Which leaves school officials in the position of having to find a way to get rid of the inevitable bad hires. Anthony Lombardi's approach at PS 49 put him at the top of the teachers-union hit list. (The union head refers to Lombardi as a "tyrant.") Lombardi placed higher demands on his teachers, requiring, for example, detailed and cogent lesson plans. (He recalls that some teachers had one-word class outlines before the new rules were put in place.) He also started showing up in class to keep tabs on what was going on. While he may not have been able to discern teaching quality from a résumé, he knew effective teaching when he saw it in the classroom. Teachers who either couldn't or wouldn't perform up to his standards were given an ultimatum: Request a transfer or get saddled with an unsatisfactory rating, leading to an onerous (for all concerned) two-year review. Since his arrival, a third of PS 49's teachers have been squeezed out through Lombardi's efforts.
Of course, this just meant they were moved to another classroom in another school, lowering the test scores of someone else's children. So while this might be a way of cleaning up PS 49, it's not much use in reforming an entire school system. New York's school chancellor, Joel Klein, has gotten rid of some teachers through a program that effectively gives them a golden parachute out of teaching—they aren't allowed into the classroom, though they stay on the payroll. But this is a very expensive Band-Aid.
What if there were a way to screen out the bad teachers before they get entrenched? Currently, New York City teachers get their union cards their first day on the job. In theory they're on probation for three years after that, but in practice very few are forced out. Lombardi suggests replacing this system with an apprenticeship program. Rather than requiring teaching degrees (which don't seem to improve value-added all that much), new recruits would have a couple of years of in-school training. There would then come a day of reckoning, when teachers-to-be would face a serious evaluation before securing union membership and a job for life.
Lombardi's proposal isn't without its problems and complications: What would the effect be on the morale of older teachers? Would the teachers unions ever agree to such a system? But none seems insurmountable. Researchers Kane and Staiger, together with coauthor Robert Gordon, have also suggested an apprenticelike system and have put forth a detailed proposal on how to implement it.
We live in an age of increasing inequality. While it's not fair to park the problem of global inequities at the doorstep of teachers unions, the continued floundering of public education in America is at least partly to blame: Education is an awfully good predictor of future earnings, and keeping bad teachers in classrooms filled with kids from poor families certainly helps to reinforce the cycle of poverty. The difference between a teacher in the 25th percentile (a very good teacher) and one at the 75th percentile (a not very good teacher) translates into a 10 percentile point difference in their students' test scores. (As a frame of reference, on the SAT, 10 percentile points translates into an 80 or so point difference in raw test score.) After a string of good teachers or bad teachers, it's easy to see how you can end up with very wide gaps in student achievement. And this is all the more tragic since at least part of the answer—doing a better job of evaluating and selecting teachers—is readily at hand.