Survivor: The TFA Edition
skoolboy remains fascinated by the way in which Teach for America, a program serving perhaps 3% of the students in the districts in which it operates, can seem like the tail wagging the dog. Like eduwonkette, I see many virtues to the program, but do not view it as a solution to the nation's challenge of developing a corps of skilled career teachers to serve our children and youth.
TFA recruits make a two-year commitment to teaching in a high-needs school, and the limited nature of this commitment is a recurring source of concern. If TFA recruits stay just two years and then leave, then the schools they serve face a revolving door of teachers shuffling in and out. TFA, for its part, cites recent evidence that TFA recruits are at least as effective in the classroom as other novice teachers. Moreover, TFA champions the enduring value of having its recruits see the challenges facing high-needs schools, if only for a few years, and claims that many recruits stay in the field of education beyond the two-year commitment.
There’s some new evidence on this latter point, emerging in the doctoral dissertation research of Morgaen Donaldson, formerly with Harvard’s Project on the Next Generation of Teachers, and now an Assistant Professor in the School of Education at the University of Connecticut. Donaldson surveyed the 2000, 2001 and 2002 cohorts of TFA recruits, obtaining 2029 responses, for a 62% response rate. Focusing on voluntary departures (approximately 16% in the sample were involuntary), she modeled the likelihood of staying in the initial placement school over time, as well as the likelihood of transferring to another school or leaving teaching altogether.
The charts below are from fitted hazard models that describe the cumulative probability of "survival" in the initial placement school across years, as well as the probability of voluntarily resigning from teaching for the first time. The first chart shows that about 90% of TFA recruits (voluntarily) remain in the initial placement school for a second year, and about 44% stay for a third year. These figures decline steadily over time, with about 22% staying in the initial placement school for a fourth year, 15% for a fifth year, and 9% for a sixth year.
The probability of voluntarily staying in the teaching profession over time is higher than the likelihood of staying in the initial placement school, since some TFA recruits, like teachers in general, transfer to other schools. The fitted models suggest that about 94% of TFA recruits remain in teaching for a second year, and 60% teach for a third year. 44% remain in teaching for a fourth year, 35% for a fifth year, and 29% for a sixth year.
It’s difficult to know whether to think of these rates of persisting in the initial school placement or in teaching at large are high or low. As usual, the question is, compared to what? TFA recruits are placed in schools that are claimed to be "hard to staff," and they may be challenging places to work, regardless of the route that brought the teachers to such schools. If the attrition rates for other novice teachers in these schools are just as high as those observed for TFA recruits, it’s harder to argue that TFA is exacerbating the problem of building a stable, high-quality teaching force in high-needs schools. Donaldson’s study doesn’t shed any light on this issue.
I’ll have a bit more to say about Morgaen Donaldson’s research on how working conditions affect the persistence of TFA recruits in their initial schools tomorrow.