Abstract
The present study examined the additive effects of age discrimination, as well as the buffering effects of psychosocial resources in reducing the negative impact of age discrimination on job satisfaction, among older employed men in Japan. Data were obtained from a national survey administered in 2016 to a probability-based sample of men aged 55 to 64 years (n = 514). The results indicated that perceived age discrimination at work was associated with a lower level of job satisfaction. Moreover, a high level of social support from supervisors and coworkers decreased the negative impact of perceived age discrimination on job satisfaction. Our findings suggest that organizational-level interventions might be necessary to mitigate age discrimination and increase social support for keeping older employees in the workforce beyond their conventional retirement age.
Introduction
As global concern about ageing populations and their implications for the labor force continues to grow, researchers and policy makers are paying greater attention to older adults in the workplace. Findings of adverse effects of age discrimination against on job satisfaction of older workers were first reported by Hassell and Perrewé’s (1993) study of employees in three different industries in the United States and Orpen’s (1995) study of Australian financial service employees in the 1990s. In the 2000s, using a larger sample from the United Kingdom, Snape and Redman (2003) indicated that perceived age discrimination led to negative consequences for affective commitment to organizations. Redman and Snape (2006), in a study of police officers in the United Kingdom, suggested that social support resources may play a buffering role, reducing the negative effects of age discrimination.
In recent years, using a more sophisticated statistical model, Taylor, Mcloughlin, Meyer, and Brooke (2013) found that everyday discrimination was a direct and indirect predictor of job satisfaction and an indirect predictor of psychological well-being, using employee survey data in Australia. Although psychosocial factors were found to be significant in the study of ageing in the workplace (Macdonald & Levy, 2016), little research has explored the role of psychosocial resources (e.g., control over one’s job, social support at work) in explaining the relationship between perceived age discrimination and job satisfaction.
Ageism has been a significant social issue for nearly 50 years. Butler (1969) introduced the term “ageism” and defined it as the “systematic stereotyping of and discrimination against people because they are old just as racism and sexism accomplish this for skin colour and gender” (Butler, 1995, p. 35). Over the last half-century, the study of ageism has expanded around the world, consistent with the growing worldwide ageing population. Ageism emerges as an interrelationship among negative stereotypes, prejudices, and age discrimination that excludes or disadvantages certain people on the basis of their chronological age (McMullin & Marshall, 2001). Regarding age discrimination in the workplace, Duncan (2003) aptly states that “age prejudice differs from other forms of discrimination in that there is no single, clearly defined oppressed group. Everyone is of an age and can be subject to age discrimination” (p. 108). At the institutional level, it can be found in hiring, training, promotion, and firing. Mandatory retirement policies represent one form of age discrimination, whereby older workers are disadvantaged strictly due to their age rather than their ability. At the interpersonal level, older workers may be subjected to microaggression, which includes prejudicial attitudes and discriminatory behavior in daily social interactions (Chou & Choi, 2011).
Although age discrimination against older workers tends to attract more attention, recent Australian data have found that perceptions of age discrimination are actually in decline among older workers, and those aged 15 to 24 years are more likely to report experience of age discrimination than those aged 55 to 64 years (Taylor & Smith, 2017). Similarly, a recent study in the United Kingdom found that those in younger age groups were more likely to report age discrimination as more serious than older groups (Sweiry & Willitts, 2012).
Because of the global importance of this issue, more cross-cultural study of age discrimination in the workplace, especially research exploring the effects of age discrimination on job satisfaction in non-Western countries, is needed. The present study, conducted in Japan, makes two main contributions in this regard. First, it enables us to examine the cross-cultural validity of previous findings based on data from Western countries. We predict that East–West differences will appear, as East Asian cultures practice filial piety, according greater respect to older adults and expecting younger family members to care for them (Levy & Macdonald, 2016). However, it is possible that this type of Japanese attitude toward the elderly might have been overvalued, in what could be called the American idealization of ageing in Japan (Koyano, 1989).
Second, Japan is a unique case because it has experienced the fastest population ageing process over the last few decades of any country in world history and because it anticipates a severe workforce shortage in the near future. The percentage of elderly persons in Japan, which ranked relatively low in the 1980s and was about average in the 1990s among developed countries, had become the highest in the world by 2005 (Cabinet Office, 2016). Therefore, this study should provide useful evidence for other countries seeking to encourage older workers to continue their labor force participation beyond the conventional retirement age.
This study aims to address two research questions: Research Question 1: Does age discrimination in the workplace have additive effects on job satisfaction in Japan, even though workforce participation rates among older men are higher there than in the United States or other Western countries? Research Question 2: Do psychosocial resources (e.g., control over one’s job, social support from supervisors and coworkers) buffer the negative impacts of age discrimination in the workplace on job satisfaction?
Background
Older Workers and the Policy Context in Japan
Taylor, Loretto, Marshall, Earl, and Phillipson (2016), in their summary of the changes in labor force participation among Organisation for Economic Cooperation and Development (OECD) countries, stated that participation rates among men aged 55 to 64 years declined in some countries between 1980 and the mid-1990s, in what could be termed a tendency toward “early exit.” However, within a short period, as policy makers recognized that early retirement was unsustainable in countries with ageing populations (OECD, 2006), the thrust of public policy changed from removing older workers from the labor market to encouraging their continued participation. Consequently, since the mid-1990s, there has been a broad increase in labor force participation among older men. However, the participation rates among men aged 55 to 64 years in Japan remained almost unchanged during this time period, falling slightly from 85.4% in 1980 to 84.8% in 1995 and 83.9% in 2010. On the other hand, the participation rates among women aged 55 to 64 years in this period continued to rise from 45.3% in 1980 to 48.5% in 1995 and 53.9% in 2010. The latest participation rate in 2017 is 87.5% for men and 63.3% for women (OECD, 2018).
With regard to the retirement process of older workers, the relationship between pension policy and employment policy has been a significant social issue in Japan. Yamada and Higo (2011) stated that the government has long intervened in the labor market in response to the anticipated increase in the fiscal burden of age-related social expenditures because of an expected severe workforce shortage in the coming decade. First, the Japanese government revised its pension scheme. Until 2000, the age of eligibility for a pension was 60. The minimum age for the flat-rate portion of the pension was raised gradually until it reached 65 in 2013. Moreover, the minimum age for the earnings-related portion of the pension is currently rising at the rate of 1 year of age every 3 years until it also reaches 65 in the year 2025 (OECD, 2004). Decreasing pension benefits will be replaced by other income sources, mainly earnings from continuing employment. As Oka (2008) noted, one of the main aims of the pension reform was to extend the working age to 65 years.
Second, the government amended the Law for the Stabilization of Employment of Older Persons to deal with the upward revision of the pension eligibility age. This amendment was expected to address the problem of a growing period without a reliable income stream for older people between their mandatory retirement age and the time when they became eligible to collect their full pension (Yamada & Higo, 2011). The amendment of the law in 2004 required companies to let their employees work beyond age 60. They can do this by (a) abolishing mandatory retirement, (b) raising the mandatory retirement age, or (c) introducing a continuous employment system to ensure stable employment for their workers until age 65. However, legal loopholes, according to Oka (2008), allowed companies to be exempted from the principle of retaining all applicants if the employer and union could reach an agreement on exceptions and penalties against violation of the law were absent.
To address these problems, the 2012 amendment abolished limitations for the subjects of the continuous employment system, and it introduced a provision to announce the names of companies that do not comply with the recommendation on employment securing measures obligation. Moreover, under the amended law, employers were permitted to continue to employ employees who wished to use the continuous employment system at any of the employers’ group companies. According to a multiple-choice survey of companies with more than 50 employees, the following employment types for continuing employees were adopted: 45.8% as regular full-time employees of the same company, 68.7% as part-time workers in the same company (including contract employees), 4.7% as full-time employees of group companies, and 8.6% as part-time employees in group companies (Cabinet Office, 2017). Thus, the law was amended to set up an environment where employment could be secured until the age of 65 years. These measures to promote the employment of elderly people may lead to reductions in age discrimination in the workplace. However, if they lead to reduced recruitment of young workers, intergenerational conflict between older and younger employees may increase.
Previous Studies of Additive and Buffering Effects
The additive effects model suggests that perceived age discrimination directly impacts job satisfaction. A growing body of evidence indicates that experiencing discrimination produces negative psychological consequences for the affected individual. However, most of this work has examined discrimination based on race or gender rather than age (Redman & Snape, 2006). Although the research on age discrimination in workplaces is limited, there is evidence that perceived age discrimination negatively influences job satisfaction, commitment, and engagement for workers in many age groups (Hassell & Perrewé, 1993; Macdonald & Levy, 2016; Orpen, 1995; Redman & Snape, 2006; Snape & Redman, 2003).
In terms of the relationships between age discrimination and job satisfaction among older workers, Hassel and Perrewé (1993) found that older workers who perceive age discrimination are less satisfied than those who do not. Redman and Snape (2006) indicated that older police officers experienced a negative effect of perceived age discrimination on job satisfaction. In a Japanese study, Yamada, Sugisawa, Sugihara, and Shibata (2005) suggested that lower levels of perceived age discrimination in the workplace and the presence of employer-sponsored programs for older employees were significantly correlated with higher organizational commitment by older employees.
From a theoretical viewpoint, the stress model provides a useful perspective for understanding the psychological consequences of perceived discrimination, as this experience can be considered a stressor. When people are unable to adapt to or cope with stressors, mental health problems are likely to ensue (Yuan, 2007). We can expect this result to be reflected in an association between perceived age discrimination in the workplace and lower job satisfaction.
The buffering effects model assumes that psychosocial resources such as job control and social support at work buffer the negative impacts of age discrimination on job satisfaction. First, the job demand–control model (Karasek, 1979) has been widely used in the study of occupational stress. Job demand refers to stressors in the workplace, such as time pressure, amount of work, and conflicting direction. Job control represents the extent of the employee’s decision-making latitude regarding their tasks. The model argues that high job demand and low control over work are likely to increase psychological distress. Yuan (2007) proposed that a high sense of control buffers the relationship between perceived age discrimination and mental health, resulting in reduced psychological distress. For individuals with higher levels of the sense of control, psychological distress is increased less by perceived age discrimination than it is in those with lower levels of the sense of control. Therefore, job control can be conceptualized as a buffer that reduces the effect of perceived age discrimination on job satisfaction, whereas job demand may exacerbate this effect. However, no research has been done on the proposed synergistic effect of high job demand and perceived age discrimination.
Stress researchers have also theorized that the psychological consequences of exposure to stressors can be reduced by social support (Cohen & Wills, 1985). In the workplace context, social support from supervisors or coworkers functions as a coping resource that reduces the negative effects of age discrimination on job satisfaction.
Prior research applying the buffering model has produced mixed findings.Redman and Snape (2006) indicated no significant moderating effects of work-based social support and found the anticipated buffering effect of nonwork-based social support only on life satisfaction, with reverse buffering for job satisfaction and normative commitment. Of the two primary sources of workplace social support proposed in previous studies, that is, supervisors and coworkers (Chou & Robert, 2008; Shimazu, Shimazu, & Odahara, 2004), Chou and Choi (2011) determined that supervisor support was more essential in alleviating perceived workplace discrimination.
Based on the aforementioned theoretical and empirical considerations, we proposed the following three hypotheses for the present study: Hypothesis 1: Perceived age discrimination is associated with lower levels of job satisfaction. Hypothesis 2: Job control buffers and job demand exacerbates the effects of perceived age discrimination on job satisfaction. Hypothesis 3: Social support from supervisors and coworkers buffers the effects of perceived age discrimination on job satisfaction so that these relationships will be weaker when social support is high.
Methods
Sample
The data used in this study were obtained from a national survey of older men in Japan. Using a stratified two-stage random sampling procedure, we identified 2,500 men aged 55 to 64 years from the basic resident registration list and conducted interviews in 2016 with those men who agreed to participate. As noted earlier, the problem of securing employment until the age of 65 years is an important issue in contemporary Japan. Moreover, significantly fewer women work full-time than men, and this study was focused on ageism rather than sexism. Therefore, we focused on men from the ages of 55 to 64 years. This survey was designed for the target age, and its results were intended to be compared with the nationwide survey conducted in 1999. All interviews were conducted face to face in participants’ homes. The response rate was 36.8% (n = 919). In the present study, we used only the data from respondents who were employed at the time (n = 514). The study protocol and informed consent procedure were approved by the ethics committee of the university.
The respondents of this study were born between October 1951 and September 1961. The flat-rate portion of the pension can be received from the age of 65 years, and the earnings-related portion of the pension can be received from either 60 to 65 years. Therefore, this survey cohort is positioned as a transitional generation in pension status.
Measures
Job satisfaction
The dependent variable, job satisfaction, was measured using items representing six facets of satisfaction: (a) making good use of one’s abilities, (b) human relationships in the workplace, (c) salaries and benefits, (d) work hours, (e) job content, and (f) work environment. Response options for these items ranged from 1 (not satisfied at all) to 5 (very satisfied), and the total score was used for analysis. Higher scores indicated greater job satisfaction. The α reliability score of this scale was .83.
Age discrimination at work
Perceived age discrimination in the workplace was measured using three items, modified from the Health and Retirement Study in the United States and Kaye and Alexander (1995). These items were also used in the nationwide survey conducted in 1999 for the same age-group. Respondents were asked to indicate whether each of three types of workplace discrimination was present at their job location: (a) employees are demoted after a certain age, (b) older workers’ opinions tend to be ignored, and (c) there is a general consensus that people should leave this company before age 60. Response options for these items ranged from 0 (not true at all) to 3 (very true), and the total score was used for analysis. Higher scores indicated more severe age discrimination. The α reliability score of this scale was .67.
Job control and job demand
Job control and job demand were measured using the items from the MONICA psychosocial study, which were translated into Japanese by Uehata (1993). Job control comprised six items: (a) being required to learn new things, (b) skill level required, (c) creativity required, (d) nonrepetitive work, (e) freedom as to how to work, and (f) freedom to make decisions. Job demand comprised five items: (a) being required to work fast, (b) being required to work hard, (c) a great amount of work to be done, (d) not enough time, and (e) conflicting demands. Response options for these items ranged from 0 (not true at all) to 3 (very true), and the total score on each subscale was used for analysis. The α reliability scores were .71 for job control and .68 for job demand.
Social support from supervisors and coworkers
Social support in the workplace was measured using three items of perceived emotional and instrumental support from (a) supervisors and (b) colleagues and subordinates. Respondents were asked to indicate how much people in each of the two relationship domains (a) consulted on job-related matters, (b) helped in accomplishment of tasks, and (c) were generous to the respondents if they were late for work or had to leave work early for family or personal reasons. Response options for these items ranged from 1 (not at all) to 4 (a great deal), and the total score on each subscale was used for analysis. If a respondent did not have a supervisor or coworker, a score of 1 was assigned. The α reliability scores were .93 for supervisory support and .89 for coworkers’ support.
Sociodemographic variables
Age, education, marital status, employment status, and occupational category were included in the analyses. Age was measured in years. Education was measured in years of schooling completed. Marital status was coded as 0 (not married) or 1 (married). Current employment status was asked by full-time employees, part-time employees, contract or entrusted employees, and dispatched workers. The latter three were put together into one category and coded as 0 (part-time) or 1 (full-time). Current occupational category was classified as (a) managers and professionals, (b) clerks, (c) service and sales, or (d) manual workers by the Japan Standard Occupational Classification.
Analytical Procedures
We tested our three hypotheses using hierarchical regression models, entering the sociodemographic variables in Model 1, with age discrimination, job control, job demand, and social support variables added in Model 2. We constructed Models 3 and 4 to test our second hypothesis, that is, whether the detrimental effects of age discrimination were buffered by job control or exaggerated by job demand. Moreover, Models 5 and 6 were used to test our third hypothesis, that is, whether the detrimental effects of age discrimination were buffered by supervisor and coworker support. Following the recommendation of Aiken and West (1991), age discrimination, job control, and job demand and social support variables were centered on their means. After the effects of all other variables had been controlled for, two-way interaction terms were entered into the equations. To avoid problems associated with multicollinearity, we examined, in separate analyses, whether the effects of age discrimination were buffered by psychosocial resources. When statistical interaction terms were present, we examined simple slope analyses and plotted age discrimination along the x-axis, for groups defined according to three levels of social support—high, average, and low—such that the buffering effects would be obvious (Aiken & West, 1991). High and low social support was defined as one standard deviation (SD) above or below the mean, respectively.
Table 1 displays means, standard deviations, and correlations among the variables in the study. It was found that 75% were full-time employees (n = 387), working 20.9 days a month and 43.0 hours a week; 25% were part-time employees (n = 127), working 19.0 days a month and 33.5 hours a week. Managers and other professionals formed 47% of the sample (n = 243), clerks were 13% (n = 69), service and sales were 10% (n = 51), and manual workers were 29% (n = 151).
Means, Standard Deviations, and Correlations Among All Study Variables (n = 514).
*p < .05. **p < .01. ***p < .001.
Results
Results from the regression analysis predicting job satisfaction are shown in Table 2. Model 1, which included sociodemographic variables, revealed that managers and professionals tended to experience higher job satisfaction whereas manual workers had lower job satisfaction.
Regression Models Predicting Job Satisfaction From Age Discrimination and Psychosocial Resources.
Note. The dependent variable was job satisfaction; Ref = reference category.
*p < .05. **p < .01. ***p < .001.
With the addition of psychosocial variables in Model 2, age discrimination and job demand exerted negative effects on job satisfaction and job control, and supervisor support and coworker support had positive effects. Moreover, the differences between occupation categories became nonsignificant after these psychosocial factors were entered into the model. Comparing the two models through determination coefficients, the amount of variance in job satisfaction explained by the model increased significantly to 28%.
Table 3 shows the coefficients of the interaction effects. The cross-product interaction terms were estimated after all the factors in Model 2 had been controlled, as indicated in the analytical procedure. The Age Discrimination × Supervisor Support term and the Age Discrimination × Coworker Support term were positively significant (Models 5 and 6).
Summary of Coefficients of Regression Models Predicting Job Satisfaction From the Interaction Terms of Age Discrimination and Psychosocial Resources.
Note. The coefficients shown are from the equations after all factors in Model 2 (shown in Table 2) had been controlled. To avoid multicollinearity, we examined, in separate analyses, whether the effects of age discrimination were buffered by psychosocial resources.
*p < .05. **p < .01. ***p <.001.
Figure 1 illustrates the interaction effect of age discrimination and coworker support, which indicates a buffering effect (Model 6), for groups defined according to the level of coworker support, high (+1 SD), average, and low (–1 SD). Age discrimination had a negative relationship with job satisfaction when social support from coworkers was either average or low.
The interaction effect of age discrimination and social support from co-workers on job satisfaction in Model 6.Note. The figure is based on the results of the simple slope analysis. Age discrimination exhibits a negative correlation with job satisfaction when social support from co-workers is either average or low.
Discussion
This study examined the additive effects of age discrimination in the workplace and the buffering effects of psychosocial resources on the negative impact of age discrimination on job satisfaction among older employed men in Japan.
The Japanese government amended the Law for the Stabilization of Employment of Older Persons to deal with the upward revision of the pension eligibility age. For the cohort of this study, the problem of how to continue working until the age of 65 years, when the full pension becomes possible, is an important issue. Many companies have taken measures to secure employment until the age of 65 by raising the retirement age or introducing a continuous employment system. A comparison of the results of this study and survey results conducted for the same age-group in 1999 show that age discrimination scores have declined. This trend suggests that various continuing employment policies may have changed the age discriminatory environment of the companies that promote to demote and retire.
However, the relationship found between age discrimination and job satisfaction indicates support for Hypothesis 1, which proposed that perceived age discrimination would be associated with lower job satisfaction. These results are consistent with past research (Hassell & Perrewé, 1993; Macdonald & Levy, 2016; Redman & Snape, 2006). This additive effect was indicated even after controlling for the effects of occupation categories and other psychosocial factors. Given the amount of time that people devote to work and its importance in their lives, this result suggests that older employed men were responding negatively to the experience of age discrimination as a stressor and a form of deprivation (Snape & Redman, 2003).
Hypothesis 2 was not supported in this study. Job control did not buffer the effects of perceived age discrimination on job satisfaction, and job demand did not exacerbate them. Although one’s sense of job control buffered the relationship between perceived age discrimination and psychological distress in previous studies (Yuan, 2007), this buffering effect was not observed in the present study’s Japanese sample. However, job demand did have a negative effect on job satisfaction and job control had a positive effect. These results suggest that stressors such as time pressure, workload, and employees’ decision-making latitude are key psychosocial factors related to improving the job satisfaction of older men in Japan.
Hypothesis 3 was supported, as social support from supervisors and coworkers buffered the negative effects of perceived age discrimination on job satisfaction. In previous studies, stress researchers have investigated both the additive effects model and the buffering effects model in this regard. Regarding additive effects, supervisor and coworker support were positively related to job satisfaction, irrespective of the level of age discrimination, in this study. This finding suggests that social support at work is a strong and significant predictor of job satisfaction, consistent with prior research (Chou & Robert, 2008; Macdonald & Levy, 2016; Redman & Snape, 2006). Older employed men who perceive that they can rely on either supervisors or other coworkers for support would likely be better able to handle work-related problems and more satisfied with their work environment.
As for buffering effects, social support at work was viewed as a coping resource reducing the negative effects of age discrimination on job satisfaction. Although previous studies have produced mixed findings concerning the buffering effects model (Redman & Snape, 2006), this study found that the negative effects of age discrimination were weaker when social support at work was high. That is, older employed men with greater coping resources would be less prone to experiencing decreased job satisfaction if they encountered age discrimination. Because perceived social support at work appears to play a key role in increasing job satisfaction, organization-level management should be deployed to address and prevent age discrimination in the workplace, at least in Japan.
This study has several limitations. First, it focused only on older employed men. Although a declining overall population (as Japan is currently experiencing) generally implies a drop in the total number of workers, the number of employed women is increasing in Japan. We need to investigate gender differences related to age discrimination in the workplace, because older women may face a double social disadvantage due to their age and gender. Second, this survey measured age discrimination, which results in demotion and retirement, focusing on the discriminatory environmental aspects of companies. In future research, we intend to focus on aspects of discrimination, including everyday discrimination, such as being ignored by colleagues or being left out of a social gathering at work (Taylor et al., 2013, p. 1137). Third, in this survey, securing employment until the age of 65 years was an issue, with our population consisting of men aged 55 to 64 years. Because the sample of this survey is drawn from the transitional generation for changes in the retirement age and the pensioning age, it is unknown whether its findings can be generalized to foreign countries where there is an early retirement age and among elderly people aged 65 years or older in that country. In fact, as of 2015, 4.58 million Japanese people age 65 and over were employed, exceeding the number of workers aged 60 to 64 years (4.38 million) for the first time (Cabinet Office, 2016). Accordingly, data obtained from workers age 65 and over should be examined in future research.
Conclusion
In recent decades, Japan has experienced the fastest rate of population ageing in world history, and the country anticipates a severe workforce shortage in the near future. For older Japanese employed men, the results indicate that perceived age discrimination at work is associated with a lower level of job satisfaction. Moreover, a high level of social support from supervisors and coworkers decreases the negative impact of perceived age discrimination on job satisfaction. Our practical implication of findings is that organizational-level interventions might be necessary to reduce age discrimination and increase social support for keeping older employees in the workforce beyond their conventional retirement age.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Japan Society for the Promotion of Science Grant-in-Aid for Scientific Research (A) (No. 26245062).