A Pollution Theory of Discrimination: Male and Female Differences in Occupations and Earnings
Occupations are segregated by sex today, but were far more segregated in the early to mid-twentieth century. It is difficult to rationalize sex segregation and "wage discrimination" on the basis of men's taste for distance from women in the same way differences between other groups in work and housing have been explained. Rather, this paper constructs a "pollution" theory model of discrimination in which occupations are defined by the level of a single-dimensional productivity characteristic. Because there is asymmetric information regarding the value of the characteristic of an individual woman, a new female hire may reduce the prestige of a previously all-male occupation. The predictions of the model include that occupations requiring a level of the characteristic above the female median will be segregated by sex and those below the median will be integrated. The historical record reveals numerous cases of the model's predictions. For example in 1940 the greater is the productivity characteristic of an office and clerical occupation, the higher the occupational segregation by sex. "Credentialization" that spreads information about individual women's productivities and shatters old stereotypes can help expunge "pollution."
Presented at the NBER-Spencer Conference, "Human Capital and History: The American Record," Dec. 7 and 8, 2012. At its inception, about two decades ago, this paper benefited from the comments of seminar participants at Columbia University, Indiana University, University of Michigan, and Princeton University and from suggestions made by Becky Blank, Nancy Folbre, Alan Krueger, Peter Kuhn and Gavin Wright. In its 2002 incarnation, it profited from comments by Francine Blau and Larry Katz and was an integral part of my Marshall Lectures, University of Cambridge, April 30 and May 1, 2002. The current version has been greatly improved through the able research assistance of Chenzi Xu, a careful reading by Stephanie Hurder, and comments of my discussant Cecilia Rouse.
My work on this project was not externally funded. Research assistance was provided through my general research support at Harvard University, but most of the work was done directly by the author.