What might be called the traditional theory of discrimination in labor markets argues that there are two types of workers who are indistinguishable in all respects save one; this can be color, sex, religion, or anything else. 'I'his difference does not affect their productivity of marketable output but it does affect the production of nonmarketable output. That is to say the two groups are perfect substitutes in the production of market goods, but they also produce subjectively valued utilities (disutilities) 1 If the value of the nonmarket goods is positive to the employer, members of such a group are said to be preferred; it is negative they are nonpreferred and it is zero they might be called neutral. Tl'he value to an employer of a particular employee is equal to the sum of the marketable plus nonmarketable outputs. Accordingly, workers with negative nonmarketable outputs receive lower wages than workers with nonnegative nonmarketable outputs. The wage differential is a measure of the extent of discrimination between groups of otherwise identical workers (see Harold Demsetz). Economists tend to shy away from questions about the formation of preferences; the discipline is more adept at deriving consequences of different preference structures. Accordingly, economic literature is mostly silent on the question of why the nonmarketable output of some workers is valued positively and others negatively. Instead economists say agents act if they preferred some outcomes (situations, goods, employees, etc.) to others. T'o understand wage differentials based on discrimination, it is useful to think in terms of costs as distinguished from prices. The wage price paid to workers is only part of the cost of employment. T'here are also nonpecuniary costs. If an employer finds the presence of some workers distasteful, he has an aversion to such workers; or, their presence is a disutility to him; then their nonmarketable output is negative and imposes a nonpecuniary cost on him. The net cost of employing such a nonpreferred worker is

Marcus Alexis, Wealth Accumulation of Black and White Families: The Empirical Evidence: Discussion, The Journal of Finance, Vol. 26, No. 2, Papers and Proceedings of the Twenty-Ninth Annual Meeting of the American Finance Association Detroit, Michigan December 28-30, 1970 (May, 1971), pp. 458-465