1
Age discrimination, job satisfaction and motivation of older
employees
Bachelor Thesis written by: Romy Janssens Student number: 10335560
Economics and Business Finance and Organization Supervisor: Sabina Albrecht 29-06-2015
2
Statement of Originality
This document is written by Romy Janssens who declares to take full responsibility for the contents of this document. I declare that the text and work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Abstract
The purpose of this thesis is to examine the effect of age discrimination on older employees’ job satisfaction and motivation to work. The consequences of age discrimination in the workplace were analysed by the use of self-reports of discrimination from respondents in the National Longitudinal Survey of Older Men. Workers with positive reports were more likely to have lower job satisfaction than workers who reported no age discrimination. In addition to that, the findings suggest that unemployed men are less likely to accept new job offers when they have been discriminated against due to their age. However, there was no evidence found that age discrimination has an effect on the retirement behaviour of older employees.
Moreover, their motivation to remain in the workforce when they reached the compulsory age of retirement appeared not to be affected by any experience of age discrimination.
4 Table of contents 1. Introduction...5 2. Literature...6 3. Data...9 4. Methodology...11 5. Results...14 6. Discussion...21 7. Conclusion...21 References...23 Appendices...25-27 - Appendix A...25 - Appendix B...26 - Appendix C...27
5
1. Introduction
There has been a substantial increase in the number of complaints of age
discrimination in the workplace (hereinafter, simply age discrimination) in recent years. For example, in 2008 the U.S. Equal Employment Opportunity Commission received 24,582 complaints from workers who felt they had been discriminated against because of their age, while ten years earlier; there were only 15,191 (EEOC, 2016). Due to the ‘baby boom’ after World War II, increases in longevity, and the low birth rates nowadays, older people aged fifty and over will play an important role in society. Specifically, the number of older people is expected to grow by more than 15% the coming years.
Furthermore, the compulsory age of retirement in the United States will rise to 67 years in 2027. This governmental measure is a response to numerous consequences of the aging population; for example, the cost of pensions and health service costs. This increase in retirement age will ensure the presence of older people in the workforce. Therefore,
organizations need to be successful in the retention of older workers and they will have to rely increasingly on the skills and experience of older workers and pay more attention to keeping aging workers motivated.
The Age Discrimination in Employment Act of 1967 was enacted to protect
individuals who are 40 or older from employment discrimination based on age. Despite this law, complaints filed by older workers and public stories about individuals who have been replaced by younger workers or older workers who were forced to leave by mandatory retirement provisions are signals that people are still facing age discrimination. However, the Age Discrimination in Employment Act cannot be fairly evaluated before the effect of age discrimination in the labour market is defined.
To analyse the impact of age discrimination on people’s behaviour it will be relevant to study older workers’ job satisfaction and motivation to continue working. Briefly, this will give a better understanding of whether these workers really like their jobs, or whether they are only motivated by the wages. According to the two-factory theory proposed by Herzberg et al. (1957), companies can achieve higher levels of performance and motivation by maximizing the satisfaction of their employees. The combination of high hygiene factors (salary, status, conditions etc.) and high motivation factors (challenging work, responsibility, importance etc.) result in the ideal situation where employees are highly motivated. Herzberg argued that job enrichment is required for intrinsic motivation and that jobs should present sufficient challenge to utilize the full ability of the employee, otherwise there will be a motivation
6
problem (Herzberg et al., 1957).
Therefore, this thesis examines the effect of age discrimination on the motivation and job satisfaction of older individuals in the workforce, as well as the attitudes of employees towards retirement. In order to analyse these impacts, self-reports of age discrimination were used. Individual responses were compared at different points in time, together with the use of other valuable information available in the data. This thesis proposes the hypothesis that older employers who claim they have experienced discrimination at work will have other outcomes in retirement behaviour, job satisfaction, and other attitudes. Results suggest that age
discrimination reduces job satisfaction and reduces the probability of older men accepting another job offer.
Some correlational studies have described the relationship between age and job satisfaction and motivation. Results imply that a successful contribution from employers to the motivation and retention of older workers is creating attractive jobs (Wang and Shultz, 2010), and older employees seem more likely to need intrinsically challenging and fulfilling jobs to stay motivated (Boumans et al., 2011). Furthermore, the company needs to take into account what kind of provisions they will offer and, more importantly, how employees stay engaged with the firm, even when they are approaching retirement age (Schalk and Van Veldhoven, 2010).
The presence of age discrimination could affect the intrinsic motivation of employees, but there has been little empirical research done on the impact of age discrimination on these attitudes. The difficulty in measuring age discrimination could be a reason for this lack of attention. Besides, investigating only employees’ earnings would not be an appropriate approach for the analysis of age discrimination, since aging is likely to affect workers’ productivity. Therefore, this thesis uses the self-reports of individuals to derive the possible consequences.
The thesis is structured as follows: Section 2 gives background information on aging people in the workforce, employee job satisfaction and motivation, age discrimination, and retirement behaviour. Section 3 presents an overview of the dataset, and Section 4 describes the hypothesis and model used. Section 5 presents results, Section 6 discusses the
implications of the results, and, finally, Section 7 presents the conclusion of the thesis.
2. Literature
7
and over. This definition is based on chronological age. Other conceptualizations of age are based on physiological, functional, organizational, or life span factors (De Lange et al., 2006). However, there is a great variety in the age category in which these workers place themselves based on these other conceptualizations. For example, some 55-year-old workers report that they feel, act, and behave much younger than their chronological age.
Some studies have described the impact of the increasing age of people on the workforce. Schalk and van Veldhoven (2010) suggested that ageing must be seen as a
multidimensional development that has positive as well as negative implications. Companies need to focus on individual preferences since these are different between generations. They should be aware of the need for more provision to older employees of training that matches their needs, and management needs to take into account what motivates and satisfies older workers in their jobs.
Furthermore, there is evidence that older workers might sometimes be more effective at work but less adaptable to change than their younger colleagues (Chiu et al., 2001). However, Stamov-Rossnagel and Hertel (2010) suggest that the motivation of older workers develops in a multidirectional and multilevel way. For example, as a function of a variety of task characteristics, the motivation of older workers declines for certain types of work tasks, despite motivation gains in other tasks. These age-related changes may be captured in a worker’s motivation profile, which is useful for positive affect regulation.
The relationship between age and job satisfaction and motivation has been described in several studies. Many studies have shown that the motivation to work and job satisfaction differ between younger and older employees. For example, Warr’s study recorded
employment-related values and particular job characteristics. Warr (2008) found that the age of the employees is significant, and that older workers report higher job satisfaction than their younger colleagues. Furthermore, older workers assess particular job characteristics as
significantly less important than younger workers. However, a recent study by Claes and van de Ven (2008) revealed that, across age groups, skill discretion and organizational fairness predicted both job satisfaction and organizational commitment. They found a negative impact of job insecurity on both job satisfaction and organizational commitment for older workers. Furthermore, Boumans et al. (2011) studied the influence of age on the relationship between work characteristics and workers’ work motivation and job satisfaction. Their findings also suggest that motivation to work is much stronger for younger than for older employees. To remain motivated, older employees seem more in need of intrinsically
8
challenging and fulfilling jobs while younger workers’ motivation increases as they are offered more career opportunities. To sum up, several studies described the relationship between age and job satisfaction and motivation.
There have been few empirical investigations into the effects of age discrimination on the behaviour of older employees. The prevalence and consequences of age discrimination in the workplace are argued by Johnson and Neumark (1997). Their study investigated the effect of age discrimination on job separations and the employment status of older workers. They found evidence that workers who reported experiencing age discrimination are more than twice as likely to separate from their employer and less likely to remain employed than workers who report no age discrimination. This estimated effect of reported discrimination remains large and significant even when controlling for the existence of mandatory retirement provisions on the current job.
Findings from Chiu et al. (2001) suggest that anti-age discrimination policies in organizations have a positive impact on employers’ beliefs about the adaptability of older workers. Moreover, this possibly affects their attitudes towards the training, promotion and retention of older workers. However, Lahey (2008) found evidence that age-discrimination laws have only slight effects on older workers. Lahey’s findings suggest that firms are not employing older workers in the first place in an anti-age discrimination environment.
Other related literature focuses on the attitudes of older employees toward retirement. There are different drivers for employees to stay in the workforce after they reach the statutory retirement age. For example, some workers are motivated by the fact that they are free from family commitments and therefore can work more hours and be more flexible, while others see financials as their main driver (Loretto and White, 2006).
However, in recent years, many organizations implemented some retirement practices that stimulated the exit of members of the workforce even before they had reached retirement age. Kooij et al. (2008) suggest the lack of a common framework for retirement behaviour. There has been some evidence found that, in general, an individual with poor health and high financial security is more likely to retire earlier. Furthermore, it can be concluded that age-related factors can have a negative impact on the motivation of older people to continue to work.
To sum up, more research is needed to understand the relationship between age discrimination and job perceptions. In the first place, it is necessary to assess the effect of age discrimination as this might play a more important role in the future aging workforce. Second,
9
analysing worker’s attitudes is needed because these might be crucial drivers according their willingness to continue work, accept another job offer, or go for retirement.
3. Data
The database used in this research is the National Longitudinal Survey of Older and Young Men (NLSOM). This two-cohort survey is part of the NLS Original Cohort project sponsored by the U.S. Bureau of Labor Statistics. The older men’s cohort used includes 5,020 men born between 1906 and 1921. The individuals in the samples were randomly chosen in selected areas and therefore they are representative of the male population in the United States. At the time of the first interview in 1966, the respondents’ ages ranged from 45 to 59. The respondents were surveyed annually between 1966 and 1969. After that, they were interviewed three years out of every five until 1983, when respondents were aged 62 to 76 years. In 1990, a final interview was conducted, and here the men ranged in age from 69 to 83 years. However, in this last period only 41.7% of the original cohort was interviewed. The data consist of information about work experience, education, household, family background, marital status, income, health, and attitudes.
The respondents were asked, in 1971, 1976, and 1981, if they, as far as work is concerned, had been discriminated against because of their age in the last five years. Therefore, this thesis gathered the information from these particular years to measure the effect of this main explanatory variable. However, the focus is primarily on 1971, since this year includes most of the observations. The outcome variable job satisfaction is measured as follows: respondents answered (1) like their job very much, (2) like it fairly well, (3)
somewhat dislike it, and (4) dislike it very much. In Table I, the results for older employees who reported lower job satisfaction, in other words categories 3 and 4, are summarized.
Table I (1971)
Observations Low Job Satisfaction Percentage
Overall 3,446 248 7.20%
Do have experience
with age discrimination 440 75 17.05%
Do not have experience
with age discrimination 3,006 173 5.76%
The t-statistics on the difference between older men who reported age discrimination and those who did not is 5.66, and therefore significant at a confidence level of 1%. This difference remains significant in the other years, 1976 and 1981 (see Appendix A).
10
Further, the unemployed respondents were asked if they would accept a job offer, and answered (1) would accept and (0) would not accept. In Table II the results for the older men who might accept another job offer are summarized. The percentage that would accept a job offer is clearly higher for unemployed men that did report being discriminated against.
Table II (1971)
Observations Accept job offer Percentage
Overall 378 85 22.49%
Do have experience
with age discrimination 44 18 40.91%
Do not have experience
with age discrimination 334 67 20.06%
Furthermore, respondents were asked if they expect to retire before the compulsory
retirement age, where they choose (1) do expect to or (0) do not expect to. In Table III, age
discrimination results for the men who expect to retire before the compulsory retirement age are summarized.
Table III (1971)
Observations Expect to retire Percentage
Overall 970 627 64.64%
Do have experience
with age discrimination 123 75 60.1%
Do not have experience
with age discrimination 847 552 65.17%
In this case, the t-statistics on the difference between older men who reported age discrimination and those who did not is 0.43, and therefore not significant at a confidence level of 5%. The respondents were also asked the opposite question: if they would like to
work past the compulsory retirement age. This was answered by (1) would like to and (0) would not like to. Again, the difference between those with an experience of age
discrimination and those without is not significant.
Further, respondents were asked, in the first survey year, about their motivation to
work. Specifically, what employees find to be the most important thing about any job. This
was answered with (1) good wages or (2) liking the work. Table IV summarizes the kinds of motivation indicated by the respondents.
11 Table IV (1966) Observations Percentage Overall 4,445 100% Good wages 1,247 28.05%
Liking the work 3,198 71.95%
Finally, the last survey, in 1990, collected information about the overall job
satisfaction of individuals over their entire work life. The remaining 1,917 respondents’
answers are classified as follows: (1) completely satisfied, (2) very satisfied, (3) somewhat
satisfied, (4) not very satisfied, and (5) not at all satisfied. Of all the respondents, only 3.65%
reported a low overall job satisfaction (category 4 or 5), and just 6.88% of them had experienced age discrimination.
Other information collected from the database concerns respondents’ wages, marital
status, employment status, and age during the survey. Wages represents the total income in dollars the respondents received from wages, salary, commissions or tips from all jobs over the year in question. Marital status is scaled from 1 to 6: (1) married, spouse present, (2)
married and spouse absent, (3) widowed, (4) divorced, (5) separated, and (6) never married. Employment status, is scaled from 1 to 8: (1) working (2) not at work (3) unemployed (4) at school, (5) retired, (6) unable to work, (7) never able to work and (8) other. The first three
categories represent individuals in the labour force. Finally, the respondent’s age is included for every survey year examined in this thesis.
4. Methodology
This chapter will explain the hypotheses and the empirical approach used to derive an answer to the research question. The basic empirical strategy of this thesis is to analyse the employment behaviour of older male workers following a self-report of age discrimination. In this way it can be determined whether or not those who report age discrimination are
subsequently more likely either to dislike their current work or have incentives to quit this job than otherwise similar workers.
Firstly, this thesis will examine whether older male worker’s yearly job satisfaction is affected by experiencing age discrimination. I expect that older men who have experienced age discrimination will report lower job satisfaction. If older employees face unequal
opportunities (for example, fewer promotions, training, pay, or job assignments compared to younger colleagues), this will affect their job satisfaction.
12
H0: The job satisfaction of older male workers is not affected by age discrimination.
Secondly, this paper will consider several attitudes of older male workers to discover if their motivation to work is affected by discrimination on the basis of age. One such attitude is unemployed older male workers’ attitudes toward a new job offer. I expect that these men who have experienced discrimination in the past will be inclined to accept a new job offer more often than who did not have this experience. One who is unemployed will be more likely to accept the offer of a new job since this will probably provide confidence and therefore, more motivation to enter the workforce again. Therefore, the second hypothesis is
H0: The likelihood that an unemployed older male worker will accept a job offer is not
affected by an experience of age discrimination.
Furthermore, I expect that older workers, who reported experience in age
discrimination during their work, will retire more often before the compulsory retirement age compared with employees who did not have these experiences. Therefore, the third hypothesis is
H0: Older male workers’ retirement behavior is not affected by experiences of age
discrimination.
Furthermore, this attitude can be analysed from the other perspective; namely, whether older men are likely to work longer than the compulsory retirement age. I expect that age discrimination will have an effect on the behaviour of employees in staying or leaving the workforce after they reach the compulsory age of retirement. This leads to the fourth hypothesis,
H0: Older male workers’ decision to stay in the work force after the age of retirement is reached is not affected by experiences of age discrimination.
Finally, besides the yearly job satisfaction this thesis will also measure the effect of age discrimination on the total satisfaction of employees’ working life. My expectation is that
13
overall job satisfaction will be affected by any experience of age discrimination during your career. Furthermore, I expect that for employees that are motivated to work when they like their job, age discrimination will have significantly more effect on their job satisfaction. Therefore, the last hypothesis is
H0: The overall job satisfaction of older male workers is not affected by any
experience in age discrimination during their career.
With these hypotheses I will be able to come up with an answer to the question of whether age discrimination has an effect on the employment attitudes and behaviour of older male workers.
Model
To test the hypotheses I used two different types of regressions. First, the probit model was used in cases where the dependent variable is binary, since it only has two possible outcomes. In other words, this method was used for analysing job offer and retirement behaviour of the employees. The main explanatory variable is whether the man has experienced age
discrimination, and I will control for: age, wage, marital status, and employment status. The outcomes are represented by a binary indicator variable Yi as follows:
Yi = 1 if Yi* > 0 and probability Pr(Yi=1) = Pr(Yi*> 0) = Pr(xiTβ+εi > 0)
Yi = 0 if Yi* ≤ 0 and probability Pr(Yi=0) = Pr(Yi*≤ 0) = Pr(xiTβ+εi ≤ 0)
Probit is based on a latent model:
= Pr(Yi= 1| x) = P(Yi*> 0 | x)
= Pr(xiβ + εi > 0 | x)
= Pr(εi > - xiβ | x)
= 1- F (-xiβ)
The error terms are assumed to be independent and normally distributed. By symmetry of the standard normal distribution the model takes the following form:
Pr(Yi=1 | x) = Pr(Yi* > 0) = Φ(xiTβ) and Pr(Yi=0 | x) = Pr(Yi* ≤ 0) = 1- Φ(xiTβ),
where Φ indicates the cumulative distribution function of the standard normal distribution. This model cannot be consistently estimated using Ordinary Least Squares so instead parameters β were estimated by the maximum likelihood theory according to Stock and Watson (2011).
14
Secondly, I used ordered probit to analyse employees’ job satisfaction, since this dependent variable is divided into more than two categories. Job satisfaction over the year is classified into four ordered categories from 1 to 4, and the overall job satisfaction of an employee’s career in ordered categories from 1 to 5, where the highest number implies the lowest job satisfaction. The ordered probit model takes the following form:
yi*= xi'β + εi,
where yi* is a latent measure of job satisfaction, a linear combination of some predictors x,
plus an error term εi, which is assumed to be normally distributed. For example, job
satisfaction measure per year takes on values according to the following scheme: yi = 1 if yi* ≤ µ1 (Men who like their job very much)
yi = 2 if µ1 < yi* ≤ µ2 (Men who like their job fairly well)
yi = 3 if µ2 < yi* ≤ µ3 (Men who somewhat dislike their job)
yi = 4 if µ3< yi* ≤ µ4 (Men who dislike their job very much)
By maximizing the log-likelihood function, the parameters β and thresholds µ can be jointly estimated. The likelihood function takes the following form:
Pr(yi = j) = Φ(µj - x'iβ) – Φ (µj-1 - x'iβ)
The outcomes can only be interpreted by their sign. Parameter β denotes the influence of variation in response to variables on the principal scale. In this case the positive sign of parameter β implies lower job satisfaction as the value of related variables increases. Furthermore, the marginal effects will be calculated to examine the effect of the independent variable on the probability of the dependent variable being in a particular category. This ordered probit model is used to test the first hypothesis about yearly job satisfaction, and the same explanatory and control variables are included as discussed above. To measure the effect of discrimination on employees’ motivation to work, this dependent variable is included when testing the final hypothesis about the overall job satisfaction of the older male workers.
5. Results
This chapter will present the results and their interpretations, and is divided into three parts. First, I will present and discuss the regressions of yearly job satisfaction of older male workers. Thereafter, the results for the older male workers’ retirement behaviour will be discussed. Finally, the outcomes for individuals’ overall job satisfaction are presented.
15
Job Satisfaction
Table V presents the results from the ordered probit regression of job satisfaction for the years 1971, 1976, and 1981. The employment and marital status of the respondents are classified in different categories but do not follow a quasi-continuous ordering. Therefore, these variables are included as dummies in the model. The outcomes married and employed are serving as bases in this regression.
Table V
Ordered probit regression of job satisfaction per year 1
Variables 1971 (2)2 1976 (2)2 1981 (2)2 Age discrimination3 0.4084*** 0.4308*** 0.3380*** 0.3542*** 0.3196*** 0.3508*** (0.0893) (0.0878) (0.0908) (0.8861) (0.1142) (0.1114) Age -0.0213*** -0.0170*** -0.0216*** -0.0141* ** -0.0454*** -0.0401*** (0.0050) (0.0049) (0.0076) (0.0071) (0.0110) (0.0110) Employment status Not at work 0.1816** 0.1597** 0.1084 0.1187 0.1675 0.1290 (0.0657) (0.0647) (0.0843) (0.0820) (0.1253) (0.1206) Marital status Married4 -0.2829 -0.2521 -0.3963 -0.5352 -0.4008 -0.3443 (0.2200) (0.2147) (0.4970) (0.3881) (0.7699) (0.1235) Widowed 0.2424** 0.2446* * 0.1566 0.2023 0.2214 0.2196 (0.1220) (0.1185) (0.1293) (0.1248) (0.1711) (0.1663) Divorced -0.0398 0.0036 -0.0607 -0.0404 0.2207 0.2022 (0.1247) (0.1210) (0.1460) (0.1437) (0.1972) (0.1951) Separated 0.1413 0.1390 -0.3519 -0.3595 0.0790 0.0778 (0.1256) (0.1228) (0.2278) (0.2257) (0.3722) (0.3708) Never married 0.1244 0.1718 0.0183 0.0229 -0.2606 -0.2716 (0.1137) (0.1120) (0.1667) (0.1639) (0.2143) (0.2122) Wage -0.00002*** -0.0000** -0.00000 (0.00000) (0.00000) (0.00000) Observations 3,075 3,172 1,770 1846 885 934 Chi-2 Test 77.69 28.07 30.18 28.07 32.53 33.58 Prob > chi2 0.0000 0.0000 0.0004 0.0005 0.0003 0.0002 Pseudo-R2 0.0123 0.0084 0.0095 0.0085 0.0210 0.0095
1 Question: ‘How do you feel about the job you have now?’ 2 Wages are omitted in column (2)
3 Age discrimination
- 1971 includes experiences in discrimination from 1966 to 1971 - 1976 includes experiences in discrimination from 1966 to 1976 - 1981 includes experiences in discrimination from 1966 to 1981
4 Married but living apart from spouse * Significant at confidence level of 10% ** Significant at confidence level of 5% *** Significant at confidence level of 1%
For all three years, the variable age discrimination is significant at a confidence level of 1%. The age discrimination coefficients are positive, which implies that on average
16
discrimination ensures higher values for job satisfaction. In other words, on average, age discrimination reduces the job satisfaction of employees. Furthermore, betas of the variables age and wage are significant, which implies that these have an influence on job satisfaction as well. The coefficients of age are negative, which indicates that as age increases, employees report higher levels of job satisfaction. This result is in line with earlier findings from Warr (2008), that found older workers, in general, reporting higher levels of job satisfaction compared to their younger colleagues.
Since the results shown in Table V can only be interpreted by their sign, I present the marginal effects of the main explanatory variable age discrimination in Table VI. The results in this table indicate the precise marginal probability of being in the higher numbered categories of job satisfaction after experiences of age discrimination; in other words, men who somewhat dislike their job and dislike their job very much. The variables wage and age are held at their means, and the categorical variables marital and employment at their first categories, which include men who are married and employed.
Table VI
Marginal effect on job satisfaction (1971)
Category Men who somewhat Men who dislike
dislike their job (3) their job very much (4)
Age discrimination1 0.0378*** 0.0141***
(0.0086) (0.0035)
Age: 55.97 (mean) Wage: 7,548.5 (mean) Employment status: Employed Marital status: Married Observations: 3,075
1 Age discrimination: includes experiences in discrimination from 1966 to 1971 * Significant at confidence level of 10%
** Significant at confidence level of 5% *** Significant at confidence level of 1%
The results can be interpreted as follows: an employed and married man at the age of 55 years with a yearly wage of $7,548.5 is 1.41% more likely to report the lowest level of job satisfaction after experiencing age discrimination. Therefore, age discrimination has a negative effect on employees’ job satisfaction. The marginal effects on job satisfaction for the other years, 1976 and 1981, are significant as well (see Appendix B).
Retirement and employment behaviour
Table VII presents the results from the probit regression of both the retirement and employment behaviour of older male workers over the year 1971. The table is divided into
17
three columns, based on the respondents’ attitudes toward early retirement, late retirement and job offers. The questions related to these attitudes are described at the bottom of the table.
Table VII
Probit regressions of retirement behaviour (1971)
Variables Early retirement1 Late retirement2 Job offers3
Age discrimination4 0.0813 -0.0288 -0.6553*** (0.1893) (0.1671) (0.2270) Age -0.0196 -0.0325*** 0.0150 (0.0109) (0.0096) (0.0216) Employment status Not at work -0.0306 -0.2452*** - (0.1200) (0.1167) Retired - - 0.5897 (0.9465) Unable to work - - 1.1930 (0.9568) Marital status Married5 0.4288 0.3787 -0.3326 (0.4834) (0.3311) (0.7056) Widowed 0.1224 -0.0534 -0.2880 (0.3117) (0.2831) (0.2941) Divorced 0.0293 0.4335 - (0.2987) (0.2294) Separated 0.4659 0.2969 -0.5209) (0.3404) (0.2513) (0.4495) Never married -0.1172 -0.1197 -0.0161 (0.2830) (0.2744) (0.3474) Wage 0.0000*** 0.0000*** 0.0000 (0.0000) (0.0000) (0.0000) Constant 0.9968 -2.4197 -0.5253 Observations 857 1,175 323 Chi-2 Test 27.25 21.93 41.89 Prob > chi2 0.0013 0.0091 0.0000 Pseudo R2 0.0246 0.0168 0.1166
1 Question: ‘Does respondent expect to retire before compulsory retirement age? 2 Question: Would respondent like to work longer than compulsory retirement age? 3 Question: ‘If you were offered a job by some employer, do you think you would take it?’ 4 Age discrimination: includes experiences in discrimination from 1966 to 1971 5 Married but living apart from spouse
* Significant at confidence level of 10% ** Significant at confidence level of 5% *** Significant at confidence level of 1%
For both types of retirement behaviour, (retire earlier or later than the compulsory retirement age), the coefficient of age discrimination is not significant. This implies that being discriminated against based on your age has no effect on either the choice to leave one’s job before the compulsory retirement age on the choice to remain in the workforce past the compulsory retirement age. However, in the Late Retirement column the variable age is significant. This negative coefficient suggests that respondents are less likely to work longer as they become older. This finding is supported by earlier evidence from Kooij et al. (2008), which suggested that, for older people, factors related to age have a negative impact on the
18
motivation to continue to work. Furthermore, the coefficient of age discrimination is significant when testing the likelihood of the acceptance of job offers by unemployed
individuals in 1971. The negative coefficient implies that discrimination has a negative effect on the likelihood of accepting job offers, meaning that job offers are less likely to be accepted if individuals have experiences of being discriminated against due to their age.
Since the results in Table VII can only be interpreted by their sign, I present the marginal effect of age discrimination in Table VIII. The results presented in this table indicate the marginal probability of early retirement and late retirement decisions. The variables wage and age are held at their means and the categorical variables marital status and employment
status at their first categories, which include men who are married and employed.
Table VIII
Marginal effects on early and late retirement decisions (1971)
Variable Early retirement Late retirement
Age discrimination1 0.0300 0.0096
(0.0699) (0.0556)
Age: 55.28 (mean) 55.53 (mean)
Wage: 10,187.69 (mean) 10,159.78 (mean)
Employment status: Employed Employed
Marital status: Married Married
1 Age discrimination: includes experiences in discrimination from 1966 to 1971 * Significant at confidence level of 10%
** Significant at confidence level of 5% *** Significant at confidence level of 1%
The marginal effects presented in Table VIII are not significantly different from zero. These results imply that age discrimination does not affect the behaviour of older male employees toward these retirement decisions. Further, Table IX presents the marginal
probability of the acceptance of job offers. The variables age and wage are again held at their means and the categorical variable marital status at the first category, married. However, the variable employment status is now held at the fifth category, which represents retired
individuals in the sample.
Table IX
Marginal effect on job offers (1971)
Variable Job offers
Age discrimination1 -0.1816***
(0.0653)
Age: 59.85 (mean) Wage: 2,809.77 (mean) Employment status: Retired Marital status: Married
19
1 Age discrimination: includes experiences in discrimination from 1966 to 1971 * Significant at confidence level of 10%
** Significant at confidence level of 5% *** Significant at confidence level of 1%
Since age discrimination is significant, the marginal effect on the acceptance of job offers can be interpreted as follows: A retired and married man at the age of 59 years with a yearly income of $2,809.77 is 18.16% less likely to accept a job offer from some employer after he has been discriminated against. This result implies that age discrimination has a significant effect on unemployed men’s decision to enter the workforce again.
Overall job satisfaction
In the last survey, in 1990, individuals reported their satisfaction with their entire working life. In Table X below, the ordered probit regression of the respondents’ overall job satisfaction is summarized. The motivation of the individuals is included in the model as a dummy variable. The outcome for individuals who are motivated by wages serves as the base in this regression.
Table X
Ordered probit regression of overall job satisfaction (1990)
Variable Job satisfaction1 (2)2
Age discrimination3 0.3948*** 0.4035***
(0.0798) (0.0752)
Motivation
Liking the work -0.1220 -0.0984
(0.0650) (0.0584) Wage4 0.18161 (0.06567) Observations 1,481 1,799 Chi-2 Test 28.13 31.36 Prob> chi2 0.0000 0.0000 Pseudo R2 0.0079 0.0072
1 Question: As you think back over your entire work life, how satisfied are you with it? 2 Wages are omitted for column (2)
3 Age discrimination: includes experiences of discrimination from 1966 to 1981 4 Wages collected in 1981
* Significant at confidence level of 10% ** Significant at confidence level of 5% *** Significant at confidence level of 1%
The coefficient of age discrimination is significant at a 1% confidence level in both of the regressions. The positive coefficient implies that the presence of age discrimination leads to an increase in the number category of job satisfaction and therefore reduces employees’ satisfaction.
The marginal effect of several variables on overall job satisfaction is presented in Table XI. The results in this table indicate the precise marginal probability of responses in the
20
higher number categories of job satisfaction after experiences of age discrimination; in other words, older male workers who are not very satisfied with and not satisfied at all with their careers. The variable wage is held at mean and the categorical variable motivated by is held first on the outcome motivated by wages, and thereafter on the outcome motivated by liking the job.
Table XI
Marginal effect on overall job satisfaction (1990)
Men who are Men who are
not very satisfied not satisfied at
with their career (4) all with their career (5)
Age discrimination1 0.0219*** 0.0094***
(0.0056)) (0.0032)
Wage: 6,783.12 (mean) Motivated by: wages
Age discrimination1 0.0180*** 0.0070***
(0.0044) (0.0023)
Wage: 6,783.12 (mean)
Motivated by: liking the work Observations: 1,481
1 Age discrimination: includes experiences in discrimination from 1966 to 1981 * Significant at confidence level of 10%
** Significant at confidence level of 5% *** Significant at confidence level of 1%
The results presented in Table XI can be interpreted as follows: an older male worker with a yearly wage of $6,783.12 in 1981 who is motivated to work by his wage is 2.19% more likely to report being not very satisfied with his career after he experiences age
discrimination. Moreover, the is 0.94% more likely to report the highest number category for job discrimination after experiencing age discrimination, which means he is not satisfied at all with his career.
On the other hand, an older male worker earning the same yearly wage, but motivated to work because he likes the job, is 1.8% more likely to report being not very satisfied, and 0.7% more likely to report the highest numbered category of job satisfaction (not satisfied at all), after any experiences of age discrimination during his career.
The results show the probabilities of lower job satisfaction are higher for employees who are motivated to work only by their wages. This result is supported by the work of Herzberg (1957), who suggests that employees who are only motivated by hygiene factors such as salary experience lower job satisfaction. Concluding, experiences in age
21
6. Discussion
In this thesis the data from the National Longitudinal Survey of Older Men was used and these respondents were randomly chosen from different areas in the United States. Therefore, the sample is likely to characterize the population of the entire country. However, the results on retirement found in this thesis cannot be generalized to others; for example, to countries in Europe. Findings from Gruber and Wise (2002) suggest that cultural differences across countries affect the labour force participation of older workers. Moreover, all countries have different social security programs that influence the retirement decisions of employees. For example, workers are significantly less likely to retire when additional work results in higher benefits for employees and their families. Besides, the statutory age of retirement for men is not uniform around the world (See Appendix C).
Furthermore, the use of self-reports is an effective tool in behavioural research because of their utility. However, they might not be entirely valid (Austin et al., 1998). For example, participants may not be honest about their feelings or do not interpret the question in the same way. It might be possible that someone report not experiencing age discrimination, while others with same experiences will report it.
Another limitation of the data would be that it only consists of observations for older men and might therefore not be representative of the effects of age discrimination on the behaviour of older female workers.
Further, the responses gathered from the database are from the years 1966 to 1990, and therefore not very recent. Moreover, there were significantly fewer observations during the last years, and that might have an effect on the results. However, the data is still very relevant, as it includes valuable information about age discrimination that is very rarely found in
previous research. 7. Conclusion
Age discrimination does seem to have a significantly effect on the job satisfaction of older men; in other words it probably reduces their job satisfaction. Therefore, the first hypothesis is rejected. This result is in line with my expectation that older male employees with experiences of age discrimination will report a lower level of job satisfaction.
Age discrimination is not significant for either older male workers choosing to retire early or to work longer than the statutory age of retirement. This result is inconsistent with my
22
expectations, since I expected that workers with experiences in age discrimination would be more likely to retire before the compulsory retirement age.
Moreover, I thought that age discrimination would also affect the likelihood of older male workers staying in the work force. However, the effect of age discrimination is
significant on the likelihood of older retired male workers when considering a job offer. This result implies that age discrimination has a negative effect on job offers, meaning that job offers are less likely to be accepted after an experience of age discrimination. This result is inconsistent with my expectation that unemployed men are more likely to accept the offer of a new job after they have experienced age discrimination.
Finally, men are more likely to report lower satisfaction with their career after experiencing age discrimination. Moreover, this probability is higher for the employees who are only motivated to work by wages. This is inconsistent with my expectation, since I expected that it would have a significantly bigger effect on workers motivated by liking their job.
This thesis considered the following research question: Does age discrimination have an effect on older employees’ job satisfaction and motivation to work? The results imply that age discrimination does have an effect on older employees’ job satisfaction. However, there is no significant evidence found that older male workers’ behaviour is changed, and their
motivation is not affected by discrimination to their age.
As discussed before, I think it would be relevant to examine the effect of age
discrimination on the behaviour of women in the workforce. Besides, there is need for more information from employees about their experiences with age discrimination since the future workforce will consist more of older people. Furthermore, if data will become available, it will be relevant to examine the effect of age discrimination on the behaviour of older workers in different European countries, since these are likely to differ from the outcomes in the United States.
23
References
Austin, E.J., Gibson, G.J., Deary, I.J., McGregor, M.J., & Dent, J.B. (1998). Individual response spread in self-report scales: personality correlations and consequences.
Personality and Individual Differences, 24, 421–438.
Boumans, N.P., de Jong, A.H., & Janssen S.M. (2011). Age-differences in work motivation and job satisfaction. The influence of age on the relationships between work
characteristics and workers’ outcomes. International journal of Aging and human
development, 73(4), 331-350.
Chiu, W.C.K., Chan, A.W., Snape, E. & Redman, T. (2001). Age stereotypes and
discriminatory attitudes towards older workers: an east-west comparison. Human
Relations, 54(5), 629-661.
Claes, R., & van de Ven, B. (2008). Determinants of older and younger workers’ job satisfaction and organizational commitment in the contrasting labor markets of Belgium and Sweden. Ageing & Society, 28, 1093-1112.
De Lange, A.H., Taris, T.W., Jansens, P.G.W., Smulders, P., Houtman, I.L.D., & Kompier, M.A.J. (2006). Age as a factor in the relation between work and mental health: Results from the longitudinal TAS survey. Occupational health psychology: European
perspectives on research, education and practice, 21-45.
Equal Employment Opportunity Commission. (2016). Annual Report for Fiscal Year 2015. Retrieved May 18, 2016, from https://www.eeoc.gov/
Gruber, J. & Wise, D.A. (2002). Social security programs and retirement around the world: Micro Estimation. National Bureau of Economic Research, 9407.
Herzberg, F., Mausner, B., & Snyderman, B. (1959). The Motivation to Work. New York: Wiley.
Johnson, R.W., & Neumark, D. (1997). Age discrimination, job separations, and employment status of older workers: Evidence from self-reports. Journal of Human Resources, 32(4), 779-811.
Kooij, D., de Lange, A., Jansen, P., & Dikkers, J. (2008). Older workers’ motivation to continue to work: five meanings of age. Journal of Managerial Psychology, 23(4), 364-394.
Lahey, J. (2008). State Age Protection Laws and the Age Discrimination in Employment Act.
24
Loretto, W., & White, P. (2006). Employers’ attitudes, practises and policies towards older workers. Human Resource Management Journal, 16(3), 313-330.
Schalk, R., & van Veldhoven, M.J.P.M. (2010). Moving European research on work and ageing forward: overview and agenda. The European Journal of Work and
Organizational Psychology, 19(1), 76-101.
Stamov-Roβnagel, C & Hertel, G. (2010). Older worker’s motivation: against the myth of general decline. Management Decision, 48(6), 894-858.
Stock, J.H. & Watson, M.M. (2011). Introduction to econometrics, 3rd edition, Pearson Wang, M., & Shultz, K.S. (2010). Employee retirement: A review and recommendations for future investigation. Journal of Management, 36, 172-206.
Warr, P. (2008). Work values: Some demographic and cultural correlates. Journal of
Occupational and Organizational Psychology, 81, 751-775.
25 Appendices Appendix A Appendix A1
Table I summarizes the reports of discrimination by older employees which reported a low job satisfaction, category 3 and 4 in 1976. The t-statistics on difference in older men, which reported age discrimination and who did not have these experiences, is 4.37 and thereby significant at a confidence level of 1%.
Table I (1976)
Observations Low Job Satisfaction Percentage
Overall 2,039 133 6.52%
Do have experience
with age discrimination 211 27 12.80%
Do not have experience
with age discrimination 1,828 106 5.80%
Appendix A2
Table I summarizes the reports of discrimination by older employees which reported a low job satisfaction, category 3 and 4 in 1981. The t-statistics on difference in older men, which reported age discrimination and who did not have these experiences, is 3.72 and thereby significant at a confidence level of 1%.
Table II (1981)
Observations Low Job Satisfaction Percentage
Overall 1,010 59 5.84%
Do have experience
with age discrimination 135 20 14.81%
Do not have experience
26
Appendix B
Appendix B1
Table I gives the precise marginal probability of being in the highest categories of job satisfaction in 1976 after experiences of age discrimination.
Table I
Marginal effect on job satisfaction (1976)
Men who somewhat Men who dislike
dislike their job (3) their job very much (4)
Age discrimination1 0.02662*** 0.0149***
(0.0076) (0.0045)
Age: 59.78 (mean) Wage: 9,585.68 (mean) Employment status: employed Marital status: Married Observations: 1,770
1 Age discrimination: includes experiences in discrimination from 1966 to 1976 * Significant at confidence level of 10%
** Significant at confidence level of 5% *** Significant at confidence level of 1%
Appendix B2
Table II gives the precise marginal probability of being in the highest categories of job satisfaction in 1981 after experiences of age discrimination.
Table II
Marginal effect on job satisfaction (1981)
Men who somewhat Men who dislike
dislike their job (3) their job very much (4)
Age discrimination1 0.0257*** 0.0074**
(0.0098) (0.0034)
Age: 64.28 (mean) Wage: 11,510.94 (mean) Employment status: employed Marital status: Married Observations: 885
1 Age discrimination: includes experiences in discrimination from 1966 to 1981 * Significant at confidence level of 10%
** Significant at confidence level of 5% *** Significant at confidence level of 1%
27
Appendix C
The map below shows the statutory retirement age for men in the private sector. The statutory retirement age is the age at which men working can retire and receive full benefits.
Worldwide retirement ages