I'd quarrel with the tone of Eric Hanushek's criticism of efforts to save teachers' jobs ("Cry Wolf!") but not the substance of his recommendations. We need teacher evaluations based on "direct evidence about teacher effectiveness," so districts could fire the weakest instructors. That should rule out test score growth for evaluations (at least in the hands of management) because the results of primitive growth models say little about whether an individual teacher is effective. Either peer review, or "the Grand Bargain" of submitting test score data to a team of teachers and administrators, would meet the goal of removing the bottom 8 to 10%. Typically, it is management that balks at these common sense methods of
