FRBSF Economic Letter
1999-22 | July 23, 1999
In recent years an increasing number of observers have argued that changes in the operation of the U.S. labor market have led to rising job instability and job insecurity for American workers. The manifestations of this include an apparent rise in layoffs and a corresponding deterioration in workers’ attitudes about the prospects for staying in their current jobs. Much of the evidence has been presented in news articles about layoffs and downsizings and results from public opinion polls. However, researchers remain in substantial disagreement regarding these issues, at least in part because formal analysis of such developments in the labor market is hindered by lags in data availability.
With the intent of updating the research and perhaps moving towards a consensus, the Russell Sage Foundation of New York City sponsored a conference that brought together scholars who have been at the forefront of research in this emerging area. The conference was organized by David Neumark, professor of economics at Michigan State University. In this Economic Letter, I discuss seven of the conference papers that were selected for publication in a special issue of the Journal of Labor Economics (forthcoming October 1999). All of these papers, along with most of the remaining conference papers, are forthcoming in a book to be published later by the Russell Sage Foundation.
Job stability refers to the duration of a typical match between a worker and a firm, which is a useful broad measure of labor market conditions. A key goal of the Russell Sage Foundation conference was to assess trends in job stability using a variety of different data sources. This was done in the papers by Neumark, Polsky, and Hansen (NPH), Jaeger and Stevens, Bernhardt, Morris, Handcock, and Scott (BMHS), and Gottschalk and Moffitt. As a group, these papers use four different data sets and make comparisons across several different pairings, thereby providing relatively comprehensive evidence regarding trends in job stability.
NPH analyzed data from various supplements to the Current Population Survey (CPS). The CPS is the monthly household survey conducted by the U.S. Bureau of Labor Statistics. The supplements contain information on employee tenure (time spent with a particular employer), which is the most commonly used measure of job stability. This data set has been a key source of information for previous studies of job instability, in part because it provides a large, representative sample of American workers.
Previous studies using the CPS supplement data found little or no evidence of a drop in job stability between the 1970s and the early 1990s. In their conference paper, however, NPH found support for the declining stability view using more recent data. They used data from the years 1983, 1987, 1991, and 1995 to examine changes over time in the percentage of workers whose tenure exceeds a four-year or eight-year cutoff. They found a mild decline in aggregate job stability during the first half of the 1990s. This trend towards declining job stability was more pronounced in each of four age groups than in the aggregate, since the aggregate trend was limited somewhat by a demographic shift in the workforce towards older, more stable workers. Moreover, the aggregate trend masked marked differences across sub-groups. Men with substantial tenure experienced a sharp decline in job stability during the first half of the 1990s; this was offset somewhat in the aggregate by an increase in job stability for less-tenured workers. Blacks of both sexes experienced a substantial decline in job stability over the full sample period. White women experienced an increase or a slight decrease in job stability depending on the period over which the comparison was made.
Jaeger and Stevens also used the CPS data on job duration. Moreover, they used another key data set that has been used in past work on job instability–the Panel Study of Income Dynamics (PSID)–and after appropriate data adjustments were able to reconcile the results from the two data sets for the period 1983-96. Their results disaggregated by age generally are consistent with those of NPH regarding declining job stability for senior workers. In particular, in both data sets Jaeger and Stevens found a decrease in the fraction of workers over age 40 who had more than ten years of tenure with their current employer, with a sharp drop in stability for men more than offsetting an increase for women. Declining stability was most pronounced during the 1990s (through 1996) in their data. Moreover, when they controlled for observable worker characteristics such as age that affect job stability and whose distribution may have changed over time, they found an even greater decline in job stability.
The paper by BMHS also supports the hypothesis of declining job stability. However, BMHS focused on young workers in the National Longitudinal Survey (NLS). They used NLS data on two cohorts of young men: (1) those born between 1944 and 1952 who were surveyed in most years from 1966 to 1981; (2) those born between 1957 and 1965 who were surveyed yearly from 1979 to 1994. In analyses that controlled for a wide variety of worker characteristics, BMHS found a marked increase in the probability of job change for young white men between the 1980s and early 1990s compared to the late 1960s and 1970s. These results are broadly consistent with NPH’s finding of rising job instability among young workers. Moreover, BMHS compared their NLS data to a similarly defined extract from the PSID and verified that the trends in the two data sets are very similar.
Among the conference participants, only Gottschalk and Moffitt found essentially no evidence to support the view that job stability or security has declined. Their work focused on data from the Survey of Income and Program Participation (SIPP), which they compared to the PSID. However, the SIPP data are available only for the years 1983-1995. For this period, and for the period 1981-1992 in the PSID, Gottschalk and Moffitt analyzed a sample of men and found no increase in job changes, either in unadjusted tabulations or in analyses that controlled for worker characteristics. Their results suggest that any increase in aggregate male job instability occurred between the 1970s and 1980s. The conflict between their results and those of NPH and Jaeger and Stevens probably is explained by subtle methodological differences and by Gottschalk and Moffitt’s focus on the aggregate trend, to the exclusion of the key subgroups analyzed by these other authors.
Although job stability is an important attribute of the employer-employee relationship, recent concern has focused as much or more on the issue of job security. Job stability depends on voluntary job change (quits) and involuntary job change (due to layoffs, plant closures, firings, etc.). In contrast, job security refers to workers’ ability to retain a desirable job–that is, to avoid involuntary job loss. The distinction is important. For example, an increase in layoffs accompanied by declining quits may produce constant job stability over time, although workers’ perceived and actual job security may have fallen due to the rising incidence of layoffs. Thus, despite the importance of job stability per se, understanding trends in job security requires separate analyses.
Among the various measures of job security, probably the most commonly cited is generated from worker responses to opinion poll questions. In her conference paper, Schmidt analyzed such data obtained from the General Social Survey, administered by the National Opinion Research Center to a nationally representative sample each year from 1977 to 1996. The results indicated that conditional on individual characteristics, industry, and region of residence, American workers were more worried about keeping their jobs and about the potential costs of job loss during the years 1990-96 than they were during years with comparable unemployment rates in the late 1970s and late 1980s.
The conference paper by Valletta went beyond worker attitudes by analyzing changes in actual job separations in the context of a behavioral model of worker-firm attachment that provides a definition of job security. In the model, workers with substantial tenure are identified as having implicit (unwritten and unspoken) agreements with firms regarding joint investment in firm-specific skills and continued cooperation by both parties. In this scenario, job security represents workers’ understanding that the employment relationship will continue as long as they continue to exert appropriate effort on the job. However, the labor market conditions that determine the stability of such implicit agreements are likely to change over time, which may produce rising incentives for the firm or the worker to terminate the employment relationship.
Valletta used PSID data for the period 1976-92 to test for such changes in the incentives to maintain implicit employment agreements. The results indicated that conditional on individual characteristics (including the wage) and the prevailing national unemployment rate, male workers with substantial job tenure experienced a rising incidence of permanent layoffs and firings between 1976 and 1992. The results also indicated that such separations are especially common in declining industries. Moreover, quits by high-tenure male workers rose between 1976 and 1992. The implied symmetry in worker and firm behavior suggests that the underlying changes in labor market operations have changed the behavioral incentives of both groups. Similar to the results from the conference papers that found reduced job stability, Valletta’s finding of reduced job security was restricted to male workers.
Another potential reason for a drop in perceived job security is deterioration in the economic consequences of job change. BMHS found that for young men the wage returns to job change have declined, which provides an added incentive for workers to worry about job security and cling to existing job pairings. Moreover, in his contribution to the conference, Farber found that during the years 1994-1997 involuntary job losers were more frequently employed in temporary and part-time jobs than were workers who had not lost jobs. Although Farber’s data did not enable comparisons across the time periods used in the other conference papers, his results suggest that the probability of being underemployed through temporary or part-time work is an important component of workers’ rising fears about job security.
The conference papers as a whole paint a relatively consistent picture of declining job stability and security among male workers. The various data sources suggest that job stability, as measured by expected job duration or separation probabilities, declined for older, senior male workers between the 1970s and the first half of the 1990s, and perhaps declined for young male workers as well. The decline in stability for senior male workers is consistent with Valletta’s results regarding termination of implicit employment agreements with such workers. A rising incidence of quits by such workers suggests that declining job attachment does not necessarily indicate a decline in workers’ well-being, although an apparent deterioration in the consequences of job loss (BMHS) indicates that concerns about workers’ well-being may be appropriate. As a whole, these results validate to some degree the rising job insecurity documented in various public opinion polls (Schmidt).
Whether these changes reflect an ongoing long-term trend is not yet clear. The collected results suggest that the largest observable decline in job stability and security occurred during the first half of the 1990s, which corresponds to a recessionary period followed perhaps by ongoing structural adjustments in industries and corporations. Whether the manifestations of declining job stability and security have extended to the second half of the decade is a crucial question, especially given the strong general labor market conditions that have prevailed in recent years.
Russell Sage Foundation conference papers, forthcoming in Journal of Labor Economics 17(October, part 2).
Bernhardt, Annette, Martina Morris, Mark S. Handcock, and Marc A. Scott. “Trends in Job Instability and Wages for Young Adult Men.”
Farber, Henry S. “Alternative and Part-Time Employment Arrangements as a Response to Job Loss.”
Gottschalk, Peter, and Robert Moffitt. “Changes in Job Instability and Insecurity Using Monthly SIPP Data.”
Jaeger, David A., and Ann Huff Stevens. “Is Job Stability in the United States Falling? Reconciling Trends in the Current Population Survey and Panel Study of Income Dynamics.”
Neumark, David, Daniel Polsky, and Daniel Hansen. “Has Job Stability Declined Yet? New Evidence for the 1990s.”
Schmidt, Stefanie R. “Long-Run Trends in Workers’ Beliefs about Their Own Job Security: Evidence from the General Social Survey.”
Valletta, Robert G. “Declining Job Security.”
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.
Please send editorial comments and requests for reprint permission to
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120