Education research has found teacher quality to be the single biggest determinant of student success. However we also know that teacher quality is not predictable based on any observables - degrees, quality of schools, etc. Which is why there is such an emphasis on ex-post teacher evaluation and performance based incentives. It has to be after a teacher is on the job that their performance is evaluated and then there must be a mechanism to promote successful ones. This is all quite logical.
But before we rush headlong on incentive schemes and hiring/firing policies we must confront two inconvenient truths in education:
1. Teacher evaluation is very hard and the great hope of using value-added measures is not working very well in practice. The measure are unstable across years, tests and even, I am told from someone with intimate knowledge of the Chilean experiment, across sections of the same test.
2. Incentive schemes, at least in the US are not working.
Just something to think about apropos the new PPS teacher contract set to be ratified by the board tonight.