own helplessness in the face of forces that do not appear amenable . . . even to collective political mechanisms of control." 68
A surprising limitation of Gordon's work, however, is his meager discussion of race and ethnicity. Although he often refers to "ethnic segregation" and to a reserve "army of the unemployed-jobless workers available for immediate employment" whose existence would "help discipline those inside the factory gates," he does not examine the racial makeup of this reserve army of labor or discuss the implications for residential settlement of a racially divided workforce. (Some efforts in this direction are contained in Green's 1984 study of antebellum Richmond, Virginia.) 69 Meanwhile, Robert Weaver observed in 1950 that
The color line in employment was well entrenched in the United States by 1929. The whole economic, political and social structure of the South dictated and supported it. In the North, long experience with importation of Negro strikebreakers had created anti-Negro sentiment on the part of the white workers. Institutionalized racial segregation in the South and spatial segregation in the North supported these attitudes. In all parts of the nation, the black worker had become a symbol of a potential threat to the white worker. This fear had grown out of the American worker's experience with an economy which has seldom had enough jobs to absorb the labor supply. 70
After World War II, despite a growing theoretical and empirical challenge, mainstream explanations of urban residential segregation continued to be based on psychological factors. For some neoclassical economists, race relations, like any other commodity, are determined by individual tastes and invisible market forces. For others, the visibility of racial minorities and whites' fears of status loss were at the root of racial residential segregation; improving the economic conditions of racial minorities (particularly blacks) without eliminating these race-connected barriers to integration could, in Taueber's view, increase segregation.
For David Gordon, urban residential segregation, at least in its economic dimension, can best be understood and analyzed in the context of capitalism and capital accumulation; qualitative changes in the method of capital accumulation lead to qualitative changes in the pattern of urban spatial settlement. In the commercial city, the small scale of economic operations allowed residential settlement patterns to be essentially random. As the scale of production and capital accumulation increased in industrial and corporate cities, qualitatively different patterns of residential segregation based on economic status of various groups emerged. Like those of the urban ecologists, Gordon's model of segregation arrays residential neighborhoods in concentric zones around the central business district with distance of each neighborhood from the center related to the income of its residents. Unlike the urban ecologists, Gordon takes some account of the deeper economic patterns driving these income distributions.