When Mathematica's thorough study on secondary math teachers from Teach for America and Teaching Fellows programs came out a couple of months back, the headlines rightly focused on the main finding: TFA corps members were more effective than the average teacher - producing the equivalent of 2.6 months of additional learning in math for a typical student.

But the study merits a close read for its other findings as well. One of these suggests that on-the-job training comes at a significant cost to teacher impact. When compared with their colleagues who weren't enrolled in coursework, the typical teacher in the study taking courses produced the equivalent of 1.5 fewer months of learning gains in math.

Now, the authors of the report caution that the study's design does not allow one to say that the coursework caused the decline in effectiveness. But it would seem to fit: teaching is hard work all on its own, and taking 219 hours of coursework in a year on top of it to become certified (the typical scenario for teachers in the study taking courses) is surely something of a drain. Unfortunately, the courses don't seem to add much in return for the time they take.

At the very least, this finding should spur the directors of alternate route programs to review each of their required courses with an eye toward cutting those that may not be delivering value.