Artigo Revisado por pares

Filtering by Race and Education in the U.S. Manufacturing Sector: Constant-Ratio Elasticity of Substitution Evidence

1976; The MIT Press; Volume: 58; Issue: 2 Linguagem: Inglês

10.2307/1924020

ISSN

1530-9142

Autores

Daniel T. Dick, Marshall H. Medoff,

Tópico(s)

Global trade and economics

Resumo

M UCH of the empirical research done on racial discrimination deals with measuring the magnitude of discrimination by determining what percentage of the equilibrium wage differential between blacks and whites can be attributed to discrimination (see, for example, Bergmann, 1971; Medoff, 1973; Sanborn, 1964; Sawhill, 1973; Thurow, 1969). An alternative view of discrimination suggests that race conveys information to employers, rightly or wrongly, and that this information is used by employers in placement and promotion decisions. That is, because information about an individual employee's productivity is costly to obtain, employers attempt to economize. In doing so they screen, or filter, persons according to easily obtainable attributes such as race, and the number of years of formal schooling completed (Arrow, 1973; Mantell, 1974, p. 158). These attributes' serve as a signaling device which helps employers evaluate the productivity of employees. Thus, filtering or screening can be said to occur when a given employee attribute prevents him from holding a job with a greater marginal product than his present job, while others without the attribute can hold the more productive job. The screening characterisic may or may not have much to do with the employee's success at the more productive job,' but in a world where information is costly, the fact that the employers think the screening characteristic conveys productivity information to them is sufficient for them to screen.2 The higher the costs of screening, the less effective is the screen. This is simply the fundamental law of demand: less of an activity is undertaken at a higher than at a lower cost. For example, screening costs are likely to be higher for religious characteristics than for racial or sexual characteristics (Demsetz, 1965, pp. 282-283). According to Arrow, a screen, or filter, is said to be complete with respect to a certain job when screening on a given attribute excludes completely all employees with the attribute from acquiring that job (1973, p. 205). For example, while a high school diploma is a screening characteristic with respect to shipping clerks, it is not complete. There are shipping clerks without high school diplomas. One would expect that filtering on high school diplomas is more complete with respect to brain surgeons.

Referência(s)