The Corner

Re: Evaluating Teachers

Jon Chait has reacted to my post on teacher evaluations, but I don’t think that he was responding to my actual claims. He seems to think that I was arguing that quantitative measurements of teacher performance (“value-added” metrics) are not achievable, and that therefore we should not use them. My actual argument was that we need such systems, but that we should be realistic about what is required to make them work.

Chait quotes a long illustrative dialogue that I used to show some of the problems that often arise from trying to use a complex regression model to measure employee performance in a corporate setting. But the sentences in my post that immediately follow the quoted dialogue are:

Not all attempts to incorporate rigorous measures of value-added fail. Let me make some observations about when and how workable systems that do this tend to be designed and implemented.

And later in the post, I also say that:

More serious measurement of teacher performance, very likely including relative improvement on standardized tests, will almost certainly be part of what an improved school system would look like.

My post wasn’t about if we should use quantitative measures of improvement in their students’ standardized test scores as an element of how we evaluate, compensate, manage and retain teachers, but rather about how to do this.

Two of the key points that I tried to make are that the metrics themselves should likely be much simpler than those currently developed by economics Ph.D.s, and that such an evaluation system is only likely to work if embedded within a program of management reform for schools and school systems. The bulk of the post was trying to explain why I believe these assertions to be true. 

An additional point that I mentioned in passing is my skepticism that such management reform will really happen in the absence of market pressures on schools. Continuous management reform, sustained over decades, that gets organizations to take difficult and unpleasant actions with employees is very hard to achieve without them. There’s nothing magic about teachers or schools. The same problems with evaluation and other management issues that plague them arise in big companies all the time. It’s only the ugly reality of market discipline that keeps them in check.

Jim Manzi is CEO of Applied Predictive Technologies (APT), an applied artificial intelligence software company.


The Latest