Worker’s expectations about losing and replacing their jobs: 35 years of change
Workers were less secure about retaining their jobs in 2010 and 2012 than in 1977 and 1978; they also were less secure about the ease with which they would find a comparable job if they were separated. As might be expected, the two measures of job security track unemployment, although other factors certainly play a role as well.
In a 1980 Monthly Labor Review article, I used data from 1977 and 1978 nationwide public opinion surveys to show that relatively few (7.7 percent) workers surveyed feared the loss of their jobs and a majority (59.2 percent) thought that they could find comparable work without much difficulty.1 Unemployment rates were 7 percent in 1977 and 6 percent in 1978.
Other findings were as follows:
blacks, those with less education, workers with lower earnings, and those with lower skilled jobs tended to experience greater insecurity of employment. They were more likely to fear being laid off in the next 12 months and were more likely to anticipate difficulty in finding a comparable job.
By contrast, workers with more education, those with higher earnings, and those in higher skilled jobs typically expressed a greater sense of security in their present jobs and were more likely to believe that finding a comparable job would be easy.2
The study was based on data from one of the most widely used academic research archives for American social behavior: the General Social Survey (GSS). The nationwide survey, broadly similar from year to year, has been conducted by the National Opinion Research Center at the University of Chicago almost every year since 1972. supported by grants from the National Science Foundation, the GSS uses face-to-face interviews at the homes of English-speaking people 18 years or older living in noninstitutionalized arrangements in the continental United States.3 The results of each survey are representative of the adult population of the United States, as evidenced by their agreement with population characteristics drawn from respondents’ answers to questions posed in surveys of the U.S. Census Bureau. The GSS may be downloaded free from the Internet.4
In 1977 and 1978, the GSS for the first time included two questions about job security, and these questions have been expressed in the same words intermittently across the years since then:
Thinking about the next 12 months, how likely do you think it is that you will lose your job or be laid off—very likely, fairly likely, not too likely, or not at all likely?
About how easy would it be for you to find a job with another employer with approximately the same income and fringe benefits you have now? Would you say it would be very easy, somewhat easy, or not easy at all?
For purposes of analysis in my 1980 article, I based the percentages for “expect to lose job” on the percentages of respondents selecting “very likely” and ”fairly likely” and the percentages for “could find another job” on the percentages for “very easy” and “somewhat easy.” I then pooled the two surveys into a single sample for analysis. A survey question about employment status allowed me to limit my analysis to respondents employed full time (35 or more hours per week).
The job security questions have been included in the GSS 19 times since 1978. In what follows, to get the longest series for comparison, I pooled and replicated the most recent surveys, those for 2010 and 2012. The questions were identical, and I repeated the analytical strategy of the earlier study. This approach made it possible to examine changes in feelings of job security across 35 years.
Compared with workers in 1977 and 1978, workers in 2010 and 2012 expressed significantly less job security. They were more afraid of losing their jobs (11.2 percent versus the earlier 7.7 percent) and were less likely to think that they could find comparable work without much difficulty (48.3 percent versus the earlier 59.2 percent). (See table 1.)
|Characteristic||Expect to lose jobs||Could find comparable jobs|
|1977 and 1978||2010 and 2012||Difference||1977 and 1978||2010 and 2012||Difference|
60 and older
Professional and technical
Managers and administrators
*Statistically significant at p < .05.
Source: Author's calculations from General Social Survey.
Comparisons also were made across 27 categories of worker characteristics over the two pairs of years, and every comparison revealed an increased fear of losing jobs, with 10 (37 percent) of the comparisons reaching statistical significance at the 95-percent confidence interval (p < .05) or better. Regarding how easy it would be to find comparable jobs, 26 of the 27 comparisons indicated that workers’ confidence had decreased, with 16 (61.5 percent) of the comparisons reaching statistical significance.
The 10 significant increases in the fear of losing jobs were for men and women; workers ages 30–39, 50–59, and 60 and older; those with elementary school educations and those with high school educations; white- and blue-collar occupations;5 and craft workers. The 16 significant decreases in how easy it would be to find comparable jobs were among men and women; Whites; those ages 30–39 and 40–49; workers with high school educations, those with some college, and those who did graduate work; white- and blue-collar occupations; managers and administrators, and clerical workers; craft workers; and workers with below-average financial conditions, those with average financial conditions, and those with above-average financial conditions.
If the foregoing measures of job security are valid, they should be correlated with the unemployment rate across the two pairs of years examined. And, indeed, figure 1 shows that the correlation between workers fearing the loss of their jobs and the unemployment rate was positive during that timeframe. The coefficient of 0.7061 is statistically significant. Figure 2 shows that the correlation between how easy it would be to find comparable jobs and the unemployment rate was negative. The coefficient is a statistically significant –0.9039. Beyond demonstrating validity, these positive and negative correlations suggest that the two measures—workers fearing the loss of their jobs and how easy it would be to find comparable jobs—should be negatively correlated. And, indeed, figure 3 shows that an inverse correlation did exist. The coefficient of this correlation is –0.6742 and significant.
The findings came from a comparison of two data points at the ends of a 35-year time series. Taking into account the entire series produced slopes of regression lines which confirm the finding that workers perceived significantly less job security in 2010 and 2012 than in 1977 and 1978. The two slopes are 0.9190 and –3.6339, indicating, respectively, that, in 2010 and 2012, more workers feared losing their jobs, and far fewer workers said that it would be easy to find a comparable job, than in 1977 and 1978. Comparing the slopes suggests that job security diminished more than 3 times as much for ease of finding comparable jobs than for the fear of losing jobs.
Across the 35 years examined, the unemployment rate fluctuated between 4.6 percent and 9.7 percent, a trajectory related chiefly to changes in major demographic characteristics of the workforce, advances in communications and transportation, the rise and subsequent popularity of computers and the internet, the decline of unions, global competition in the manufacturing sector, and growing inequality of wealth. Moreover, the nature of recoveries from recessions varied. Some recessions, such as that in 1980, were followed by relatively rapid recoveries, so that the cycle was V shaped, whereas other recessions, such as the one beginning in December 2007 and lasting through June 2009, were more extended, resulting in a U-shaped cycle.
Because many factors that drove the unemployment rate intensified across the 35 years from the first study to this one, one could speculate that the correlations between the unemployment rate and the two job security measures would be higher during more recent years. To examine this possibility, I pooled the 21 surveys that were conducted over the 35-year period into two groups: 10 from 1977 through 1991 and 11 from 1993 through 2012. Then I computed correlations between each security measure and the unemployment rate in both periods. The correlations for the fear of losing jobs were 0.9200 for 1977–1991 and a lesser 0.6542 for 1993–2012, but both were significant. The slopes were each positive: 1.536 for the first period and 0.673 for the second. The correlations for how easy workers thought it would be to find comparable jobs were similar and significant in both periods: –0.9505 and –0.9119, respectively. The slopes were each negative: –4.200 and –3.443. Because the correlations were not higher in the later period, it appears that job security may be more related to workers’ knowledge of the realities of the labor market than to the factors that underlie those correlations.
another way of assessing workers’ decrease in confidence in their job security is to assess their answers to the two questions across the years between 1977 and 1978, on the one hand, and 2010 and 2012, on the other. In between, the questions were asked on these near-annual survey 17 times, at least every 2 years. It was therefore possible to compare the 11.2 percent of all workers who said that it was “very likely” or “fairly likely” that they would lose their jobs, and the 48.3 percent of all workers who thought it would be “very easy” or “somewhat easy” to find a comparable job, in 2010 and 2012 with the corresponding values across the 17 surveys conducted from 1982 through 2008. The result was that the 11.2 percent is greater than the percentage obtained in 12 of the 17 surveys and the 48.3 percent is less than the percentage obtained in 15 of the 17 surveys. Like the analysis set forth in the previous paragraph, this one indicates that the level of worker confidence was lower in 2010 and 2012 than in 1977 and 1978.
It seems plausible that workers’ expectations about their job security might be a direct reflection of changes in the unemployment rate. I tested this hypothesis by comparing, on the one hand, the two overall measures across the 21 years (1977 through 2012) in which the corresponding questions were included in the GSS with, on the other hand, the unemployment rate reported in labor force statistics from the Current Population Survey. The correlation of workers’ fear of losing their jobs with the unemployment rate was .70: moderate, positive, and statistically significant within a 99.9-percent confidence interval, or p < .001. The correlation of ease of finding comparable jobs was –.90: strong, inverse, and also statistically significant within a 99.9-percent confidence interval, or p < .001. These correlations constitute strong evidence that workers’ expectations about their job security are highly accurate reflections of the realities of unemployment.
Because the 2010 and 2012 surveys were administered soon after the worst economic downturn, from December 2007 to June 2009, since the Great Depression, their results might be seen as anomalous—in line with the basic downward trend reported in this article, but statistically of little value. That is, some may infer that the lower job security felt by Americans in 2010 and 2012 was an aberration, based upon the unusual conditions presented by the recent recession. But the reality is that the downward trend in feelings of job security has been going on for the last 35 years, apart from the “extra push” it has received from the “Great Recession,” as evidenced by my earlier ordinal comparisons of the 2010 and 2012 responses with the percentages obtained from the 17 surveys conducted from 1977 and 1978 until then.
ACKNOWLEDGMENT: The author expresses appreciation to the executive editor of the Monthly Labor Review and to economists of the Bureau of Labor Statistics for suggestions incorporated into this article.
1 Charles N. Weaver, “Worker expectations about losing and replacing their jobs,” Monthly Labor Review, April 1980, pp. 53–54.
2 Ibid., p. 53.
3 James A. Davis, Tom W. Smith, and Peter V. Marsden, General Social Survey: 1972–2012 Cumulative Codebook (Chicago: National Opinion Research Center, 2013).
5 It is useful to have a global measure of occupations that includes all occupations beyond the four most common white-collar and four most common blue-collar occupations shown in table 1. Consequently, I coded occupations into white- and blue-collar classifications from approximately 23,000 occupational titles assigned by the U. S. Census Bureau and available in the GSS. For further details, see 1970 Census of Population: Alphabetical Index of Industries and Occupations (U.S. Department of Commerce, June 1971); and 1980 Census of Population: Classified Index of Industries and Occupations (U.S. Department of Commerce, October 1980).