There is a new paper by Papageorge, Ronda, and Zheng, with a very interesting thesis, namely that preparing rowdies for better schooling results may not help their long-term prospects in life:
Prevailing research argues that childhood misbehavior in the classroom is bad for schooling and, presumably, bad overall. In contrast, we argue that childhood misbehavior reflects underlying traits that are potentially valuable in the labor market. We follow work from psychology and treat measured classroom misbehavior as reflecting two underlying non-cognitive traits. Next, we estimate a model of life-cycle decisions, allowing the impact of each of the two traits to vary by economic outcome. We show the first evidence that one of the traits capturing childhood misbehavior, discussed in psychological literature as the externalizing trait (and linked, for example, to aggression), does indeed reduce educational attainment, but also increases earnings. This finding highlights a broader point: non-cognition is not well summarized as a single underlying trait that is either good or bad per se. Using the estimated model, we assess competing pedagogical policies. For males, we find that policies aimed at eliminating the externalizing trait increase schooling attainment, but also reduce earnings. In comparison, policies that decrease the schooling penalty of the externalizing trait increase both schooling and earnings.
For the pointer I thank the excellent Kevin Lewis.