WOMEN'S EARNINGS REEXAMINED

Abstract

Since a discussion of details may obscure important general issues, we start by restating what we consider to be the two major contributions of our 1974 study [5]: 1. The methodological contribution lies in the specification of an earnings function when human capital investment occurs at different rates (including zero) over various segments of the life cycle. Such specifications are appropriate for persons when market work does not continue throughout the usual working age, or for analyses of continuous work histories segmented by some relevant criteria such as job change and occupational or geographic mobility. Applied work on these subjects continues [1], and specifications are undergoing refinement. In retrospect, the model we used to analyze earnings of women is not the best, but it is serviceable, which permits us to refrain from injecting new matters into the present discussion. 2. The substantive interest in our study of earnings of women focuses on several propositions: (a) Even with identical current characteristics, women whose past experience was shorter have currently lower wages, as their total job-investments were smaller. (b) Total investments are smaller not only because they are spread over fewer years of work experience, but also because persons who spend less time in market work invest less per year worked than persons with longer work experience. (c) If work experience is interrupted for any significant length of time, job skills tend to erode, so the human capital stock is reduced by depreciation or atrophy. The relevance of these propositions to the analysis of the sexual wage gap is obvious. In part, women's wages are lower than men's because their labor force attachment is shorter and often discontinuous. Also, the gap widens over the life cycle, since initial differences in experience are necessarily small.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 84 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.