From Jesse Rothstein:
Recent proposals would strengthen the dependence of teacher pay and retention on demonstrated performance. One intended effect is to attract those who will be effective teachers and repel those who will not. I model the teacher labor market, incorporating ability heterogeneity, dynamic self-selection, noisy performance measurement, and Bayesian learning. Simulations with plausible parameter values indicate that labor market interactions are important to the evaluation of alternative teacher contracts. Reasonable bonus policies create only modest incentives and thus have very small effects on selection. Tenure and firing policies can have larger effects, but must be accompanied by substantial salary increases. Both bonus and tenure policies pass cost benefit tests, though the magnitudes of the benefits are quite sensitive to parameters about which little is known.