RELATIVE EARNINGS MOBILITY OF WOMEN IN THE UNITED STATES

Theoretical and empirical studies of earnings mobility have generally focused on relative cross-sectional earnings distributions or on absolute longitudinal changes in individual earnings. Recently, interest has increased in relative earnings mobility, the degree to which individuals' earnings change relative to the earnings of others over time. The pioneering research in this area, however, has been largely empirical. This dissertation proposes a theoretical framework for the study of relative earnings mobility and provides empirical tests of that theory. Relative earnings mobility occurs, in a general sense, when the age-earnings profiles of individuals intersect. Such intersections occur when the slopes of individual age-earnings profiles vary. Many variables are hypothesized to influence the height and slope of these profiles, and thus the likelihood of their intersection. Among these variables are economic class, occupation, industry, unionization, region, cyclical variations, sex, and race. Examining the occupational and industrial distributions of women, their rates of unionization, and so forth, the study hypothesizes that the age-earnings profiles of women are lower, flatter, and less variable than those of men, thus women would be expected to exhibit less relative earnings mobility than men. Subsamples of women by region, industry, race, and age would be expected to exhibit distinct patterns of earnings mobility. Empirical tests are done using the Social Security Administration's Longitudinal Employer-Employee Data file (LEED). For a sample of women continuously employed from 1957 to 1971, a first-order Markov process yields estimates of the rate at which women move between rank-ordered state spaces of a transition probability matrix. The test results are consistent with many of the theoretical propositions. Testing of subsamples shows that rates of relative mobility do vary with sex, and with age, race, industry, and region, as hypothesized.