Abstract
Noncognitive skills have provoked substantial research interest in recent years. However, existing evidence for their importance relies heavily upon reduced-form regressions that do not account for the simultaneous production of cognitive and noncognitive skills. Moreover, existing constructs of noncognitive skill are typically chosen for their convenience, or else for their explanatory power over psychological survey items. In response, we analyze a decision-theoretic model of educational development to derive five candidate noncognitive skills, and we test their importance by adapting an established longitudinal and structural econometric model of multidimensional skill formation. We find: first, that noncognitive skills matter for cognitive development; second, that different aspects of noncognitive skills matter to different degrees; and third, that once a child begins to take her own decisions, her propensity to think analytically becomes the most important noncognitive determinant of her ongoing cognitive development.
Iain Embrey
Microeconomic research toward equal opportunity for disadvantaged children.