Stone points out that the gains made by Board-certified teachers amounted to fourteen percent of one standard deviation. What does that mean? Anything less than ten percent of a standard deviation is basically trivial. Stone asserts that such gains should be considered only slightly above trivial.
Of note, the math gains produced by Board-Certified teachers were nearly identical to that of Teach For America (TFA) teachers according to the recently-released Mathematica study. While the gains produced by both sets of teachers may be identical, the policy implications are quite different. For one, if National Board certification was not available to teachers, by all indications these National Board teachers would still be teachers and would still presumably be producing the same student gains. Goldhaber and Anthony did not, for example, find that designating a teacher as Board Certified made the teacher any more effective than he or she would have been otherwise. On the other hand, if TFA did not exist, it's unlikely that most of its young recruits would have entered teaching and the gains of their students would not have taken place.
Further, when you consider the fact that many states pays $4,600 for each board-certified teacher (since states pay for all applicants and only one out of two are successful) and then offers significant salary bonuses, serious questions remain over whether the high cost of NBPTS merits such hefty public investment. There is some possibility that NBPTS reduces turnover, but that claim has yet to be empirically tested. Meanwhile, TFA costs districts relatively little, a cost which is often offset by the amount they save in having to recruit and support new teachers.