Racial and gender disparities found in most other societies are particularly magnified in South Africa where the marginalized social group constitutes a numerical majority of the population. These factors, along with region, are dominant axes of inequality in the country. However, empirical knowledge of the interplay between these systems of social inequality in determining employment outcomes remains somewhat scant. This dissertation addresses that gap by studying occupational sex segregation across various racial groups using multilevel modeling techniques. Individual-level data from the 2001 Census and magisterial-level data from survey data aggregations and published sources are used.

I first study the influence of (1) individual (education and migration), (2) household characteristics (family status), and (3) contextual factors (urbanization, former homeland residence, industrial composition, and culture) on women's likelihood of being in white- and blue-collar male-dominated occupations. I then investigate whether the effects of these individual and contextual characteristics on occupational placement vary across the four main racial groups (i.e., black African, Coloureds, Asian-Indians, and Whites)? That is, do these factors interact differently for different racial groups?

Results from the analyses indicate that high concentrations of service industries tend increased women's opportunities for holding white-collar male-dominated occupations. On the other hand, while black African women's placement in male-dominated jobs is not influenced by urbanization, women of other races, particularly Coloureds and Whites, fare better in urban districts. In fact, residence in and around homelands was particularly significant for black Africans who are still trying to gain a foothold in mainstream South African society. In the unique case of Indian women, labor supply factors such as education, have greater predictive power than macro-level demand factors.

In terms of human supply variables, educational attainment improves women's chances of holding male-dominated occupations among white collar workers across all racial groups; the effect is not as strong among blue-collar workers. However, returns to education are not as high as expected. Migration is significant for only black Africans, highlighting the legacy of apartheid. Finally, women's marital status and associated short-term child bearing obligations do not act as impediments to their occupational choices.