The Times’ Room for Debate blog tackles teacher evaluations today, in particular the news that New York City plans to introduce a dozen new tests in order to gather data for said evaluations. Participants include Linda Darling-Hammond, Kevin Carey, Marcus Winters, and yours truly, among others. Here’s my submission; read the whole package here.
Improving teacher evaluations is one of the most important reforms encouraged by the federal “Race to the Top” initiative — and one of the central components to making our schools better. No one can defend today’s evaluation systems which, by and large, find every teacher to be above average (if not superior) even as our student achievement results lag our international competitors.
If pay and employment decisions are to be based on teacher performance, at least in part, we need evaluations that can stand up to scrutiny (and to lawsuits). Simply put, we won’t make much progress in terminating our least effective teachers (either for cause or because of budget pressures) until we have evaluation systems that are fair, trustworthy and rigorous. And it’s only common sense that one element of those evaluations should be an assessment of how much students are learning under the teacher’s charge.
However, there’s a real downside in moving to centralized, rules-based, bureaucratic evaluation models, as indicated by New York City’s decision to add a dozen new tests to collect more teacher performance data. I feel for the city; if you want evaluations to be grounded in data, then it’s not crazy to assess students in subjects for which children are not currently tested by the state. But talk about attacking a fly with a sledgehammer. There’s already a ton of testing in our schools. Isn’t there another alternative?
There is an option that neither reformers nor the unions want to consider: trust the principal. In most of American life, individuals are evaluated by their managers, who have a lot of discretion over their employment, their salaries, and any bonuses they might receive. In the best organizations, those managers collect plenty of data before making their decisions — peer reviews, outcome data, etc.— but none of this is meant to substitute for human judgment. It’s not a perfect system, and without safeguards it can be open to abuse, but if you believe in matching authority with accountability, it’s the least worst alternative.
If you trust the principal, then there’s no need for new citywide tests. Good administrators that want to evaluate their social studies teachers, for example, might spend more time in their classrooms. They might look at the quality of student work, or get feedback from peers and parents. Maybe they’d want an objective assessment of student growth in the subject over the course of the year; give them the option. But don’t make it mandatory.
And if administrators actually have the authority to link their evaluation decisions to something meaningful — firing bad teachers, bumping the salaries of their superstars — they will have reason to take the evaluation process more seriously.
Reformers who are pushing for statewide or even district-wide evaluation systems are saying out loud: we can’t trust principals to make these decisions on their own. And they are creating pressure for districts like New York’s to spend countless hours and dollars trying to gather data. If we can’t trust school leaders to identify their best and worst teachers, then the whole project of school reform is sunk. Not all the additional tests or teacher evaluations in the world can change that.