and 1980s, used data assembled from a sequence of Current Population Survey (CPS) tenure supplements. They use these CPS supplements to construct artificial cohorts, with which job retention rates (i.e., the probability of retaining one's job) over various spans of years can be computed and compared over time. This is preferable to looking at distributions of incomplete tenure spells at a point in time, which can be affected by, among other things, new entrants to the labor market, and hence obscure information on how long jobs are lasting.
Through the 1980s to January 1991, they found little or no change in average job stability for the workforce as a whole in the U.S. economy (see Table 3.1). Columns (1), (2), and (4) in Table 3.1 summarize these results for 4-year retention rates. Columns (1) and (2) report the estimated rates, and column (4) reports the changes. As shown in the first row, the overall change is a very slight decline of .002. As shown in the other rows, at a disaggregated level, job stability fell for those with 2–9 years of tenure and rose for the least and especially the most tenured workers (15 or more years of tenure). By age group, job stability fell only for the youngest workers. Column (7) reports changes in estimated 10-year retention rates, over a longer sample period. Here, too, the overall change is negligible (-.001), although the pattern by age and tenure group is somewhat reversed.
Columns (3), (5), and (6), as well as column (8), report newer results from Neumark et al., (1997), updating the estimates through 1995 (Wyatt Company, 1995). The estimates reported in the table reflect a relatively consistent picture: first, 4-year retention rates still show essentially no overall change, if anything rising slightly compared with earlier 4-year spans. However, in the 1991–1995 period, job stability declined rather substantially for workers with 9 or more years of tenure, a finding that may reflect the types of stories that have appeared in the media regarding changes in job stability for "career" employees. In the 8-year retention rate results reported in column (8), similar (although less sharp) declines for more-tenured workers appear. Moreover, for the period considered, 8-year retention rates fell modestly overall, by .021. Finally, the results disaggregated by age indicate non-trivial declines for all age groups except those age 55 and over. This suggests that conditional on age, job stability has declined