High teacher attrition is painted as a very bad thing, causing disruption to schools and siphoning dollars away from student needs toward recruiting and training replacements. Yet emerging research suggests there may be a gold lining: recent studies from Eric Hanushek and Dan Goldhaber have found that new teachers who leave are likely to be less effective than the teachers who stay. A new NBER working paper adds to this body, specifically looking at attrition in New York City.

Who Leaves? Teacher Attrition and Student Achievement uses value added student achievement data to look at the effectiveness of new teachers who leave and those who stay in both high performing and low performing schools. The study also looks at where the leavers go: to another school within NYC, another district within NY state, or leave the system all together (which could mean they teach in another state or that they are no longer teaching at all).

While the study finds that less effective first-year teachers are more likely to leave--especially if they are working in a low performing school--a significant number of these leavers simply transfer to another school in the district, where they generally aren't any more effective than in their first school.

Reinforcing previous research, all leavers--regardless of effectiveness--tend to move to higher performing schools with lower poverty and minority populations.

The study concludes that broad teacher retention policies are misguided; they may work to keep teachers who are harmful to student achievement. Retention strategies should discriminate among teachers based on their effectiveness, with particular attention on the impact of district transfer policies.