The Short-and Long-Run Impact of Comparative Noncognitive Skills

This study documents a new fact about educational production: Students' relative standing in noncognitive skills has lasting effects on educational trajectories that are distinct from those of absolute skill levels and cognitive achievement. Using administrative data covering the universe of public school students in Greece, I exploit quasi-random classroom assignment to identify the causal impact of comparative noncognitive skills-measured as students' grade 10 classroom rank in grade 9 unexcused absences. A worse ranking in noncognitive skills generates persistent and nonlinear effects on short-run academic performance and longer-run educational trajectories. A worse behavioral rank predicts weaker achievement in Greek Language and Mathematics for both boys and girls in grades 10 and 11. Starting in grade 11, however, responses diverge by gender. A worse ranking in noncognitive skills predicts stronger sorting into vocational schools and less competitive academic tracks for girls, whereas boys sort into more competitive tracks. A worse ranking in noncognitive skills is also associated with greater participation in and higher performance on national examinations for boys only. Boys with a worse ranking in noncognitive skills also apply more frequently to, and are more likely to gain admission to, higher-earning degree programs, while girls' admissions and expected earnings are unaffected. Gender differences in comparative noncognitive skills account for 37 percent of the gender gap in expected post-college salaries. Complementary evidence from a survey experiment shows that comparative behavioral labels systematically shift teachers' expectations and attribution patterns for otherwise identical students. This suggests that relative-standing effects operate through belief-driven institutional responses.