Abstract
This paper develops a series of information-theoretic measures to consider the systemic effects on individual incomes of complex patterns of social and economic discrimination by race, ethnicity, and gender, in the U.S. It derives coefficients of joint, conditional or incremental, and mutual information that offer non-parametric characterizations of the relative influence of economic and social-identity characteristics in the determination of individual income for different groups. It reports on estimates of those coefficients obtained using large-scale cross-sectional data from that economy. Those estimates support two sets of conclusions. First, the informational significance of social identity in the determination of incomes differs clearly and persistently across social-identity groups. For some groups social identity exerts a significant informational influence in the determination of income. Other groups enjoy greater scopes for individual differentiation by factors other than social identity. Second, the informational influence of educational attainment on income is deeply shaped by social identity. Among other expressions of this, the paper finds that some identity groups see the comparative measure of informational association between their incomes and educational attainment rise steadily with levels of educational attainment. In contrast, other groups see those comparative measures fall as educational attainment rises. These observations point to the economic effects systems of discrimination impose on certain groups, and to the relative privileges enjoyed by those not subjected to them.
