Background
To encounter the pressure of population ageing on the social security system and the expected shortage of workers in the next decades in many developed countries [
1], there is a societal need for workers to prolong their working life. Many European governments responded to this need by increasing the statutory retirement age and discouraging early exit from the workforce. Working beyond retirement may also provide in the need for prolonged working lives. Although previous research has identified a variety of factors that influence early exit from the labor market, still little is known about the factors that impact working beyond retirement.
Retirement does not necessarily mean a final farewell to the labor market. In the Netherlands, a growing number of retirees engage in work activities between the end of their career employment and complete labor force withdrawal, which is often referred to as bridge employment. The percentage of people in the age group 65 to 70 years that works for at least 12 h per week, increased from 3.4 % in 2001 to 8.6 % in 2014 [
2]. In comparison, these numbers are much higher in the United States where 30.8 % of the age group 65 to 69 worked at least 35 h per week, in 2010 [
3].
Although previous research has shown that health, job characteristics, skills and knowledge, and financial and social factors may push or pull workers from work to (non-disability) early retirement [
4‐
10], still little is known about the factors that impact working beyond retirement. It is unclear whether this is just predicted by a beneficial status of factors that influence early retirement, or whether also additional factors play a role. Second, it is unclear whether different factors predict working beyond retirement among retirees who retire ‘on time’ compared to retirees who retire early. Third, occupational epidemiologists often focus on health, work ability and physical job demands in relation to work force participation, whereas psychologists focus on motivational processes and psychosocial job demands, and economists focus on financial factors. To understand the complexity of working beyond retirement a broader perspective is needed.
Theoretical background
The life course perspective provides the opportunity to get a better understanding of the different factors that influence working beyond retirement. According to this perspective, it is expected that transitions, i.e. changes related to entering or exiting roles, such as the transition from work to retirement and vice versa, are embedded within multiple interdependent trajectories, i.e. within the life spheres health, work, family and leisure [
11]. Working beyond retirement cannot be fully understood in isolation of a person’s health, work and motivation to work, skills and knowledge, and financial and social situation (i.e. family and other forms of social participation). In the present study we will use the life course perspective as a framework to understand determinants of working beyond retirement by focusing on the life course principles ‘human agency’, ‘linked lives’, and ‘timing’ [
11].
‘
Human agency’ implies that individuals actively create their own lives and choices, but that this is done within a set of opportunities and constraints that come with one’s history or that exist in one’s environment. Previous research showed that work motivation and motives are important predictors of early retirement intention and working beyond retirement, respectively [
12,
13]. A study showed that the most important reason to work beyond retirement, is that someone enjoys working [
13]. Health problems can be considered as a constraint to continue working, since it predicts early retirement [
10,
14]. Job characteristics, such as physical work demands and high work pressure, may also hinder continued working [
8,
15]. On the other hand, psychosocial job characteristics, such as appreciation at work, could facilitate continued working [
5]. Financial factors can also be considered as a constraint or an opportunity to stop working. Previous research showed that the financial possibility to retire early strongly contributed to early retirement [
5].
The life course perspective also emphasizes the social embeddedness of transitions, by pointing to the principle of
‘linked lives’. This principle refers to the interdependence of lives and it states that the lives of individuals are bound to the lives of others [
11]. Transitions from work to early retirement are shaped to a large extent by social relationships, i.e. within the family and at work. Previous research showed that having a partner increased the likelihood of early retirement [
8]. Also support of the partner to continue working or to retire early predicted retirement at older age and early retirement [
5,
16].
Another principle within the life course perspective is ‘
timing’, which refers to the age at which an experience occurs and how it is experienced [
11]. Transitions can occur ‘on time’ or ‘off time’. Transitions that occur off time ensure that persons do not have the chance to go through anticipatory socialization, and that they lack peers to provide social support and to share experiences regarding the transition [
11]. Therefore, 'off time' transitions might be experienced as involuntary. In recent decades, there has been a strong ‘early exit culture’ in the Netherlands [
17]. However, several pension system reforms have been implemented, such as the increase of the statutory retirement age from 65 years in 2012 to 67 in 2021 [
18]. In addition, there is an extensive public debate on the need to prolong working life, and hence, the social norm may have become more favorable with respect to prolonged careers. Retiring at the statutory retirement age can be considered as ‘on time’ retirement, and retiring before the statutory retirement age can be considered as ‘off time’ retirement. Although early retirement is often considered as one of the more ‘voluntary’ pathways of early exit from the workforce, especially as compared to disability pension and unemployment, a recent qualitative study showed that early retirement is not always experienced in this way [
19]. In this study employees with poor health sometimes retired early because they felt pushed out by their employer, although they themselves did not experience a reduced ability to work. In addition, such involuntary transitions from work to early retirement may have negative financial consequences, in situations where there was actually a financial need to continue working until the statutory retirement age.
Research questions and hypotheses
In the present study we investigate which factors influence working beyond retirement from a life course perspective. We aim to gain insight in how the timing of retirement could define subgroups of retirees regarding influential factors. Our first research question is: What is the influence of individual characteristics, work motives and motivation, health, job characteristics, skills and knowledge, and the financial and social situation on working beyond retirement? Our second research question is: Are there differences in the factors which influence working beyond retirement between those retirees who reached the statutory retirement age of 65 years, and thus retired ‘on time’, and those who did not yet reach the statutory retirement age, and thus retired ‘off time’ (early)? In line with the principle ‘timing’ of the life course perspective, we expect that ‘on time’ retirees are intrinsically motivated to work beyond retirement, e.g. because they enjoy working, find their work meaningful or interesting, or think it offers opportunities for learning and development, whereas ‘off time’ retirees are extrinsically motivated to work beyond retirement, e.g. for financial reasons or because others expect them to do so.
Discussion and Conclusion
This study aimed to gain insight in the influence of individual characteristics, work motives and motivation, health, job characteristics, skills and knowledge, and the financial and social situation on working beyond retirement and to investigate whether there are differences between ‘on time’ and ‘off time’ retirees.
Regarding the first research question, our study showed that work motivation, health, and the financial situation influence working beyond retirement. This is in line with the principle ‘human agency’ of the life course perspective and with previous studies on predictors of early retirement [
5,
6,
9,
10,
12]. The results showed that persons who were highly engaged in their work were more likely to work beyond retirement, which confirms findings from a previous study that showed that intrinsic motivation was related to the willingness to continue working [
31], and a study that showed that a lower motivation to work was associated with intention to retire early [
12]. Previous research showed that self-perceived poor health predicts early retirement, but that specifically mental health problems are not related to early retirement [
10]. Indeed, the opposite seems true with regard to working beyond retirement; good physical health predicts working beyond retirement and good mental health does not. Hence, even though our study population probably is a selection of the more healthy employees, since the unhealthy ones partly have left the workforce through disability benefits in previous years (‘healthy worker effect’ [
32]), physical health predicts working beyond retirement. Contrary to what we expected, work motives, job characteristics, and skills and knowledge did not predict working beyond retirement multivariately, whereas the enjoyment motive, appreciation at work, interesting work and developmental proactivity predicted working beyond retirement in the univariate analyses. This may be explained by work engagement, which was moderately interrelated with these variables (Pearson’s correlations 0.36, 0.30, 0.34, and 0.40, respectively).
In line with the principle ‘linked lives’ of the life course perspective, we expected that having a partner, and employment status of the partner are related to working beyond retirement. However, our study found no effect. This contradicts previous research that showed that having a partner increased the likelihood of early retirement [
8], and that the attitude and support of the partner about continuing to work or early retirement predicted retirement at older age and early retirement, respectively [
5,
16]. However, persons who participated in voluntary work were more likely to work beyond retirement, which is an indication that working beyond retirement is embedded in someone’s social situation to a certain degree, and that it cannot be seen in isolation from other social relationships.
Regarding the second research question, our study showed that determinants of working beyond ‘off time’ retirement and working beyond ‘on time’ retirement were largely comparable. However, two remarkable differences should be mentioned; in the group of ‘off time’ retirees, not having the financial possibility to retire early predicted working beyond retirement, whereas in the group of ‘on time’ retirees persons who did not have the financial possibility to retire early were less likely to work beyond retirement. This partially supports our hypothesis that ‘off time’ retirees are extrinsically motivated to work beyond retirement to a larger extent than ‘on time’ retirees. This might indicate that the group of ‘off time’ retirees partly consisted of persons who experienced their transition from work to early retirement as involuntary, for example because they felt pushed out by their employer, whereas they actually wanted to continue working or could not financially afford to retire early, and therefore started to work beyond retirement. Unfortunately we did not have information on voluntariness of the transition. Future research might investigate whether the retirement context, i.e. the reason for retirement and voluntariness of retirement could be important characteristics to define subgroups of retirees with regard to working beyond retirement.
Persons who were highly engaged in their work, were more likely to work beyond retirement, both in the group of ‘off time’ retirees and the group of ‘on time’ retirees. This is contrary to our hypothesis that ‘on time’ retirees would be intrinsically motivated to a larger extent than ‘off time’ retirees. Although timing of retirement could partially define subgroups of retirees, the pattern regarding intrinsic and extrinsic motivation is not as clear as we would have expected based on the timing principle of the life course perspective. This may indicate that there are also other characteristics that could define subgroups of retirees, such as educational level. It may be that within our study population, educational level explains why we found that both intrinsic factors and extrinsic factors predicted working beyond retirement. We tested this possibility by post-hoc analyses stratified by educational level (data not shown). The results indicated that within persons with a higher educational level, having the financial possibility to retire early predicted working beyond retirement, whereas within persons with a lower educational level having the financial possibility was not related to working beyond retirement. However, the role of work engagement was comparable for persons with a low educational level and those with a high educational level. Hence, also educational level could only partially explain which factors predicted working beyond retirement. This may imply that either there are no clear subgroups of retirees regarding working beyond retirement, or that other characteristics, which were not incorporated in the present study, play a role. Although our study involved a broad range of variables in a diversity of domains, we did not consider different ‘types’ of retirement. We were not able to distinguish between retirement schemes, e.g. an early retirement scheme via employer or sector, retirement because someone reached a job-specific retirement age, part-time, or full-time retirement. Especially the degree to which working beyond retirement is financially attractive might differ between different retirement schemes. Future research might further investigate whether education and ‘type’ of retirement are relevant characteristics to define subgroups of retirees.
Strengths of the present study are the longitudinal character of the large dataset, and the fact that variables frequently studied in different areas of expertise, i.e. work motives and motivation, health, job characteristics, skills and knowledge, and the financial and social situation, were all included in this study. Moreover, we had low drop-out in the present study; 79 % of the participants of interest at T1 also participated at T2, T3, and T4, which can be considered as a high response in longitudinal research. However, this study also has limitations. First, all data relied on self-reports. A drawback of measuring employment status relying on self-report is that it is dependent on the interpretation of the participant and it is thus not defined by, for example, the main source of income as registered by a tax institution or the government. A second limitation is the relatively short follow-up period of three years, which concerns only a small part of the whole life course. From a life course perspective one might expect that also experiences earlier in life impact decisions regarding work and retirement. This was supported by a study of Damman et al. that showed that mid-life experiences, i.e. educational investments, job changes, late transitions into parenthood, and late divorces, are associated with weaker retirement intentions, whereas mid-life health problems are associated with stronger retirement intentions [
4]. In addition, this study showed that later labour market entry and late transitions into parenthood were associated with a lower likelihood of early retirement, and that part-time work before the age of 50 years resulted in a higher likelihood of early retirement. Third, selection bias may have occurred as a result of selective entry in this internet panel. Certain groups of persons, i.e. without internet access, illiterate persons and persons who do not master the Dutch language, may be underexposed in an internet panel. The findings of the present study may not be generalizable to these groups. Selection bias may also have occurred as a result of selective non-response at follow-up. However, the response of the study was high, i.e. 79 %, and differences between respondents and non-respondents on baseline characteristics were small, and considered as not relevant.
As previously described, future research is needed to investigate the role of the retirement context, i.e. the reason for retirement and voluntariness of retirement, as well as differences in educational level and ‘type’ of retirement with regard to working beyond retirement. Besides, we used the statutory retirement age of 65 years to distinguish between ‘on time’ and ‘off time’ retirement. Since not only institutional but also cultural and individual norms determine the ‘right’ time of a transition, future research is needed to determine where the line should be drawn. Furthermore, future research should explore further possibilities of using the life course perspective to understand factors that influence working beyond retirement. There is still little understanding on how different determinants influence working beyond retirement and why persons work beyond retirement. Especially the group of persons who works beyond early retirement is of interest, since an alternative career path for these persons may have been to retire at later age. Finally, it is unclear whether working beyond retirement mostly occurs in an employment contract with same employers and occupations or different employers and occupations, and whether it can be seen as a sustainable way of prolonging working life. We recommend future research to give more insight in these characteristics of working beyond retirement.
In conclusion, our results indicated that especially the motivation to work, physical health, the financial situation of the household, and participation in voluntary work played an important role in working beyond retirement. This supports the idea that the human agency principle of the life course perspective is useful to understand factors that impact working beyond retirement. Since social situation did not play a role, except for voluntary work, less evidence was found for the linked lives principle. Also timing seemed to be less relevant, since hardly any differences between ‘off time’ and ‘on time’ retirees were found.