Teacher Quality: Compensation Or Self-Selection?

Now here’s something that ought to be used as input into the National Education Blueprint (abstract):

Teacher Quality Policy When Supply Matters
Jesse Rothstein

Recent proposals would strengthen the dependence of teacher pay and retention on performance, in order to attract those who will be effective teachers and repel those who will not. I model the teacher labor market, incorporating dynamic self-selection, noisy performance measurement, and Bayesian learning. Simulations indicate that labor market interactions are important to the evaluation of alternative teacher contracts. Typical bonus policies have very small effects on selection. Firing policies can have larger effects, if accompanied by substantial salary increases. However, misalignment between productivity and measured performance nearly eliminates the benefits while preserving most of the costs.

Deciphering the Econo-speak results into Econo-English:

1. For a variety of reasons, performance-based teacher compensation policies have very little impact on education outcomes (good teachers are born, not made);
2. The ability to fire ineffective teachers is actually more important, but requires a considerable jump in pay to compensate for the added risk of job insecurity;
3. The caveat for the above is that if performance measurements can be manipulated (e.g. exam-oriented teaching, instead of knowledge-oriented teaching), almost all the increase in education outcomes is lost.

The whole point of the paper is to examine a teacher hiring, retention and compensation framework that while performance-based is also geared towards self-selection i.e. we want potentially good teachers to come forward, but discourage bad teachers from ever entering the classroom.

That’s a worthwhile goal and a pretty novel appraoch to the question, but from the look of things there’s very little in the way of a real consensus on the issue of improving teacher quality. Part of the problem is simply measurement – exam results are at best a “noisy” signal of how good a given teacher actually is, and subjective evaluations are subject to observer bias.

Another problem is the time span – in the 11-15 years of formal education, what is the individual impact of any given teacher in a child’s education outcome, especially if there are interim changes to curriculum or school policies?

Tough questions, very few answers.

Technical Notes

Rothstein, Jesse, "Teacher Quality Policy When Supply Matters", NBER Working Paper No. 18419, September 2012