Changes of life satisfaction in the transition to retirement: a latent-class approach
Martin Pinquart & Ines Schindler
-
This study was to take a close look at the variability in changes of life satisfaction during the transition to retirement. We assumed that the effects of retirement vary depending on a person's developmental context and that, therefore, contradictory (contradictory positions have been advanced as to whether retirement has negative, positive, or no effects on subjective well-being) theoretical positions may be descriptive of different subpopulations of retirees.
-
Contradictory positions have been advanced as to whether retirement has negative, positive, or no effects on subjective well-being. The authors investigated in life satisfaction in 1,456 German retirees. Using latent growth mixture modeling, the authors found 3 groups of people who experienced retirement differently. In Group 1, satisfaction declined at retirement but continued on a stable or increasing trajectory thereafter. Group 2 demonstrated a large increase in satisfaction at retirement but overall declining satisfaction. In Group 3, satisfaction showed a temporary very small increase at retirement. Groups differed by retirement age, gender, socioeconomic status, marital status, health, unemployment before retiring, and region. Thus, retirement is not a uniform transition, and resource-rich individuals are less likely to experience retirement-related change in satisfaction.
-
"Retirement is an important milestone in later adulthood that is associated with changes in daily routines, social roles, social contacts, and income" (Pinquart and Schindler, 2007). Because of growing life expectancy, older adults spend an increasing amount of time in retirement (Lee, 2001/ cited directly from Pinquart and Schindler, 2007). In many countries, available pension and saving plans have eased the financial burdens of retirement.
-
How and whether the transition to retirement impacts on a person's subjective well-being (SWB) has remained the subject of much debate (Reitzes, Mutran, & Fernandez, 1996). Retirement has been conceptualized as either a transition accompanied by psychological distress or a transition contributing to enhanced or stable SWB. Rather than trying to decide between those perspectives, our central aim in this study was to take a close look at the variability in changes of life satisfaction during the transition to retirement. We assumed that the effects of retirement vary depending on a person's developmental context and that, therefore, contradictory theoretical positions may be descriptive of different subpopulations of retirees. Using longitudinal data from the German Socioeconomic Panel (Haisken-DeNew & Frick, 2003), we were able to test for groups of retirees showing differential change in life satisfaction and for predictors of group memberships (e.g., demographic data). Further, we compared group-specific trajectories with existing theoretical positions and empirical findings regarding the effects of retirement.
-
Competing theoretical positions and empirical findings on the effects of retirement
1. Sociological role theory
Sociological role theory points to employment as a fundamental role that is central to an individual's identity (J.E. Ki & Moen, 2001). Thus, retirees are expected to feel that they have lost an important role, and such feelings could lead to psychological distress.
2. Continuity theory
Similarly, continuity theory would suggest that retirement may be psychologically stressful because of the discontinuity in a central role or because individuals experience rolelessness if they cannot replace the lost work role with new activities (Richardson & Kilty, 1991). Thus, continuity theory would suggest an overall decline of SWB after retirement. Nonetheless, continuity in other social roles may buffer the effects of retirement on SWB (Reitzes & Mutran, 2002). Individuals may maintain or improve their SWB by spending time with their families, developing new interests or hobbies, learning new things, or volunteering (Wu, Tang, & Yan, 2005).
In addition, retirement also entails culturally transmitted rights, such as the right to economic support and the right to autonomy concerning the management of one's time, which could be a source of positive well-being (Atchley, 1976). Finally, personality traits, such as high levels of emotional stability, may contribute to stable levels of SWB during the retirement transition (Reis & Pushkar-Gold, 1993).
-
The retirement transition is a process that begins in preretirement and continues into postretirement. Atchley (1974, 1976) has suggested a process model of retirement adjustment.
1. Anticipatory attitudes about retirement are formed in the preretirement phase.
2. Before retirement, SWB may decline, as individuals may worry about the anticipated loss of the work role.
3. After the retirement event, retirees may experience a short honeymoon phase marked by vacations and new interests or a rest-and-relaxation phase of brief respite from the obligations of work.
4. These positive feelings may change to disenchantment and impaired SWB as individuals face the reality of everyday life in retirement. This should be followed by a reorientation stage characterized by the development of a realistic view of the social and economic opportunities and constraints of retirement.
5. Finally, stability occurs once the retirees accommodate and adjust to retirement. Thus, after the retirement event, a temporary increase of SWB followed by a marked decline and then by a smaller increase was expected.
Nonetheless, Atchley (1974) cautioned that these stages do not occur in an orderly and consistent fashion for all individuals.
-
Other theories also suggest that there may be different trajectories of retiree's SWB. For example, Antonovsky and Sagy (1990) suggested a series of retirement-related developmental tasks, such as searching for active involvement, reevaluating one's world outlook, and developing a healthy lifestyle. How retirees complete these tasks determines whether positive or negative psychological consequences of retirement dominate. Similarly, stress and coping theory suggests that adaptation to retirement depends on ways of coping with the transition (Smith, Patterson, & Grant, 1992). Thus, subgroups of retirees would show decline, stability, or increase in SWB.
- EMPIRICAL RESEARCH ON THE ASSOCIATION BTW R & L-S
Empirical research on the association between retirement and life satisfaction is contradictory. Some studies reported higher levels of life satisfaction in retirees than in their working peers (e.g., Isaksson & Johansson, 2000), whereas others found no differences (e.g., Warr, Butcher, Robertson, & Callinana, 2004) or a lower level of life satisfaction in retired participants (Richardson & Kilty, 1991; Riddick, 1985). In a longitudinal study, Mayring (2000) found no significant changes in global life satisfaction between 6 months before retirement and 18 months after retirement. Similarly, Isaksson and Johannson (2000) observed no significant change in retirement.
-
The heterogeneity of previous findings may, in part, result from differences in research methodology. For example, studies that compare retirees with working peers do not allow for assessing whether group differences would be a cause of retirement, a consequence of retirement, or the effect of an unmeasured third variable (e.g., differences in health status). In addition, differences between the results of individual studies may also be due to varying intervals between the event of retirement and the time of measurement, as would be suggested by process models of retirement adaptation (e.g., Atchley, 1974, 1976; Richardson & Kilty 1991). Third, differences between individuals may be masked by averaging data. Retirement may be associated with more gains in some older adults and with more losses in others, and individuals may, therefore, differ in the direction and amount of change in life satisfaction. For example, Richardson and Kilty (1991) identified different trajectories for subgroups of retirees, such as stable (34%), initial decline followed by no further change in life satisfaction (18%), and initial decline followed by an increase in life satisfaction (14%). Fourth, retirement-related changes may overlap with unspecific age-associated changes. Because of age-associated losses, such as declines in health and physical performance, several longitudinal studies have found a small decline in life satisfaction in the second half of life that was accelerated in the old-old (e.g., Schilling, 2006).
-
The German Socioeconomic Panel data used here have several advantages over those used in previous studies and, therefore, allowed us to address some of the critical methodological issues we have just raised. Most important, it is a nationally representative longitudinal study that started before the event of retirement and followed the retirees over longer time intervals, that is, beyond the immediate response to the retirement event. This design allowed us to disentangle retirement-related changes in life satisfaction from changes in satisfaction that may have occurred for other reasons (e.g., declining satisfaction in response to deteriorating health in older age). Moreover, we tested for differential trajectories of life satisfaction rather than focusing on average change across the whole sample. As previous studies have demonstrated, it may not be fruitful to investigate the effect of retirement on satisfaction but rather to determine subsamples of respondents for whom the effects of retirement vary (Richardson & Kilty, 1991). On the basis of available theories, we expected to find subgroups of retirees whose life satisfaction declined (continuity theory), temporarily increased (Atchley's model), and stabilized and improved because of retirement-associated gains (Wu et al., 2005). To identify these groups, we analyzed trajectories of life satisfaction from preretirement to postretirement with latent growth mixture models (cf. B. Muthen, 2001, 2004; B. Muthen & Shedden, 1999). First, we determined whether different gorups of participants showed different satisfaction trajectories and how many groups were present in our data. Second, we identified predictors of membership in the different trajectory groups.
-
Predictors of Change in Life Satisfaction During Retirement
Scholossberg (1995) and Fouquereau, Fernandez, and Mullet (2001) suggested that there were four groups of variables that may influence adaptation to retirement: a) personal characteristics (e.g., gender), b) resources before and after retirement (e.g., level of income), c) coping responses, and d) situational variables (e.g., transition from unemployment to retirement). We restrict our discussion to aspects of a, b, and d that have been assessed in the present study.
Age at retirement
In Germany, the mandatory retirement age is 65. However, similar to the United States and many other western countries (e.g., Shaw, Patterson, Semple, & Grant 1998), Germany has seen a trend toward early retirement (cf. Kohli, 1996). For example, because of the downsizing of companies and higher flexibility in pension schemes, many workers choose to or have to leave their jobs before the age of 65 years. Several available studies have reported that early retirement is associated with less satisfaction than late retirement is (e.g., Plamore, Fillenbaum, & George, 1984; Williamson, Rinehart, & Blank, 1992), a result that may be based on an overrepresentation of involuntary retirement and health problems among early retirees. However, some cross-sectional studies reported that higher age at retirement was associated with lower levels of life satisfaction (McGoldrick & Cooper, 1994). This may have been caused by an age-associated decline of life satisfaction (e.g., due to deteriorating health) rather than to the timing of retirement.
Gender
Socioeconomic Status (SES)
Although workers with jobs of high occupational prestige and high income may have more to lose when retiring, they also have more resources for adapting to retirement. For instance, they have a higher pension coverage that increases the social and leisure opportunities in retirement (Dorfman, 1989) and a higher educational attainment that provides individuals with the social skills and self-direction to appreciate the opportunities in retirement for meaningful and intrinsically satisfying activities (Reitzes & Mutran, 2004). In fact, several studies have shown that higher educational attainment, income, and occupational prestige and high levels of autonomy in the workplace are associated with higher levels of life satisfaction and morale after retirement (e.g., Belgrave & Haug, 1995; Dorman, 1989; George, Fillenbaum, & Palmore, 1984; Mayring, 2000; Richardson & Kilty, 1991).
However, other studies found no association of indicators of SES with retirees' satisfaction and other indicators of SES with retirees' satisfaction and other indicators of psychological well-being (e.g., S. Kim & Feldman, 2000; Seccombe & Lee, 1986; Wu et al., 2005) or a positive association only in a subsample of retirees (Szinovacz & Davey, 2005).
(...) Richardson and Kilty (1991) found that high occupational status rather than income or education predicted whether retirees showed improvement or decline in well-being after retiring.
Marital status
Although married adults tend to report higher levels of life satisfaction than do unmarried peers, differences in life satisfaction between married and unmarried respondents are generally small (Pinquart, 1998). It has been shown that married adults have more positive attitudes toward retirement (Mutran, Reitzes, & Fernandez, 1997) and higher levels of retirement satisfaction (Price & Joo, 2005) than do unmarried adults. However, many studies found no influence of marital status on global and domain-specific indicators of life satisfaction in retirees (Austrom et al., 2003; Dorfman, 1995; S. Kim & Feldman, 2000; Seccombe & Lee, 1986)
Physical health
Physical health is an important resource for adjustment to retirement. Restrictions in health can limit the pursuit of activities in retirement, and workers in bad health are more likely to retire involuntarily. Several studies found positive associations of retirees' life satisfaction with indicators of objective health (Dorfman, 1995) and subjective health (e.g., Austrom et al., 2003; S. Kim & Feldman, 2000; McGoldrick & Cooper, 1994; Szinovacz & Davey, 2005; Wu et al., 2005). Only a few studies have found no significant association between health and psychological well-being in retirement (Seccombe & Lee, 1986).
Commitment to leisure activities before retirement
After retirement, the time previously spent in the work role has to be filled with new activities. Thus, individuals who have already developed some leisure interests before retirement may show positive changes in life satisfaction during the retirement transition. Two different studies found that having hobbies and spending more time on these activities were related to higher levels of life satisfaction in retirement (Austrom et al., 2003; S. Kim & Feldman, 2000).
> my opinion: those with leisure activities well before retirement, they might tend to have a high level of life satisfaction regardless of retirement event. Leisure activities in and of themselves give a rather enduring positive impact on one's life satisfaction.
Transition from unemployment to retirement
In Germany, economic crisis led to the downsizing and closing of many companies, and unemployed persons aged 55 years or older had lower chances for reemployment than did younger individuals. Thus, an increasing number of older adults have been shifting from unemployment to retirement. In this case, retirees face less of an income loss or may even increase their income after retirement. In addition, retirement may be a positive role change for the formerly unemployed because unemployment is negatively evaluated by the society, for instance, as indicating a personal failure (Dekker & Ester, 1992; Furnham, 1983). Furthermore, less change in social contact and daily routines would be expected for the formerly unemployed versus the formerly employed, leading to lower levels of stress in the transition to retirement. These factors may lead to more positive consequences of retirement for the formerly unemployed. Nonetheless, at least one factor could reduce retirement satisfaction in this group: The longer these persons have been unemployed before retirement, the lower their pensions will be. In a comparative study, Warr et al. (2004) found lower levels of life satisfaction in unemployed adults than in employed and retired peers. In addition, Belgrave and Haug (1995) found that those who retired from a situation of unemployment showed higher levels of SWB than did their retired peers.
-
Comments
Post a Comment