Teach for America’s Impact Is Modest at Best

by Jason Richwine

What happens to student test scores when super-high-achieving college graduates spend a couple of years teaching in low-income schools? Well, not a whole lot, according to a new random-assignment study from the policy-research group Mathematica.

Most NR readers are probably familiar with the Teach for America (TFA) program. TFA is a non-profit whose mission is to close the socioeconomic achievement gap by recruiting especially smart, driven, and idealistic young people for a two-year teaching commitment. Previous randomized experiments have found that TFA teachers had a small positive effect (compared to regular teachers) on math scores but no effect on reading.

Five years ago TFA endeavored to scale-up its operations with the help of $50 million from the federal government. Mathematica’s new report evaluates the effectiveness of TFA in the midst of that expansion. Scale-up is a well known problem with social interventions. A small demonstration project may show gains for certain students in a certain environment, but scaling up the program to multiple sites often results in the average treatment effect falling to zero.

TFA was already a fairly large operation, and the expansion was not dramatic. Nevertheless, Mathematica found that TFA’s effect on the grades pre-K–5 that it studied during the scale-up was statistically indistinguishable from zero. In other words, TFA teachers performed about the same as regular teachers on average. Looking at subgroups, TFA had a small but significant positive effect on reading scores in grades pre-K–2. But reading impacts in grades 3–5 were not significant, nor were impacts on math at any level.

What to make of these results? On the bright side, TFA teachers appear to be at least as effective as regular teachers when it comes to teaching reading and math to elementary students. The fact that TFA requires only a five-week crash course in pedagogy — rather than traditional teacher certification — is another reason to question the value of an education degree. In addition, we know that teachers tend to improve with a few years of experience, so TFA volunteers who stay in the teaching profession may make a more substantial impact down the road. (Mathematica could not detect any outcome differences for TFA teachers versus novice regular teachers, but the sample size was too small to draw a meaningful conclusion.)

Still, even the most optimistic reading of the data suggests that TFA’s impact on student learning is no more than modest. Some advocates claim that our education system could be fundamentally transformed if teaching were an elite field populated by the best and brightest. TFA is to some extent a test of that proposition, and the results suggest no transformations are at hand. Teachers matter, of course, but even the smartest and most driven teachers will not be able to systematically convert below-average students into above-average ones.

The Corner

The one and only.