This paper reexamines the link between career interruptions and subsequent wages. Using a rich new Swedish data set, we. are able to disaggregate time out of work into several components. In both cross-sectional and panel estimations, regressing log wages on total time out results in a negative coefficient on total time out, which has been interpreted in other studies as evidence for human capital depreciation. However, we find that different types of lime out have different effects on wages and that these effects vary by gender. This suggests that human capital depreciation is not the entire explanation for the negative effect of career interruptions on subsequent wages.

James W. Albrecht and Susan B. Vroman are professors of economics at Georgetown University. Per-Anders Edin is a professor of economics at Uppsala University and NBER. Marianne Sundstrom is on the staff of the Demography Unit, Stockholm University. They thank the Demography Unit of Stockholm University for permission to use and assistance with the data and the Swedish Council for Social Research for supporting work on the data set. Albrecht and Vroman thank the American Scandinavian Foundation, the Institute for Social Research of Stockholm University, and the Swedish Council for Work Life Research, and Sundstrom thanks the Swedish Research Council for Humanities and Social Sciences for financial support. Tile data used in this article can be obtained for a fee from Statistics Sweden, Stockholm.