Home Home | About Us | Sitemap | Contact  
  • Info For
  • Professionals
  • Students
  • Educators
  • Media
  • Search
    Powered By Google

Mean Job Satisfaction Levels Over Time: Are Things Bad and Getting Worse?

Nathan A. Bowling, Michael R. Hoepf, and David M. LaHuis, Wright State University
Lawrence R. Lepisto, Central Michigan University

Job satisfaction is a topic that has captured the interest of organizational researchers, managers, and laypeople alike. It is often reported in the popular literature that most workers are dissatisfied with their jobs and that dissatisfaction has become increasingly common in recent years (see Franco, Gibbons, & Barrington, 2010; for an opposing opinion, see SHRM, 2009). These authors have suggested, for instance, that:

  • “job satisfaction is at an all time low” (Job Satisfaction, 2010).
  • “more people out there are miserable in their jobs than fulfilled by them” (Lencioni, 2007, p. 219).
  • “In many countries throughout the world, there seems to be a kind of widespread dissatisfaction at work. . . And things seem to be getting worse” (Lama & Cutler, 2003, p. 17).
  • “Even Americans who are lucky enough to have work in this economy are becoming more unhappy with their jobs” (Americans’ Job Satisfaction, 2010).

Although statements such as these often grab news headlines, they are inconsistent with scientific theorizing about the general happiness levels of the “average person” (Cacioppo, Gardner, & Berntson, 1999; Diener & Diener, 1996). This theorizing suggests that, because positive emotion is both biologically and psychologically adaptive, evolution has produced generally high levels of happiness within the human species. Consider the potential competitive advantage of being happy. First, being happy contributes to high levels of approach motivation, thus allowing one to better capitalize on personal opportunities and to build one’s personal resources (Fredrickson & Losada, 2005). As just one example, happiness helps to strengthen existing friendships and it facilitates the development of new friendships, thus helping to nurture one’s interpersonal resources. Second, happiness may be adaptive in that it contributes to one’s ability to readily detect threatening situations (Diener & Diener, 1996). That is, a contrast effect may occur in which negative—and potentially damaging events—become more conspicuous among people who are accustomed to primarily experiencing happiness and other positive emotions. Although these mechanisms were originally offered to explain why the average person is relatively satisfied with his or her life overall, one could also interpret them as predicting the existence of relatively high mean job satisfaction levels. Furthermore, we expect that despite the occurrence of ambient work-relevant events that impact entire societies (e.g., economic downturns), the general human predisposition toward being happy causes mean job satisfaction levels to be uniformly high from one year to the next. Thus, we offer two hypotheses, which we tested using three multiwave archival datasets:

Hypothesis 1: Workers will generally report job satisfaction levels that are higher than the satisfaction scale’s midpoint.
Hypothesis 2: Mean job satisfaction levels will consistently remain higher than the scale’s mid-point from one year to the next.

Method

Participants

We used three independent archival datasets to test Hypotheses 1 and 2. These datasets were The General Social Survey (GSS; The National Data Program for the Sciences, 2006), the Wisconsin Longitudinal Study (WLS; University of Wisconsin System, 2005), and the Adult Longitudinal Panel (ALP; Lepisto, 1997). We used these particular datasets because each included multiple waves of global job satisfaction data and they included participants from a cross-section of different occupations.

GSS. The GSS (The National Data Program for the Sciences, 2006), which uses a nationally representative sample to address basic scientific questions regarding the structure and development of American society, contains a standard core of demographic, behavioral, and attitudinal questions. Many of the core questions have remained unchanged since 1972 and appear on multiple waves of the GSS, thus facilitating the examination of trends over time. The GSS assessed global job satisfaction in 26 separate waves between 1972 and 2006 using the single item: “On the whole, how satisfied are you with the work you do: Would you say you are very satisfied, moderately satisfied, a little dissatisfied, or very dissatisfied?” We coded participant responses such that high scores (4) represented high job satisfaction and low scores (1) represented low job satisfaction. The ns ranged from 944 (for the 1972 wave) to 2,338 (for the 1994 wave). The average age across all waves was 46 years old, and the sample was 54.3% female and 81.8% Caucasian.

WLS. The WLS (University of Wisconsin System, 2005), which uses a random sample of 10,317 participants who graduated from Wisconsin high schools in 1957, includes items addressing a diverse set of topics, including family relationships, educational experiences, work experiences, and physical and mental health. Study participants first completed surveys during their senior year of high school when they were 17–18 years old (in 1957) and again at ages 36 (in 1975), 53–54 (in 1992–1993), and 64–65 (in 2003–2004). Of the 10,317 original sample members, 9,139 (88.6%) were interviewed in 1975, 8,493 (82.3%) in 1992–1993, and 6,278 (61%) in 2003–2004. As of 2004, 1,297 (12.6%) of the original participants were deceased. The sample is broadly representative of older White Americans with at least a high school education (U.S. Census Bureau, 2003). On average, the sample was 51.6% female.

We focused our analysis on the years in which job satisfaction was assessed, which were three waves: 1975, 1992/1993, and 2004. We should note that the job satisfaction item was phrased slightly differently in each wave. The 1975 item was “How satisfied are you with current/last job as a whole?” the 1992/1993 item was “All things considered, how satisfied are you with your job as a whole?” and the 2004 item was “All things considered, how satisfied were you with your job as a whole, were you very satisfied?” The scale ranged from 1= very unsatisfied 4 = very satisfied. The N for each wave was 4,090.

ALP. The ALP includes waves collected in 1996, 2001, and 2006 (for a review of the ALP, see Lepisto, 1997). Although this panel primarily focuses on developmental changes in consumer attitudes and behavior, the most recent three waves included measures of job satisfaction. A total of 292 of the 1,008 individuals who responded to the T1 questionnaire held full-time employment and were thus eligible for inclusion in the current study. Of these 292 individuals, 120 remained employed throughout the duration of the study and provided reasonably complete job satisfaction data for all three waves. These 120 participants were thus used in our analyses. The average ALP participant was 48 years old at T1. Sixty-one percent were male and 93% were Caucasian.

Job satisfaction was assessed within the ALP dataset using three items from Hackman and Oldham (1980). Each item was on a seven-point scale from 1 = strongly disagree to 7 = strongly agree. A sample item is, “Generally speaking, I am very satisfied with my job.”  The internal consistency reliabilities (alphas) for the job satisfaction scale were .75 for T1, .76 for T2, and .71 for T3. 

Analyses

We tested Hypothesis 1 by comparing mean levels of job satisfaction against the neutral point of the scale. We tested Hypothesis 2 by comparing mean levels of job satisfaction across time. For the GSS, we conducted a one-way ANOVA with job satisfaction as the outcome and the year of the wave as the independent variables because it was not possible to reliably link job satisfaction scores to particular participants across different years. For the WLS and ALP samples, we conducted repeated measures ANOVAs. We focused on the effect sizes (η2) because the large sample sizes could cause practically insignificant effects to be statistically significant. For each of the three datasets, we also plotted mean job satisfaction by year, thus creating a linear trend line. The results of these analyses are reported below.

Results

Tables 1, 2, and 3 present the means and standard deviations for job satisfaction across time for the GSS, WLS, and ALP, respectively. To test Hypothesis 1, which predicted that workers would generally report job satisfaction levels that were higher than the satisfaction scale’s midpoint, we compared the mean level of job satisfaction against the neutral point of the scale. For both the GSS and WLS samples, the “neutral’ score was 2.5. The “neutral” score for the ALP data was 4. In general, most of the mean job satisfaction scores were nearly one standard deviation unit above the neutral points of their respective scales. In addition, the means for the GSS and WLS samples generally fell in between the fairly satisfied (3) and very satisfied (4) scale points. Taken together, these results support Hypothesis 1.

Table 1
Means, Standard Deviations and Standardized d Scores for the GSS Sample    

Year M SD N Standardized d
1972 3.30 0.80 944 1.00
1973 3.33 0.80 1,141 1.03
1974 3.29 0.83 1,223 0.95
1975 3.37 0.81 1,165 1.07
1976 3.34 0.82 1,185 1.02
1977 3.32 0.77 1,262 1.06
1978 3.33 0.82 1,280 1.02
1980 3.25 0.85 1,246 0.88
1982 3.26 0.85 1,224 0.89
1983 3.32 0.81 1,333 1.00
1984 3.20 0.91 1,208 0.77
1985 3.29 0.82 1,235 0.96
1986 3.35 0.75 1,162 1.13
1987 3.22 0.84 1,165 0.85
1988 3.29 0.80 1,153 0.99
1989 3.27 0.82 1,206 0.95
1990 3.28 0.80 1,041 0.99
1991 3.26 0.80 1,149 0.96
1993 3.23 0.82 1,228 0.89
1994 3.29 0.79 2,338 1.00
1996 3.25 0.82 2,313 0.92
1998 3.31 0.79 2,216 1.03
2000 3.30 0.77 2,162 1.03
2002 3.33 0.80 1,061 1.04
2004 3.32 0.82 1,397 1.01
2006 3.32 0.80 2,177 1.03
Total 3.29 0.81 36,214 0.98

Note. Standardized d scores were computed against the “neutral” score of 2.5. 

Table 2
Means, Standard Deviations and Standardized d Scores for the WLS Sample    

Year M SD Standardized d
1975 3.50 0.66 1.52
1992 3.42 0.69 1.33
2004 3.51 0.69 1.46

Note. N = 4,090. Standardized d scores were computed against the “neutral” score of 2.5.
 

Table 3
Means, Standard Deviations and Standardized d Scores for the ALP Sample    

Year M SD Standardized d
1996 5.22 1.11 1.10
2001 5.21 1.24 0.98
2006 5.50 1.10 1.36

Note. N = 120. Standardized d scores were computed against the “neutral” score of 4. 

Figures 1, 2 and 3 plot job satisfaction means as a function of time for the GSS, WLS, and ALP, respectively. Together, these three figures suggest that within each of the archival datasets, job satisfaction scores were relatively high, and there appeared to be no systematic increases or decreases in job satisfaction over time. We tested for mean differences using ANOVAs. The ANOVA results for the GSS data indicate that the job satisfaction mean did vary across time (F(25, 36188) = 3.94, p < .01), but effect size was very small (η2= .002). In addition, as shown in Table 1, the mean job satisfaction scores ranged from 3.20 to 3.37. This suggests that there was not much change in mean job satisfaction levels over time. Similar results were found for the WLS sample (see Table 2). The repeated measures ANOVA revealed significant differences (F (2, 8178) = 26.52, p < .01), but the effect size was again very small (η2= .014). The means ranged from 3.42 to 3.51. Finally, the results for the ALP data were consistent with the other two data sets (see Table 3). Although the repeated measures ANOVA suggested significant mean differences (F(2, 238) = 4.25, p = .02), the effect size (η2= .034) suggested that effect was small. In addition, the means ranged from 5.21 to 5.50. As a whole, these results provide support for Hypothesis 2.

Discussion

Analyses conducted using three independent multiwave datasets supported our predictions that mean job satisfaction levels would be relatively high (Hypothesis 1) and would not differ systematically across time (Hypothesis 2). Specifically, across the datasets, mean job satisfaction scores were roughly half-way between the neutral point of the scale and the scale point reflecting the high possible level of satisfaction. Furthermore, although job satisfaction varied to a small extent from wave to wave, these changes appear to be random fluctuations rather than systematic increases or decreases. It is of note that these findings are contrary to the assumptions of many laypeople, which are illustrated by the quotes at the beginning of the current paper.

Theoretical and Practical Implications

The finding that mean job satisfaction levels are uniformly high from decade to decade has important implications: It suggests that workers are highly adept at adjusting to their environments (see Bowling, Beehr, Wagner, & Libkuman, 2005; Landy, 1978). Over the course of the three longitudinal archival studies, several changes have occurred within the world of work. These changes, which include countless technical innovations, increases in the average American’s wealth, general economic upturns and downturns, changes in employment law, and societal changes, such as women’s increased role within the workplace, have drastically transformed the nature of the workplace. Although one might have expected that mean job satisfaction levels would have changed as a result, this was clearly not the case. In short, workers as a whole seem to have adapted to these environmental changes.

The possibility that people generally adjust to large-scale environmental conditions has particular relevance to today’s workers, given recent global economic conditions. On one hand, many people may assume that the generally negative mood produced by the recent economic recession may “spill over” into employees’ attitudes toward their jobs. Furthermore, the recent recession could result in negative changes to one’s work environment, such as increased workloads and pay cuts, which in turn contribute to job dissatisfaction. Such effects, however, might be largely counteracted by a competing process in which many workers may be more satisfied with their jobs during an economic downturn because they are grateful to simply be employed (Agell & Lundborg, 1995; Akerlof & Yellen, 1990).   

Limitations and Future Research

We should note two limitations of this research. First, the GSS and WLS each utilized single-item job satisfaction measures. Although single-item measures in general have been the target of some criticism, previous research suggests that job satisfaction can be effectively assessed with a single item (Wanous, Reichers, & Hudy, 1997).         

Second, all of the data used in the current research were collected from American workers. Future research should attempt to replicate our findings within non-US samples. Such research might be especially insightful if conducted within nations that have experienced considerable cultural, political, or economic change. Although the human predisposition toward generally high levels of happiness (Cacioppo et al., 1999; Diener & Diener, 1996) may typically cause mean job satisfaction levels to remain more or less stable from year to year, it remains to be seen if job satisfaction can remain stable in the face of fundamental society-level changes, such as a transition from a communist to a capitalist economy or from a dictatorial government to a democracy.   

References

Agell, J., & Lundborg, P. (1995). Fair wages in the open economy. Economica, 62, 335–351.
Akerlof, G. A., & Yellen, J. L. (1990). The fare wage-effort hypothesis and unemployment. The Quarterly Journal of Economics, 105, 255–283.
Americans’ job satisfaction falls to record low. (2010). Retrieved from
http://www.usatoday.com/ money/workplace/2010-01-05-job-satisfaction-use_N.htm
Bowling, N. A., Beehr, T. A., Wagner, S. H., & Libkuman, T. M.  (2005). Adaptation-        level theory, opponent process theory, and dispositions: An integrated approach to the stability of job satisfaction. Journal of Applied Psychology, 90, 1044–1053.
Cacioppo, J. T., Gardner, W. L., & Berntson, G. G. (1999). The affect system has parallel and integrative processing components: Form follows function. Journal of Personality and Social Psychology, 76, 839–855.
Diener, E. & Diener, C. (1996). Most people are happy. Psychological Science, 7, 181–185
Franco, L., Gibbons, J., & Barrington, L. (2010). I can’t get no ... job satisfaction, that is: America’s unhappy workers. (Research Report No. 1459-09-RR). New York, NY: The Conference Board.  
Fredrickson, B. L., & Losada, M. F. (2005). Positive affect and the complex dynamics of human flourishing. American Psychologist, 60, 678–686.
Hackman, J. R., & Oldham, G. R. (1980). Work redesign. Reading, MA: Addison-Wesley.
Job Satisfaction Is at an All Time Low. We’re Shocked. Shocked. (2010). Retrieved from http://www.onlineinvestingai.com/blog/2010/01/12/job-satisfaction-is-at-an-all-time-low-       were-shocked-shocked/
Lama, D., & Cutler, H. (2003). The art of happiness at work. New York, NY: Riverbead.
Landy, F. J. (1978).  An opponent process theory of job satisfaction.  Journal of Applied Psychology, 63, 533–547.
Lencioni, P. (2007). The three signs of a miserable job. San Francisco, CA: Jossey-Bass.
Lepisto, L. (1997). The adult longitudinal panel: a research program to study the aging process and its effects on consumers across the life span. In L. Kahle & L. Chiagouris (Eds.), Values, lifestyles, and psychographics (pp. 317–333). Mahwah, NJ: Erlbaum.
Society for Human Resource Management. (2009). 2009 Employee job satisfaction: Understanding the factors that make work gratifying. Retrieved from http://www.shrm.org/ Research/SurveyFindings/Articles/Documents/09-0282_Emp_Job_Sat_Survey_FINAL.pdf
The National Data Program for the Sciences. (2006). General Social Survey (GSS), 1972-2006, Variable satjob (job or housework) [Data file]. Retrieved from http://www3.norc.org/ GSS+Website/Browse+GSS+Variables/Subject+Index/
University of Wisconsin System. (2005). Wisconsin Longitudinal Data (WLS), 1975-2005, custom dataset (variables jbcrsa, rg044jjc, and gg044jjc) [Data file]. Retrieved from http://www.ssc.wisc.edu/wlsresearch/data/subset/
U.S. Census Bureau. (2003). Current population survey: Annual social and economic supplement. Washington, DC: Author.
Wanous, J. P., Reichers, A. E., & Hudy, M. J. (1997). Overall job satisfaction: How good are single-item measures? Journal of Applied Psychology, 82, 247–252.