Occupational Mobility and the Distribution of Occupational Success Among Young Men

Abstract

Interpretation of the distribution of occupational success depends crucially on an understanding of the process of mobility. If the holders of low-paying jobs this year are just as likely as anyone else to hold good jobs next year, then the lower tail of the earnings distribution is not a matter of great social concern. Further, if the distribution of success is primarily determined by a random process, the prospects for a significant reduction in the inequality of earnings are unfavorable. On the other hand, high mobility in the short run may conceal systematic stratification of the labor force-it may be that the same individuals spend most of their time in the worst jobs. If so, luck and random events may play a small role in the distribution of well being, inequality may be a major issue, and the elimination of stratification may hold out the promise of an important reduction in inequality. Simple tabulation of distributions of success and of mobility among categories of success cannot answer the basic question of the role of stratification. In this paper, we develop a model and statistical method for studying mobility and relating it to characteristics upon which stratification may be based. Our study rests on annual interviews conducted from 1966 through 1969 with about 3,000 white males born between 1942 and 1952.1 We find that mobility is closely related to personal characteristics, especially intelligence and education. Men with high IQs and extensive education are more likely to move to high-paying jobs and are more likely to remain in them. Nevertheless, random events dominate the distribution of occupational success in the short run. Our results strongly support the finding of Christopher Jencks and his collaborators that redistribution of the determinants of earnings would do almost nothing to reduce the observed inequality of annual earnings.

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 — 164 citations exist upstream.