Cognitive Ability, Wages, and Meritocracy
Intelligence, Genes, and Success
This paper presents new evidence from the NLSY on the importance of meritocracy in American society. In it, we find that general intelligence, or "g"--a measure of cognitive ability--is dominant in explaining test score variance. The weights assigned to tests by "g" are similar for all major demographic groups. These results support Spearman's theory of "g." We also find that "g" and other measures of ability are not rewarded equally across race and gender, evidence against the view that the
... he view that the labor market is organized on meritocratic principles, Additional factors beyond "g" are required to explain wages and occupational choice. However, both blue collar and white collar wages are poorly predicted by "g" or even multiple memures of ability. Observed cognitive ability is only a minor predictor of social performance. White collar wages are more "g" loaded than blue collar wages. Many noncognitive factors determine blue collar wages. In their controversial book The Bell Curve, Richard Herrnstein and Charles Murray summarize an impressive body of research on the correlations between social outcomes and scores on tests of cognitive ability. A remarkable finding of the research they survey is that one linear combination of tests -called "g" -predicts performance almost as well as the full battery of tests. 1 Charles Spearman first proposed that general intelligence, or "g", is a common ability that explains petiormance on all tests of intelligence. General intelligence was also thought to be heritable although that is a completely separate matter.2 Both assumptions have been questioned in the scholarly literature. Theories of multiple abilities go back to Thurstone (1947) . Carroll (1993) provides a comprehensive discussion of the evidence. The theory of the heritability of intelligence is simplified by, but does not require, unidirnensioml ability. The Bell Curve embraces both "g" and heritability. Moreover, it extends Spearman and attempts to demonstrate that differences in "g" explain discrepancies in social outcomes across race. This paper examines the arguments for, and the empirical evidence about, g. Using the NLSY (National Longitudinal Survey of Youth) data employed by Murray and Herrnstein we demonstrate that "g" explains a majority of the variance in test scores. Other combinations explain at most a fifth of what "g" explains. Moreover, the weights of "g" on the constituent tests are remarkably similar across race and gender. The classical theory of "g" is alive and well in the NNY. Ironically, while Herrnstein and Murray embrace the theory of "g", they use a different (though highly correlated) measure of ability in their analysis. 1 "g" is formed by taking principal components of the correlation matrix of test scores. me component associated with the largest eigenvalue is multiplied by the test scores to form g. Prediction is measured by R-squared-i, e. the proportion of variance explained. 'See Gould (1979) for a disparaging review of the early psychometric literature. Carroll (1993) presents a more balanced discussion.