Race, Children's Cognitive Achievement and The Bell Curve

In The Bell Curve, Herrnstein and Murray demonstrate that a mother's score on the Armed Forces Qualification Test is a powerful predictor of her child's score on a cognitive achievement test. We replicate this finding. However, even after controlling for maternal scores, there are significant gaps in the scores of black and white children which suggests that maternal scores are not all that matter. In fact, both maternal education and income are important determinants of child test scores, conditional on maternal AFQT. We argue that racial gaps in test scores matter because even within families, children with higher scores are less likely to repeat grades. However, conditional on both child test scores and maternal AFQT, maternal education and income also affect a child's probability of grade repetition. We conclude that, even if one accepts test scores as valid measures of 'nature', both nature and nurture matter. Finally, we show that the effects on child test scores of maternal test scores, education, and income differ dramatically depending on the nature of the test, the age of the child, and race. The results suggest that understanding the relationships between different aspects of maternal achievement and child outcomes may help us unravel the complex process through which poverty is transmitted across generations.