Seniority-based layoffs losing ground--again

See all posts

There's new evidence of the negative impact on students from seniority based layoffs.

Economist Dan Goldhaber with fellow researcher Roddy Theobald conducted a unique and interesting simulation, comparing the different learning outcomes that occurred after a series of real layoffs across school districts in Washington State with a hypothetical group of teachers who would have been laid off if effectiveness, rather than seniority, were the primary consideration.

The traditional seniority system, at least in the state of Washington, seems particularly immune to the type of manipulation that smart principals routinely seem capable of pulling off in other personnel decisions, such as transfers. As evidence, there was little overlap in the teachers who were laid off in reality and those laid off in the simulation: only 16 percent of the teachers who actually got laid off were "laid off" in the simulation. Either principals have no tricks up their sleeves to protect particularly effective teachers from a layoff--or it might be that their idea of who is really ineffective doesn't match with what the test scores say.

If the most ineffective teachers--as judged by test scores--had been let go, Washington school districts could have avoided what Goldhaber estimates to be a roughly 3-month loss in student learning in classrooms staffed by weak teachers who managed to avoid the actually layoff under the seniority system. Districts in the state could also have made up their deficits by laying off 10 percent fewer teachers in the state, given that younger teachers cost relatively less money than older teachers, so more have to be laid off to erase any deficit.

In any case, the learning loss alone makes a persuasive argument for moving to an effectiveness-based layoff system, but with a few caveat emptors. The ability to identify ahead of time who the worst teachers are going to be in any given school year is still rife with errors--though retrospectively, it is clearly a breeze. Districts may make some of the right choices, but are unlikely ever to reach a level of accuracy in layoffs that would allow them to reliably produce three months of increased learning.

Simulation aside, we actually found some parts of Goldhaber's analysis of the actual layoffs pretty interesting, such as the finding that after controlling for experience, teachers who hold master's degrees are much more likely to hold onto their jobs, for reasons that aren't clear. While the best job to hold when lay-offs threaten appears to be in special education, the worst are in physical education and health. Of course, no teachers, no matter credentials, are more likely to get cut than 1st-year teachers.

See NCTQ's analysis of how states are handling this issue.