mead cohen berger shevtsova garfinkle michta grygiel blankenhorn
Education Gets Smart
Surprise: Firing Bad Teachers Improves Student Performance, Study Says

Much of our education policy is built around the assumption that teacher turnover is generally undesirable—that training new teachers is too costly, that tough accountability standards deter good people from entering the profession, and that it’s best if teaching is a lifelong career protected and underwritten by unions. And, indeed, there is some evidence that, all things being equal, schools with high rates of teacher turnover do worse than those in which teachers stay for longer periods of time.

But a new NBER working paper from economists at Stanford and the University of Virginia suggests that, when done right, one kind of teacher turnover, at least, can be highly effective: programs for aggressively replacing bad teachers. The authors collected data from a unique Washington, D.C. program called IMPACT, which assesses teachers based on student outcomes and ratings from their peers, rewards those who perform well, and replaces those who persistently perform poorly. In a nutshell, it worked: The teachers pushed out for poor performance were consistently replaced with teachers who performed significantly better. “Under a robust system of performance assessment,” the authors write in their conclusion, “the turnover of teachers can generate meaningful gains in student outcomes, particularly for the most disadvantaged students.”

As we’ve written before, the idea that all teachers must be teachers for life needs to be questioned more often. That’s especially true when one is talking about replacing poorly performing teachers. Despite the protestations of unions, the evidence has been mounting that identifying teachers with passion and skill is eminently possible, and that pumping new blood into the system can be good for students. This doesn’t mean policymakers should construct arcane accountability metrics for the sake of looking like they are doing something. But it does mean that we need more experimentation in teacher assessment (and in our education system more broadly)—and that programs that work, like IMPACT, should be adapted and built upon elsewhere.

Features Icon
Features
show comments
  • FriendlyGoat

    “This doesn’t mean policymakers should construct arcane accountability metrics for the sake of looking like they are doing something.”

    We’re already pretty far down that road in a number of places, but the “opt-out” movement on standardized testing may end up putting in the monkey wrench. From what I understand, you’ve got some parents who are generally on the political side which wishes to bash teachers with test-based “evaluations” —–and yet—-they want to opt their kids out of the testing for all kinds of other reasons. Oops.

  • Andrew Allison

    Shouldn’t ““Under a robust system of performance assessment,” the authors write in their conclusion, “the turnover of teachers can generate meaningful gains in student outcomes, particularly for the most disadvantaged students.” be self-evident?

  • johngbarker

    More effective teaching and learning would ensue with less waste if there were higher standards for those entering the profession ( on that account teaching might even become real profession instead of a semi-profession). In Finland only ten percent of the applicants to teacher training programs are accepted. I believe but I cannot remember the source of this information. Much better to select at the front end than in the middle.

© The American Interest LLC 2005-2016 About Us Masthead Submissions Advertise Customer Service