the import of my findings and the findings of other scientists who have looked at the student background versus teacher input argument. High quality teachers are critical to student achievement. In fact, the evidence suggests that increasing the quality of the teacher workforce is one of the most important social investments we can make in the long-run fight against poverty.
Using the fact that much of what explains achievement differences can be traced to factors outside of schools as an argument against rigorous teacher evaluation, a component of which may include test score growth measures, is bad science. Without getting into the statistical weeds, let us explain by analogy. Class size may be a factor in influencing student achievement, but if there are no differences in the number of students per class across classrooms, then class size does not explain the gaps we see in student achievement – it cannot since class size would be the same for all students. But the fact that class size does not explain achievement does not mean it cannot be manipulated to affect it.
The public policy issue isn’t whether family and community factors influence what students learn (of course they do). The public policy issue is whether differences between teachers produce large differences in what students learn. As every parent who’s worried about which teacher their child’s assigned to knows—and as a large body of statistical research affirms—some teachers are incredibly effective, some are quite ineffective.
The only reason we know that teachers are so very important to student achievement is because they often differ significantly from one another, and we can see empirically that these differences have educationally meaningful impacts on students. Thus, recognizing the importance of teachers goes hand in hand with an acknowledgement that teachers are not interchangeable, identical, widgets. Yet this is exactly how most traditional evaluation systems treat them since there is little to no differentiation in teacher performance ratings. 98 percent of teachers are simply marked “effective.”
This brings us back around to Connecticut’s Senate Bill 24. We might not know perfectly how new performance evaluation systems ought to be structured, but we do know that recognizing the differences between teachers is the first step in crafting human resource policies that upgrade the overall quality of the teacher workforce. We must also not let the perfect be the enemy of the good, and look to what other states and localities have already done on this front as examples of what is possible.
Pretending that all teachers are average is bad science. It’s also bad politics because it divides groups that should be on the same side on this topic even if they aren’t on others. Recognizing superb teachers is a first step toward rewarding them (financially, with career advancement, etc., but not just praise.)
And for a teacher evaluation system to recognize superb teachers (yes, and to identify ineffective teachers as well), the system has to be recognized as sensible and fair. That means multiple components of evaluation. It means recognizing that some groups of students walk into the classroom with disadvantages. We surely don’t want to give teachers poor evaluations for working with students with the greatest needs!
But the bottom line is that some teachers are more effective than others…just like in every other profession. More effective teachers should be identified both so they can be rewarded and so they can serve as role models for others…just like in every other profession. There’s room for disagreement of opinion and politics about how to best evaluate teachers. The science of what’s the best “how” isn’t fully settled. But moving toward a smart and meaningful evaluation system is a move in the right direction.
Dan Goldhaber is the director of the Center for Education Data & Research and a professor in Interdisciplinary Arts and Sciences at the University of Washington Bothell and Richard Startz is Professor of Economics at the University of California, Santa Barbara and the University of Washington.