Last month, in a conversation, Amanda Ripley noted that test score based teacher evaluations might not be “worth it.”

Amanda’s argument was that the loss of trust and morale associated with these evaluations might cause them to have an overall negative impact on public education (apologies to Amanda if I’m getting her argument wrong, is a tough medium for policy nuance …. anyway, others have made similar arguments).

I understand this line of thinking. When I led NSNO, I sometimes put the brakes on internal initiatives where the risk of negatively affecting culture outweighed any potential gains in efficiency.

So what to make of government mandated, test score based, teacher evaluations?

Here’s how I approach the issue: 

1. I think it’s a poor idea to pass a law or policy that provides a one size fits all type of solution for something as complicated as employee evaluations.

2. If, as I prefer, non-profits operated all the schools in this country, the issue would be a non-starter for me. I would oppose all such top down evaluation efforts.

3. The issue is complicated by the fact that 95% of public school students attend government operated schools.

4. The issue is further complicated by the fact that, as the Widget Effect demonstrated, many of these school systems do a weak job of evaluating employees.

5. The issue is even further complicated by the fact that, for the foreseeable future, the vast majority of public school students will attend government operated schools.

6. And to top it all off: the statistical measures used, while likely an improvement on existing systems, lack high degrees of precision (this is the other major critique of these systems, for a balanced perspective, see DiCarlo’s take here). 

I am very torn on the issue.

My ideal solution to the issue of poor management of teachers is to transition our public education system to a non-profit operated system of schools.

Under this system, as with most sectors of our economy, different organizations could develop different evaluation systems based on their own values and strategy. And instead of legislators having to dream up the perfect evaluation system, schools would do this work and then be held accountable for delivering results.

But the transition to non-profit schooling is not occurring anytime soon.

As such, I’m left trying to figure out some utilitarian calculation of whether the pros of top down teacher evaluation regimes outweigh the cons.

This type of technocratic endeavor often ends poorly, as most complex systems make such analysis futile.

But here goes:

1. I think student achievement based evaluation systems probably work pretty well at identifying the worst teachers (even if there is much more noise closer to the median).

2. I’m open to Eric Hanushek’s idea that we could significantly increase student achievement if we moved the bottom 5% of teachers out of the workforce and we could replace them with more effective teachers (both by increasing retention of higher-performers, as well as increasing the size of teacher pipelines that are turning out teachers that do well on student achievement measures).

3. I am mildly skeptical that school districts will actually fire the bottom 5% of their teachers. Chad Alderman and Carolyn Chuong of Bellwether Consulting just came out with a report that noted that “the ‘widget effect’ continues in all but a few places, and the ability of teachers and principals to improve student outcomes has little bearing on their evaluation outcomes.” 

4. I do not know whether Amanda is right that these types of reforms reduce teacher morale. My guess is that individual managers have more of an affect on morale than evaluation policies, but I’m not sure.

5. I think (but admittedly have no hard data to support the idea) that top down teacher evaluation types of reforms are actually positive for the charter movement. From the progressive perspective, charters become a way to free yourself from corporate technocrats. From the conservative perspective, charters become a way to free yourself from government intrusion.

So, as much as it pains me to say it, I guess I’m (somewhat) in support for top down, test based teacher evaluations.

My preferred solution isn’t happening any time soon.

Implementing these evaluations might actually speed up the adoption of my preferred solution.

And, all things considered, I think the benefit of (potentially) moving out the bottom 5% of teachers outweighs the (potential) cost of reducing morale.