Targeting Teachers

The mantra of the current school reform movement in the United States is that high- quality teachers produce high-achieving students. As a result, we should hold teachers accountable for student outcomes, offering bonus pay to the most effective teachers and shoving the least effective ones out the door. Of course, in order to implement such a policy, you need a valid and reliable measure of teacher quality, and the reformers have zeroed in on one such measure, which is known as the value-added approach. According to this method, you calculate the effectiveness of individual teachers by the increase in test scores that students demonstrate after a year in their classroom.

Propelling this trend is a flood of research purporting to show that differences in teacher quality can lead to huge differences in the outcomes of schooling, both for students and for society. For example, in a 2010 study for the National Bureau of Economic Research, Eric Hanushek argues that a teacher judged to be strong by the value-added measure (one standard deviation above the mean) might raise the lifetime earnings of a student by $20,000. From this perspective, improving the quality of teaching promises to increase individual opportunity for the disadvantaged—which will reduce social inequality—and at the same time to increase human capital, which will promote economic growth and national competitiveness. Sounds great. Of course, this calculation is based on the assumption that test scores measure the economically useful knowledge of the future worker, which is far from obvious. But arguments like these provide a big incentive to generate usable data on who’s a good teacher and who’s not.

All of this makes the current effort to develop a simple and statistically sound measure for good teaching quite understandable. But it doesn’t make the effort justifiable. The problem with this approach is that teaching is an extraordinarily complex and demanding form of professional practice whose quality is impossible to capture accurately in a simple metric. The push to develop such a metric threatens to reduce good teaching—and good education—to whatever produces higher scores on a standardized test. As a result, the value-added measure of teacher quality may end up promoting both the wrong kind of teaching and the wrong kind of schooling.

In this article, I explore three major questions that arise from this development. Why did the value-added measure of teaching emerge at this point in the history of American education? What are the core characteristics of teaching as a professional practice that make it so hard to perform effectively and so hard to measure accurately? And under these circumstances, what are the likely consequences of using the value-added measure of teaching?

Roots of the Value-Added Measure of Teacher Quality

Until the last thirty years, Americans had been comfortable measuring the effectiveness of the...