Working Conditions,Psychological Well-Being of Older Employees and Retirement Timing: Results From Linked Survey and Register Data in Finland

Abstract

As populations age, supporting longer working lives becomes increasingly important. We investigate the relationships between working conditions, pre-retirement psychological well-being, and retirement timing using survey and register data on 3,261 Finnish employees who retired between 2019 and 2021. Rather than early retirement, our focus is to examine retirement at the statutory age versus retirement postponed by at least one year. Our results show that job autonomy and good leadership are associated with better psychological well-being, while psychosocial job demands are linked to poorer well-being. Poor well-being increases the likelihood of earlier retirement. Schedule flexibility and job autonomy reduce the risk of earlier retirement, whereas physical and psychosocial demands decrease the likelihood of later retirement – psychosocial demands also indirectly through their impact on well-being. Well-being mediates the effect of poor self-rated health on retirement. These results underscore the importance of working conditions and well-being in shaping retirement behavior.

Keywords

well-being self-rated health working conditions retirement careers

Introduction

Retirement ages are rising in many countries, making it increasingly important to support older employees and help them stay at work. A major aspect of this support is promoting employees’ psychological well-being, which is more heavily shaped by environmental and life circumstances than physical health. People spend a large part of their life at work, so it can be assumed that workplace factors have a considerable influence on psychological well-being. Organizational opportunities and constraints also shape retirement decisions. However, the question of exactly how working conditions impact upon psychological well-being and how these, in turn, impact upon retirement timing, remains an under-researched area.

While the connection between adverse working conditions and higher stress levels and poorer mental health in later life has been studied (e.g., Almroth et al., 2021; Leijten et al., 2015), the association between working conditions and pre-retirement psychological well-being has gone largely unexamined. Henkens and Leenders (2010) showed in the Netherlands that job demands, such as a physically demanding job and workload, were conducive to emotional exhaustion among older workers, while social support decreased the likelihood of exhaustion. Recently, Zhang et al. (2023) found in Sweden that negative psychosocial working conditions, especially low job control, were associated with poor well-being before retirement. Descriptive evidence from Finland suggests that public sector employees with poorer psychosocial working conditions (high job demands, low decision authority) more often experience psychological distress before retirement (Lahdenperä et al., 2022).

Focusing on the connection between working conditions and psychological well-being in the general working population rather than in older workers, Häusser et al. (2010) found that job resources such as higher job control and social support were associated with better well-being, whereas higher job demands were related to poorer well-being. Schütte et al. (2014) investigated the relationship between psychosocial working conditions and well-being among employees in 34 European countries. Several psychosocial work factors, such as high quantitative demands, low social support, and poor leadership, predicted poor well-being. There were almost no differences in these associations between countries or occupations.

The association between psychological well-being and retirement timing have multiple possible explanations. Enhanced well-being has been associated with higher job satisfaction and engagement (Bandyopadhyay, 2018; Khatun et al., 2022), which can motivate postponing retirement. Conversely, poorer well-being has been linked to lower job satisfaction, lower engagement, weaker organizational commitment, and adverse health outcomes (Diener & Chan, 2011; Schultz et al., 2012). Those with poor well-being may therefore be more inclined to retire earlier to change their life circumstances. Moreover, psychological well-being influences how people perceive their capacity to continue working. Even if physical health is adequate, poor well-being can lead individuals to feel less capable of continuing at work. In general, individuals with better well-being may have a more positive outlook on aging and work, seeing continued employment as meaningful, while those with poor well-being may prioritize rest and personal time, leading to earlier retirement.

Retirement transition can have a positive effect on psychological well-being (Bonsang et al., 2023; Lahdenperä et al., 2022; Latif, 2011), but the link between pre-retirement psychological well-being and retirement timing remains relatively unexplored. Most studies examining retirement timing use measures of physical health; measures of psychological health are used less often (Topa et al., 2018). It has been shown that poor (self-rated) health is one of the major determinants of early retirement (e.g., Fisher et al., 2016). Indicators of poor psychological well-being (e.g., emotional exhaustion or mental strain) have sometimes been related to earlier preferred or intended retirement, but evidence on the connection between well-being and actual retirement is limited (e.g., Geuskens et al., 2012; Henkens & Leenders, 2010; Sousa-Ribeiro et al., 2021; Virtanen et al., 2014).

The relationship between working conditions and retirement timing has received a fair amount of research, but most studies have focused on early or premature retirement rather than later retirement (Browne et al., 2019). Moreover, later retirement has typically referred vaguely to ‘retiring later’ or ‘continuing to work’, or to postponing retirement beyond statutory retirement age (Andersen et al., 2021; Carr et al., 2016; Stengård et al., 2021; van Solinge & Henkens, 2013). The role of working conditions in encouraging or discouraging individuals to work for longer after reaching statutory retirement age remains unclear. Moreover, there is no systematic evidence on the link between working conditions and later retirement: the sets of work factors examined and operationalizations of working conditions have varied widely, and the results have been mixed (e.g., Browne et al., 2019; Woźniak et al., 2022). Some studies have found evidence of a connection between working conditions and retirement timing, others have not.

Combined evidence from several studies suggest that job resources, including work-time control (Virtanen et al., 2014), higher job autonomy or decision authority (Andersen et al., 2021; Carr et al., 2016; Stengård et al., 2021), and social support (Stengård et al., 2021; van Solinge & Henkens, 2013) are associated with later retirement.¹ There is less evidence on the role of job demands, including higher physical and mental demands and time pressure, as a determinant of actual retirement (e.g., Anderssen et al., 2021; Ardito & Fleischmann, 2023; Carr et al., 2016; Nivalainen, 2022; Stengård et al., 2021; van Solinge & Henkens, 2013).

Assuming that working conditions are linked to psychological well-being, and that psychological well-being is a determinant of retirement timing, it is possible that working conditions affect retirement timing not only directly but also indirectly via psychological well-being. Henkens and Leenders (2010) found that emotional exhaustion acted as a mediator in the relationship between working conditions and retirement intentions, but as far as we know, the mediating role of psychological well-being has not been tested in the context of actual retirement. If working conditions influence retirement timing both directly and indirectly, their importance in retirement decision-making might be greater than previously thought.

Study Contribution and Scope

We base our study on the Job Demands-Resources (JD-R) model (Demerouti et al., 2001), a well-known theoretical framework that seeks to explain how working conditions affect employees’ health, well-being, and work performance. The model distinguishes between job demands and job resources. Job demands refer to aspects of a job that call for continuous physical or psychological effort (e.g., high workload) and that therefore may lead to physiological or psychological costs. Job resources, then, are aspects of a job that motivate employees and help them achieve their work goals (e.g., social support) (Bakker & Demerouti, 2017). The JD-R model posits two processes: the health impairment process and the motivational process. High job demands can lead to mental and physical exhaustion and so undermine health, psychological well-being, and work engagement, resulting in early retirement, whereas job resources support motivation and thus promote engagement and well-being, potentially encouraging individuals to work for longer (Bakker et al., 2023).

Our analysis considers a broader range of working conditions than most previous studies and uses several measures of job resources, such as employer support for continued employment, appreciation of older workers, schedule flexibility, job autonomy, and high-quality leadership, as well as job demands, such as physically and mentally demanding work, time pressure/workload, and poor atmosphere at work. There is limited research into the determinants of pre-retirement psychological well-being. Our study contributes to this body of knowledge by investigating the association between various job resource and job demand variables and pre-retirement psychological well-being.

As for retirement timing, our focus is on retiring at the statutory retirement age or postponing retirement by at least one year. The role of psychological well-being in earlier versus later retirement has received little attention so far. Therefore, unlike most studies, our investigation uses measures of both physical health and psychological well-being. Most previous studies have failed to establish a link between job demands and actual retirement, and the evidence regarding the role of job resources in retirement has been mixed. Our study offers novel evidence on the importance of working conditions as predictors of earlier versus later retirement. Finally, as far as we know, the determinants of pre-retirement well-being and retirement timing have not been investigated in a single study. Therefore, it has not been possible to investigate the indirect effect of job demands on retirement timing via psychological well-being, which we do in this study.

Figure 1 illustrates our conceptual model. Our research question is this: How do job resources and demands influence psychological well-being before retirement, and how do they relate to the timing of retirement? We test the following hypotheses:

Hypothesis 1: Job resources are related to better pre-retirement psychological well-being, whereas job demands are associated with poorer psychological well-being.

Hypothesis 2: Poorer pre-retirement psychological well-being is related to earlier retirement.

Hypothesis 3: Job resources are related to later retirement, while job demands are associated with earlier retirement.

Hypothesis 4: Job demands indirectly affect retirement timing via poorer pre-retirement psychological well-being.

Our case study is Finland. Finland has one of the fastest-aging populations in the world (Population Reference Bureau, 2020), which makes extending working careers a timely and pressing issue. Like the United States and some other countries, Finland has a flexible retirement age, allowing people to start drawing their pension within a certain age range. The statutory retirement age, which increases with each successive cohort, is the minimum age at which full pension benefits can be claimed.² Those who delay claiming their pension beyond the statutory age will receive a 4.8% increase to their pension for each subsequent year, with no upper age limit. They will also accrue a new pension at a rate of 1.5% of their earnings. As Finland is a Nordic welfare state, continuing to work longer is rather a choice than necessity.

The structure of the article is as follows: The second section describes the data and measures, the third section presents the results, and the fourth and fifth section provide the discussion and conclusion, respectively.

Data, Measures, and Method

Data

We use survey data collected by the Finnish Centre for Pensions in autumn 2022. The questionnaire was sent to 5,000 randomly selected individuals who had retired on an old-age pension between 2019 and 2021 and who were in paid employment at the end of the year preceding retirement. The individuals were to have earned at least 5,000 euros during the year before retirement. They also had to have transitioned to an old-age pension by the age of 68. The sample is representative of employees who retired between 2019 and 2021.

Responses were received from 3,378 individuals, giving a response rate of 68%. After accounting for item non-response (see Method-section), 3,216 observations remained. The dataset includes weights, and the respondents represent the survey sample by age and gender. The survey data is supplemented with register data from the Finnish Centre for Pensions (e.g., date of birth and date of retirement) and Statistics Finland (highest level of education).

The data contains information such as health, family situation, and financial circumstances. The working conditions described in the survey refer to the job held immediately prior to retirement. Many previous studies, especially those investigating retirement, measure working conditions up to several years before actual retirement (e.g., Carr et al., 2016; Nivalainen, 2022; van Solinge & Henkens, 2014).

The survey respondents’ average age was 66.2 years in 2022 (Table 1). The majority (95%) were born in 1957 or earlier. The statutory retirement age was 63 years for 17.0% of respondents, 63 years 3 months for 20.2%, 63 years 6 months for 32.2%, and 63 years 9 months for 25.6%. 5% of respondents had specific retirement rules and retired before statutory age, which was 64 years or higher.

Table 1.

Descriptive Statistics of Sample and the Outcome Variables

Variable	Value/% of respondents
Sample characteristics
Average age (2022)	66.2 years
Distribution of statutory retirement age (birth year)
63 years (1950–1954)	17.0
63 years 3 months (1955)	20.2
63 years 6 months (1956)	32.2
63 years 9 months (1957)	25.6
64 years or higher (1958 and after)	5.0
Outcome variables
Poor psychological well-being	30.1
Retirement timing
At the statutory age or earlier	39.8
1‑11 months later	26.8
1+ year later	33.4

Measures

Outcome Measures

Poor Psychological Well-Being

This measure is based on a single survey item that used a hard-to-translate Finnish term (jaksaminen) that describes aspects of energy, resilience, and the ability to cope. We refer to this here as ‘psychological well-being’: ‘How was your psychological well-being before retirement?’ The response categories were: 1 = Good; 2 = Fairly good; 3 = Moderate; 4 = Fairly poor; and 5 = Poor. Responses 3, 4, and 5 were defined as poor psychological well-being. 30.1% of respondents belonged to this group (Table 1).

According to our definition, a person with poor psychological well-being lacks energy and resilience and finds it difficult to cope when things are not going well. They also tend to feel drained and tired and feel they must put in extra effort to get things done.

Retirement Timing

Retirement timing is based on register data on the date of old-age retirement, date of birth, and statutory retirement age. First, the age of old-age retirement was calculated. Secondly, based on this figure and the statutory retirement age of the cohort in question, a variable indicating retirement timing was constructed. Retirement timing is a three-class variable indicating whether a person retired at the statutory age or earlier (39.8% of respondents), postponed retirement by 1–11 months (26.8%), or by one year or more (33.4%) (Table 1).

Independent Variables

Table 2 presents the independent and control variables, their respective shares (% of respondents) in the dataset, the coding algorithms applied, the wording of the survey questions, and the data source.

Table 2.

Job Resources and Demands and Control Variables, Their Means, Coding Algorithms, and the Wording of the Survey Questions and Data Source

Variable	Share (% of respondents)	Coding algorithm	Wording/data source
Job resources
Support for continued employment	53.6	Dummy variable. Responses 1 and 2 dichotomized	‘Employees were supported to continue until old-age retirement age.’ 1 = Strongly agree; 2 = Somewhat agree; 3 = Somewhat disagree; 4 = Strongly disagree
Appreciation of older employees	66.1	Dummy variable. Responses 1 and 2 dichotomized	‘Older employees were appreciated at the workplace.’ 1 = Strongly agree; 4 = Strongly disagree
Schedule flexibility	61.6	Dummy variable. Responses 1 and 2 dichotomized	‘My work time was flexible.’ 1 = Strongly agree; 4 = Strongly disagree
Job autonomy	71.2	Dummy variable. Responses 1 and 2 dichotomized	‘I had good opportunities to influence my work.’ 1 = Strongly agree; 4 = Strongly disagree
High quality of leadership	60.9	Dummy variable. Responses 1 and 2 dichotomized	‘My work was well managed.’ 1 = Strongly agree; 4 = Strongly disagree
Job demands
Physically demanding job	20.1	Dummy variable. Responses 1 and 2 dichotomized	‘My job was physically too demanding.’ 1 = Strongly agree; 4 = Strongly disagree
Mentally demanding job	35.7	Dummy variable. Responses 1 and 2 dichotomized	‘My job was mentally too demanding.’ 1 = Strongly agree; 4 = Strongly disagree
Time pressure/workload	52.4	Dummy variable. Responses 1 and 2 dichotomized	‘My work was too busy or the workload was too high.’ 1 = Strongly agree; 4 = Strongly disagree
Poor atmosphere at workplace	19.5	Dummy variable. Responses 3 and 4 dichotomized	‘The atmosphere at the workplace was good.’ 1 = Strongly agree; 4 = Strongly disagree
Control variables
Male	41.3	Dummy variable. 0 = Female; 1 = Male	‘What is your gender?’ 1 = Female; 2 = Male; 3 = Other/Prefer not to answer
Education		Factor variable. 1 = Basic education, up to 9 years of education; 2 = Secondary education, up to 12 years of education; 3 = Lower tertiary education, up to 15 years of education; 4 = Higher tertiary education, 16+ years of education	Register data. Highest level of education at the end of the year before retirement
Basic	10.6
Upper secondary	38.1
Lower tertiary	34.4
Upper tertiary	16.9
Public sector	48.9	Dummy variable. 0 = Private sector; 1 = Public sector	Register data. Employment sector at the end of the year before retirement
Spouse’s status		Factor variable. Response 4 = 1. Response 1 = 2. Responses 2 and 3 = 3	‘What was the main type of activity of your spouse before you retired?’ 1 = Spouse was working; 2 = Spouse was retired; 3 = Spouse was not working or retired; 4 = I have no spouse
No spouse	29.0
Spouse working	30.9
Spouse not working	40.1
Poor self-rated state of health	22.7	Dummy variable. Responses 3, 4 and 5 dichotomized	‘How was your state of health before retirement?’ 1 = Good; 2 = Fairly good; 3 = Moderate; 4 = Fairly poor; 5 = Poor
Ability to make ends meet		Factor variable. Responses 1 and 2 = 1. Response 3 = 2. Response 4 = 3. Responses 5 and 6 = 4	‘How did you make ends meet before retirement?’ 1 = With great difficulty; 2 = With difficulty; 3 = With some difficulty; 4 = Fairly easily; 5 = Easily; 6 = Very easily
With difficulty	9.2
With some difficulty	21.1
Fairly easily	35.7
Easily	34.0
Had saved for retirement	52.7	Dummy variable. Responses 1 and 2 dichotomized	‘Had you saved for retirement?’ 1 = Yes, significantly; 2 = Yes, to some degree; 3 = I had saved but not for retirement; 4 = I had not saved much
Work very important in life	39.1	Dummy variable. Response 1 dichotomized	‘My work was a very important area in my life.’ 1 = Strongly agree; 2 = Somewhat agree; 3 = Somewhat disagree; 4 = Strongly disagree
Social relations at the workplace very important	24.0	Dummy variable. Response 1 dichotomized	‘How important were social relations at the workplace before retirement?’ 1 = Very important; 2 = Fairly important; 3 = Not very important; 4 = Not at all important
Very satisfied with the amount of leisure time	17.4	Dummy variable. Response 1 dichotomized	‘How satisfied were you with your amount of leisure time before retirement?’ 1 = Very satisfied; 2 = Fairly satisfied; 3 = Somewhat satisfied; 4 = Not very satisfied; 5 = Not at all satisfied
N	3,261

Job Resources

We examine the impact of job resources using the following five single-item measures: employer’s support for continued employment; appreciation of older workers; schedule flexibility; job autonomy; and high-quality leadership. Schedule flexibility and job autonomy represent different dimensions of job control.

Higher social support and higher job control in general are typically connected with better psychological well-being (Aronsson et al., 2017; Häusser et al., 2010; Zhang et al., 2023). Likewise, high quality of leadership has been found to be a predictor of better well-being (Schütte et al., 2014).

The evidence on the connection between retirement timing and job resources is mixed. Some studies find that social support encourages later retirement, others report no connection (Carr et al., 2016; Nivalainen, 2022; Stengård et al., 2021; van Solinge & Henkens, 2014). Low recognition or appreciation increases early retirement or has no effect (Anderssen et al., 2021; Carr et al., 2016; Sousa-Ribeiro et al., 2021). Some studies indicate that flexible work-time increases later retirement, but others have found no association (van Solinge & Henkens, 2014; Virtanen et al., 2014). Job autonomy has been reported to decrease the likelihood of early retirement, but not all studies agree (Andersen et al., 2021; Carr et al., 2016; Nivalainen, 2022; Stengård et al., 2021). Poor leadership quality is sometimes associated with early retirement, but this has not been extensively researched (Thorsen et al., 2016).

Job Demands

Our model includes four single-item job demand measures: physically demanding job; mentally demanding job; time pressure/workload; and poor atmosphere at work. The latter three reflect dimensions of psychosocial demands.

Higher job demands, such as physical and emotional/mental demands, as well as time pressure and workload, have been associated with poorer well-being or emotional exhaustion (Häusser et al., 2010; Henkens & Leenders, 2010; Shütte et al., 2014). Workplace discrimination or bullying can be assumed to signal a poor atmosphere at work, and both have been linked to poor well-being (Shütte et al., 2014).

The relation between job demands and retirement timing is less clear. Typically, higher physical demands are not linked to retirement (Andersen et al., 2021; Carr et al., 2016; Stengård et al., 2021). Furthermore, many studies have failed to find evidence for the role of mental job demands or time pressure (Carr et al., 2016; Stengård et al., 2021). While the importance of the social climate or atmosphere at work to retirement timing has received some discussion, no strong support has emerged for this connection (de Wind et al., 2014; Thorsen et al., 2016).

Control Variables

As per existing literature, we include several control variables associated with either psychological well-being or retirement timing (Table 2). These include demographic and socioeconomic characteristics, as well as health, family, and financial factors.

There may be gender differences in psychological well-being as men and women are exposed to different working environments and perceptions of stress vary by gender, with older women more likely to experience emotional exhaustion (Helkavaara et al., 2011; Hsu, 2019; Marchand et al., 2018). However, no association between gender and retirement timing has been found in Finland (Nivalainen, 2022; Nivalainen, 2022).

The link between socioeconomic status and psychological well-being is not obvious. Those with a higher level of education may have more resources to cope with stress, and sometimes a lower level of education is linked to higher emotional exhaustion (Hsu, 2019). However, some studies have reported the opposite, with highly educated people experiencing emotional exhaustion more often (Helkavaara et al., 2011; Henkens & Leenders, 2010; Hulsegge et al., 2023). Regarding retirement timing, it is well established that highly educated individuals and those with higher socioeconomic status are more likely to work for longer (Nivalainen, 2022; Virtanen et al., 2014).

The association between employment sector and well-being is unclear. In Finland, public sector workers tend to retire later than others (Nivalainen, 2022).

Poorer self-rated health is associated with well-being when measured by emotional exhaustion (Hsu, 2019). Health is also central in explaining retirement timing. Better self-rated health is conducive to later retirement, while poorer health has the opposite effect (Nivalainen, 2022; van Solinge & Henkens, 2014).

Marital status is not strongly linked to well-being (Henkens & Leenders, 2010; Hsu, 2019). However, having a spouse can be an important factor in retirement, as people without a partner tend to retire later (van Solinge & Henkens, 2014). The spouse’s retirement status can also be relevant and has been linked to early retirement (Nivalainen, 2022).

Financial difficulties may be associated with psychological problems (e.g., Weich et al., 1998). However, the evidence on finances and retirement is not conclusive. Some research suggests that a better financial situation leads to later retirement (Zappalà et al., 2008), whereas others have reported the opposite (Nivalainen, 2022).

Two variables describe personal values: the importance of work in life and social relations at work. Low perceived importance of work and low sense of community, including few good friends at work, are associated with poorer well-being (Schütte et al., 2014). Those who consider work very important tend to retire later, while the importance of workplace relationships has received less research (e.g., Nivalainen, 2022).

Additionally, we investigate the role of leisure time in psychological well-being, based on the assumption that high satisfaction with leisure time signals a good work-life balance. Leisure activities are beneficial to well-being, and a poor work-life balance is linked to poorer well-being (Häusser et al., 2010; Michèle et al., 2019).

Method

The descriptive statistics of the data and the outcome variables are presented in Table 1. Table 2 introduces the independent and control variables. Individuals with missing information on covariates were included in the lowest categories for dummy variables. However, for variables describing gender, employment sector, and ability to make ends meet, those with missing information were excluded. The item non-response rate was very low (at most 2.0%, or n = 66).

Table 3 investigates the determinants of poor psychological well-being and Table 4 the determinants of retirement timing. We first ran a regression including only the independent variables (working conditions) and then a regression including the control variables. A logit model was used for psychological well-being and a multinomial logit model for retirement timing. In both cases, the results are presented as average marginal effects. The marginal effect shows the change in the probability of the event if the explanatory variable changes by one unit (e.g., if a dummy variable changes from 0 to 1).

Table 3.

Determinants of Poor Psychological Well-Being, Average Marginal Effects of Logit Models

	Specification 1	Specification 2	Specification 3
	Marginal effect (std.error)	Marginal effect (std.error)	Marginal effect (std.error)
Personal characteristics
Male		−0.042^**	−0.039^*
		(0.015)	(0.015)
Education (ref. basic)
Upper secondary		0.037	0.028
		(0.021)	(0.022)
Lower tertiary		0.043	0.027
		(0.022)	(0.023)
Upper tertiary		0.054^*	0.033
		(0.027)	(0.027)
Public sector		0.005	0.011
		(0.014)	(0.014)
Spouse’s status (ref. no spouse)
Spouse working		−0.025	−0.026
		(0.018)	(0.017)
Spouse not working		−0.004	−0.007
		(0.016)	(0.016)
Poor self-rated state of health		0.490^***	0.476^***
		(0.020)	(0.020)
Ability to make ends meet (ref. with difficulty)
With some difficulty		−0.008	−0.011
		(0.027)	(0.026)
Fairly easily		−0.022	−0.029
		(0.025)	(0.025)
Easily		−0.042	−0.042
		(0.026)	(0.026)
Work very important in life		−0.056^***	−0.048^***
		(0.014)	(0.014)
Social relations at the workplace very important		−0.032^*	−0.022
		(0.016)	(0.016)
Very satisfied with the amount of leisure time			−0.138^***
			(0.017)
Job resources
Support for continued employment	−0.030	−0.021	−0.017
	(0.018)	(0.015)	(0.015)
Appreciation of older employees	−0.005	0.002	0.002
	(0.018)	(0.016)	(0.016)
Schedule flexibility	−0.024	−0.010	−0.002
	(0.016)	(0.014)	(0.014)
Job autonomy	−0.076^***	−0.034^*	−0.033^*
	(0.020)	(0.017)	(0.016)
High quality of leadership	−0.059^**	−0.039^*	−0.038^*
	(0.019)	(0.016)	(0.016)
Job demands
Physically demanding job	0.059^**	−0.002	−0.002
	(0.020)	(0.017)	(0.017)
Mentally demanding job	0.183^***	0.150^***	0.146^***
	(0.019)	(0.017)	(0.016)
Time pressure/workload	0.126^***	0.096^***	0.083^***
	(0.017)	(0.015)	(0.015)
Poor atmosphere at workplace	0.025	−0.021	−0.019
	(0.021)	(0.017)	(0.017)
N	3,261	3,261	3,261

Notes. ^*p < .05, ^**p < .01, ^***p < .001.

Table 4.

Determinants of Retirement Timing, Average Marginal Effects of Multinomial Logit Models

	Specification 1			Specification 2
	At the earliest age	1–11 months later	1+ year later	At the earliest age	1–11 months later	1+ year later
	Marginal effect (std.error)	Marginal effect (std.error)	Marginal effect (std.error)	Marginal effect (std.error)	Marginal effect (std.error)	Marginal effect (std.error)
Personal characteristics
Poor psychological well-being				0.067^**	0.011	−0.078^***
				(0.022)	(0.021)	(0.021)
Male				0.016	−0.007	−0.009
				(0.020)	(0.018)	(0.018)
Education (ref. basic)
Upper secondary				−0.024	0.029	−0.005
				(0.030)	(0.027)	(0.026)
Lower tertiary				−0.120^***	0.052	0.065^*
				(0.031)	(0.028)	(0.028)
Upper tertiary				−0.190^***	0.026	0.160^***
				(0.035)	(0.032)	(0.033)
Public sector				−0.082^***	0.027	0.056^**
				(0.019)	(0.017)	(0.017)
Spouse’s status (ref. no spouse)
Spouse working				0.100^***	−0.007	−0.095^***
				(0.023)	(0.021)	(0.022)
Spouse not working				0.054^**	0.025	−0.079^***
				(0.020)	(0.019)	(0.020)
Poor self-rated state of health				0.034	−0.024	−0.009
				(0.023)	(0.021)	(0.022)
Ability to make ends meet (ref. with difficulty)
With some difficulty				0.069^*	−0.049	−0.020
				(0.032)	(0.032)	(0.032)
Fairly easily				0.049	−0.013	−0.036
				(0.030)	(0.031)	(0.031)
Easily				0.088^**	−0.054	−0.034
				(0.031)	(0.032)	(0.031)
Has saved for retirement				0.029	0.046^**	−0.075^***
				(0.017)	(0.016)	(0.016)
Work very important in life				−0.071^***	−0.062^***	0.130^***
				(0.018)	(0.017)	(0.018)
Social relations at the workplace very important				−0.010	−0.004	0.014
				(0.021)	(0.019)	(0.019)
Job resources
Support for continued employment	−0.002	−0.030	0.028	−0.002	−0.027	0.029
	(0.020)	(0.018)	(0.019)	(0.020)	(0.018)	(0.019)
Appreciation of older employees	0.004	−0.014	0.009	0.007	−0.011	0.004
	(0.021)	(0.019)	(0.020)	(0.020)	(0.019)	(0.020)
Schedule flexibility	−0.048^*	0.009	0.039^*	−0.054^**	0.018	0.035^*
	(0.018)	(0.017)	(0.018)	(0.018)	(0.017)	(0.017)
Job autonomy	−0.072^**	−0.001	0.072^***	−0.043^*	0.008	0.034
	(0.022)	(0.020)	(0.020)	(0.022)	(0.020)	(0.021)
High quality of leadership	0.013	0.014	−0.027	0.015	0.022	−0.037
	(0.021)	(0.019)	(0.020)	(0.020)	(0.019)	(0.019)
Job demands
Physically demanding job	0.101^***	−0.015	−0.086^***	0.049^*	0.001	−0.050^*
	(0.023)	(0.020)	(0.021)	(0.023)	(0.022)	(0.022)
Mentally demanding job	0.082^***	−0.007	−0.074^***	0.076^***	−0.019	−0.057^**
	(0.020)	(0.018)	(0.019)	(0.020)	(0.018)	(0.019)
Time pressure/workload	0.003	0.035^*	−0.038^*	0.019	0.031	−0.050^**
	(0.019)	(0.017)	(0.018)	(0.019)	(0.017)	(0.018)
Poor atmosphere at workplace	−0.003	−0.009	0.012	−0.004	−0.004	0.008
	(0.025)	(0.022)	(0.024)	(0.024)	(0.022)	(0.023)
Observations		3,261			3,261

Notes. ^*p < .05, ^**p < .01, ^***p < .001.

Next, we employed the Karlson–Holm–Breen (KHB) method (Stata Version 18: khb) to formally test whether poor psychological well-being acts as a mediator in the relationship between job demands and retirement timing. Mediation occurs when the effect of an independent variable on a dependent variable is transmitted through a third variable, known as the mediator. In our case, poor well-being is assumed to be the mediator. The KHB method provides unbiased decompositions of the total effect of job demands on retirement timing into direct and indirect effects (via the mediator) for nonlinear probability models (Breen et al., 2013). The results are presented in Table 5.

Table 5.

Total, Direct and Indirect Effect of Mentally Demanding Job, Time Pressure/Workload and Health on Postponing Retirement by 1+ Years

	1+ year later
	Reduced model (total effect)	Full model (direct effect)	Difference (indirect effect)	Indirect effect (%)
	Marginal effect	Marginal effect	Marginal effect^a	Indirect effect (%)
Mentally demanding job	−0.069***	−0.057**	−0.013**	18
Time pressure/workload	−0.057**	−0.050**	−0.007*	13
Poor self-rated state of health	−0.049*	−0.009	−0.039***	81
Poor psychological well-being		−0.078***

Notes. ^*p < .05, ^**p < .01, ^***p < .001.

Controls: the same as in specification 2 in Table 4. Difference is not necessarily exactly the difference between Reduced model and Full model due to rounding of figures.

^aStandard error of difference not known for marginal effects. The significance of the coefficient is assumed to hold.

Results

Poor Psychological Well-Being

The results of the logit models for poor psychological well-being are presented in Table 3 and reported as marginal effects. Specification 1 only includes working conditions. Among the job resource variables, job autonomy and high-quality leadership are significant and negative. As an example of interpretation, job autonomy and high-quality leadership decrease the probability of poor psychological well-being by 7.6 and 5.9 percentage points, respectively. In other words, these job resources are associated with better well-being. However, other job resource variables — support for continued employment, appreciation of older employees, and schedule flexibility — are not significantly associated with well-being. Among the job demand variables, a physically demanding job, a mentally demanding job, and time pressure/workload are significant and positive. Therefore, these job demand variables are related to poorer psychological well-being. A poor atmosphere at work is not significantly associated with well-being.

In Specification 2, control variables are added. Job autonomy and high-quality leadership still decrease the probability of poor psychological well-being, whereas the other job resource variables remain insignificant. Therefore, in terms of job resources, Hypothesis 1 was partially supported: job autonomy and high-quality leadership are associated with better psychological well-being. A mentally demanding job and time pressure/workload continue to increase the probability of poor well-being, while a physically demanding job is no longer significant. A poor atmosphere at work remains insignificant. Therefore, in terms of job demands, Hypothesis 1 was partially supported: a mentally demanding job and time pressure/workload are associated with poorer well-being.

When it comes to control variables, women are more likely than men to experience poor psychological well-being. The highly educated also have problems with well-being more often than others. If self-rated state of health is poor, the probability of poor well-being is clearly higher. Although a correlation (r = 0.5) was observed between psychological well-being and self-rated health, no collinearity was detected (VIF = 1.09). An individual’s values also contribute to their well-being. If work or social relations at work are very important to a person, the probability of poor well-being decreases.

Specification 3 is an additional specification that tests the importance of satisfaction with the amount of leisure time. If a person is very satisfied with their leisure time, poor psychological well-being is less likely. However, when this variable is added to the model, education and the importance of social relations at work are no longer significant. The explanation for this is unclear. However, it is possible that satisfaction with leisure time varies according to education and the importance of social relations, and that the leisure time variable captures that variation.

Retirement Timing

The results of the multinomial logit model for retirement timing are presented in Table 4 and reported as marginal effects. Specification 1 only includes working conditions. Among the job resource variables, schedule flexibility and job autonomy are significantly related to retirement timing. To illustrate the interpretation, schedule flexibility and job autonomy decrease the probability of retiring at the earliest age by 4.8 and 7.2 percentage points, respectively, and increase the probability of postponing retirement by at least one year by 3.9 and 7.2 percentage points. Support for continued employment, appreciation of older employees, and high-quality leadership are not associated with retirement timing. Regarding the job demand variables, both physical and mental demands increase the probability of retiring at the earliest age and decrease the probability of continuing to work for at least one more year. Time pressure/workload also decreases the probability of continuing to work for longer and increases the probability of retiring soon after the earliest age. A poor atmosphere at work is not significantly related to retirement timing.

Specification 2 adds control variables. Schedule flexibility and job autonomy continue to decrease the probability of retiring at the earliest age. Schedule flexibility also increases the probability of postponing retirement by at least one year, whereas job autonomy is no longer significant in this respect. Support for continued employment, appreciation of older employees, and high-quality leadership remain insignificant.³ Therefore, in terms of job resource variables, Hypothesis 3 is partially supported: schedule flexibility and job autonomy reduce the probability of retirement at the earliest age and in that sense are related to later retirement.

Regarding job demand variables, the results are similar to those of Specification 1. Both physical and mental demands increase the probability of retiring at the earliest age and decrease the probability of continuing to work for at least one more year. Time pressure/workload also decreases the probability of continuing to work. A poor atmosphere at work is not significant. Therefore, in terms of job demand variables, Hypothesis 3 is partially supported: physically and mentally demanding jobs and time pressure/workload decrease the probability of continuing to work longer and in that sense are related to earlier retirement.

Supporting Hypothesis 2, poor psychological well-being increases the probability of retiring at the earliest age and decreases the probability of continuing to work. The results also show that higher education increases the probability of postponing retirement and decreases the probability of retirement at the earliest age. Public sector employees are more likely to postpone retirement and less likely to retire as soon as possible. Those with no spouse are more likely to continue working longer, while those with a spouse are more likely to retire at the earliest age, regardless of the spouse’s work status. Interestingly and somewhat surprisingly, health is not significantly associated with retirement timing.

Those who can make ends meet easily or with little difficulty are more likely than others to retire at the earliest age. Those who have saved for retirement are less likely to continue working longer and more likely to retire soon after reaching the minimum retirement age. If work is very important to a person, they are more likely to continue working and less likely to retire as soon as possible. On the other hand, the importance of social relations in the workplace is not associated with retirement timing.

Mediation Analysis

Working conditions can indirectly affect retirement timing via poor psychological well-being. The potential suspects are the job demand variables that were significant in explaining poor well-being: a mentally demanding job and time pressure/workload. We used the Karlson–Holm–Breen (KHB) method to test whether poor well-being mediated the relationship between a mentally demanding job and time pressure/workload, and retirement timing. The outcome of interest was postponing retirement by at least one year (‘1+ year later’).

The results of the mediation analysis are presented in Table 5. The first column shows the total effect of a mentally demanding job and time pressure/workload on postponing retirement, i.e., the effect when psychological well-being is not controlled for. The total effect of both job demand variables is significant, and both are negatively associated with the probability of postponing retirement. The second column shows the direct effect of a mentally demanding job and time pressure/workload on postponed retirement, i.e., the effect when psychological well-being is controlled for. The direct effect of both job demand variables is significant, as already seen in Table 4, but smaller than the total effect. The third column presents the indirect effects of a mentally demanding job and time pressure/workload on postponing retirement, i.e., the effects mediated through poor well-being. Both indirect effects are significant and negative, thus supporting Hypothesis 4. The indirect effects of a mentally demanding job and time pressure/workload account for 18% and 13% of the total effect, respectively.

Table 4 showed that poor self-rated health is not a significant determinant of retirement timing. Since health is usually a key factor in retirement timing, we suspected that its effect might be partly mediated through psychological well-being. We tested this and present the results in Table 5. The total effect of poor health on postponed retirement is significant and negative, albeit smaller in magnitude and less significant than that of poor well-being. When well-being is controlled for, the effect of health on retirement disappears. The indirect effect of poor health on the probability of postponed retirement is significant and negative. As much as 81% of the total effect of poor health on postponed retirement is mediated via poor psychological well-being.

Discussion

This study investigated the interplay between working conditions, pre-retirement psychological well-being, and retirement timing among older employees in Finland. As retirement ages are rising in many countries, it is important to understand how working conditions influence the well-being of older employees and how well-being and working conditions are associated with retirement timing.

Our results show that job resources, particularly job autonomy and high-quality leadership, are associated with better psychological well-being before retirement. These findings lend support to the JD-R model, which posits that job resources enhance well-being. Our results align with the work of Zhang et al. (2023) on job control among older individuals and the work of Schütte et al. (2014) on leadership quality among working-age employees. This suggests that these resources remain critical in the later stages of working life, implying that organizational practices that promote autonomy and leadership skills may serve as protective factors against poor well-being before retirement.

Furthermore, our results indicate that psychosocial demands, particularly a mentally demanding job and time pressure/workload, are associated with poorer well-being, hence supporting the strain hypothesis of the JD-R model. Our finding is in line with Henkens and Leenders (2010), who found that workload increases the emotional exhaustion of older workers, but contradicts Schütte et al. (2014), who found that emotional demands do not predict poorer well-being among working-age employees. This suggests that there may be age-related differences in coping capacity when it comes to emotional job demands, and that job redesign or task redistribution should be considered for older employees to reduce mental overload. An important additional finding is that high satisfaction with the amount of leisure time significantly decreases the risk of poor well-being, suggesting that leisure activities have a positive impact on well-being (Michèle et al., 2019). This indicates that policies promoting flexible work arrangements or gradual retirement could enhance well-being by allowing more time for leisure.

We also found that poor psychological well-being increases the probability of earlier retirement. This adds to the limited evidence on the impact of well-being on retirement timing. While a few earlier studies have found that psychological well-being is related to retirement preferences, there is less evidence on its importance in determining actual retirement (e.g., Henkens & Leenders, 2010; Sousa-Ribeiro et al., 2021). The result underscores the importance of early identification and support for employees experiencing psychological distress.

According to our findings, job control – schedule flexibility and job autonomy – decrease the probability of earlier retirement. The result for schedule flexibility supports the findings of Virtanen et al. (2014), and the result concerning autonomy is consistent with Carr et al. (2016) and Stengård et al. (2021). Our finding supports the idea that control over one’s work is an important motivator for older workers to remain in the workforce and indicates that organizations seeking to retain older workers should emphasize flexible scheduling and participatory decision-making practices. Physical and mental demands and time pressure/workload reduce the probability of later retirement. This represents a novel contribution, as many others (e.g., Carr et al., 2016; Stengård et al., 2021; van Solinge & Henkens, 2014) have failed to establish a link between physical or psychosocial demands, and actual retirement. Our findings indicate that high job demands may accelerate labor market exit among older workers. Consequently, reducing physical and mental strain and managing workload may be effective strategies for supporting the continued employment of aging employees.

The KHB method supported the idea that poor psychological well-being partially mediates the negative effect of job demands (mental demands and time pressure/workload) on later retirement. This indirect effect highlights the importance of these job demand variables in the retirement process. As far as we are aware, this is a novel finding which suggests that interventions targeting job demands could delay retirement indirectly by preserving well-being.

We found that poor self-rated health does not significantly affect retirement timing when psychological well-being is taken into account. This finding challenges the traditional view that health is a key determinant of retirement timing (e.g., Fisher et al., 2016; Nivalainen, 2022). It is important to note that both state of health and psychological well-being were self-reported using similar response categories, suggesting that the differences observed in the strength of the associations between health measures and retirement timing cannot be explained away by differences in assessment methods. Moreover, since state of health and well-being were asked as separate items, it is likely that respondents interpreted state of health as referring primarily to physical aspects of health. The moderate correlation observed between these two measures indicates that they are distinct constructs, even though they are conceptually related. This means that they share some underlying factors (e.g., emotional states, stress) but also capture distinct aspects of individual experience. The lack of multicollinearity between the two measures allows us to examine their independent contributions to retirement timing, highlighting the importance of considering both physical and psychological dimensions when studying retirement behavior. Our results imply that poor psychological health may contribute more to retirement decisions than poor physical health.

In terms of limitations, our analysis uses single-item measures of working conditions. These seem to capture important aspects of the work environment, but multi-item scales might provide better coverage of job resources and demands. In addition, health and work characteristics were evaluated retrospectively. However, studies have shown that retrospective evaluations of working conditions predict well-being or exit from the labor market (e.g., Sundstrup et al., 2018; Zhang et al., 2023). Moreover, all respondents had worked until retirement, so the people in our sample are likely in better physical and psychological health than the older workforce in general. This could underestimate the relationship between working conditions, psychological well-being, and retirement timing. However, it is precisely this group that has the option of extending their careers beyond pensionable age. Therefore, the results reveal important information about the role of job resources and demands in the decision to continue working for longer.

Conclusion

Enhancing job resources and altering job demands can be seen as vital to improving the quality of work environments for older employees. Our results also suggest that even though monitoring the health of employees is important, it might be equally important to follow their psychological well-being. This would facilitate pre-emptive measures before employees start to experience problems. Early screening and discussions with older employees should be an integral part of good leadership.

The results of this study are important in the context of promoting extended working lives. As retirement ages continue to rise, the impact of working conditions on psychological well-being later in life will likely become an increasingly prominent topic. Promoting job resources and limiting job demands can improve employee well-being, particularly in the context of late retirement. At the organizational level this could involve a focus on age-sensitive leadership skills, increasing older employees’ decision-making authority, redesigning jobs to reduce mental and physical workloads, and adjusting schedules. Workplace modifications to alter job demands can delay retirement both directly and indirectly, through their impact on well-being.

Footnotes

ORCID iD

Satu Nivalainen

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

This study uses individual-level survey data collected by the Finnish Centre for Pensions combined with register data from the Finnish Centre for Pensions. Due to legal restrictions and data protection regulations, the author does not have the permission to make sensitive personal data available.*

Notes

Author Biography

Satu Nivalainen holds a PhD in Economics and works as an economist in the Research Department at the Finnish Centre for Pensions. Her research focuses on old-age retirement, retirement intentions, continued employment, partial old-age pension, and the behavioral effects of pension reforms. E-mail: satu.nivalainen@etk.fi.

References

Almroth

Hemmingsson

Sorberg Wallin

Kjellberg

Burstrom

Falkstedt

(2021). Psychosocial working conditions and the risk of diagnosed depression: A Swedish register-based study. Psychological Medicine, 52(15), 1–9. https://doi.org/10.1017/S003329172100060X

Andersen

L. L.

Thorsen

S. V.

Larsen

Sundstrup

Boot

C. R. L.

Rugulies

(2021). Work factors facilitating working beyond state pension age: Prospective cohort study with register follow-up. Scandinavian Journal of Work, Environment & Health, 47(1), 15–21. https://doi.org/10.5271/sjweh.3904

Ardito

Fleischmann

(2023). Health, working conditions and retirement. In Burnay

Ogg

Krekula

Vendramin

(Eds.), Older workers and labour market exclusion processes. Life course research and social policies, vol 14. Springer. https://doi.org/10.1007/978-3-031-11272-0_9

Aronsson

Theorell

Grape

Hammarström

Hogstedt

Marteinsdottir

Skoog

Träskman-Bendz

Hall

(2017). A systematic review including meta-analysis of work environment and burnout symptoms. BMC Public Health, 17(1), 264. https://doi.org/10.1186/s12889-017-4153-7

Bakker

A. B.

Demerouti

(2017). Job demands–resources theory: Taking stock and looking forward. Journal of Occupational Health Psychology, 22(3), 273–285. https://doi.org/10.1037/ocp0000056

Bakker

A. B.

Demerouti

Sanz-Vergel

A. I.

(2023). Job demands–resources theory: Ten years later. Annual Review of Organizational Psychology and Organizational Behavior, 10, 25–53. https://doi.org/10.1146/annurev-orgpsych-120920-05393

Bandyopadhyay

(2018). Well-being and its determinants. In Gatchel

R. J.

Schultz

I. Z.

(Eds.), Handbook of occupational health and wellness (pp. 123–145). Springer. https://doi.org/10.1007/978-1-4614-4839-6_10

Bonsang

Garrouste

Perdrix

(2023). Retirement and well-being. In Zimmermann

K. F.

(Ed.), Handbook of labor, human resources and population economics (pp. 1–14). Springer. https://doi.org/10.1007/978-3-319-57365-6_391-1

Breen

Karlson

K. B.

Holm

(2013). Total, direct, and indirect effects in logit and probit models. Sociological Methods & Research, 42(2), 164–191. https://doi.org/10.1177/0049124113494572

10.

Browne

Carr

Fleischmann

Xanthopoulou

Head

(2019). Working conditions and retirement: A systematic review. Journal of Occupational Health, 61(5), 345–360. https://doi.org/10.1007/s11368-023-03632-6

11.

Carr

Hagger-Johnson

Head

Shelton

Stafford

Stansfeld

Zaninotto

(2016). Working conditions as predictors of retirement intentions and exit from paid employment: A 10-year follow-up of the English longitudinal study of ageing. European Journal of Ageing, 13(1), 39–48. https://doi.org/10.1007/s10433-015-0357-9

12.

Demerouti

Bakker

A. B.

Nachreiner

Schaufeli

W. B.

(2001). The job demands-resources model of burnout. Journal of Applied Psychology, 86(3), 499–512. https://doi.org/10.1037/0021-9010.86.3.499

13.

de Wind

Geuskens

G. A.

Ybema

J. F.

Bongers

P. M.

van der Beek

A. J.

Burdorf

(2014). Health, job characteristics, skills, and social and financial factors in relation to early retirement: Results from a longitudinal study in the Netherlands. Scandinavian Journal of Work, Environment & Health, 40(2), 186–194. https://doi.org/10.5271/sjweh.3393

14.

Diener

Chan

M. Y.

(2011). Happy people live longer: Subjective well-being contributes to health and longevity. Applied Psychology: Health and Well-Being, 3(1), 1–43. https://doi.org/10.1111/j.1758-0854.2010.01045.x

15.

Fisher

G. G.

Chaffee

Sonnega

(2016). Retirement Timing: A Review and Recommendations for Future Research. Work, Aging and Retirement, 2, 230–261. https://doi.org/10.1093/workar/waw001

16.

Geuskens

G. A.

Oude Hengel

K. M.

Koppes

L. L. J.

Ybema

J. F.

(2012). Predictors of the willingness and the ability to continue working until the age of 65 years. Journal of Occupational and Environmental Medicine, 54(5), 572–578. https://doi.org/10.1097/JOM.0b013e318248057a

17.

Häusser

J. A.

Mojzisch

Niesel

Schulz-Hardt

(2010). Ten years on: A review of recent research on the job demand-control (-support) model and psychological well-being. Work & Stress, 24(1), 1–35. https://doi.org/10.1080/02678371003683747

18.

Helkavaara

Saastamoinen

Lahelma

(2011). Psychosocial work environment and emotional exhaustion among middle-aged employees. BMC Research Notes, 4, 101. https://doi.org/10.1186/1756-0500-4-101

19.

Henkens

Leenders

(2010). Burnout and older workers’ intentions to retire. International Journal of Manpower, 31(3), 306–321. https://doi.org/10.1108/01437721011050594

20.

Hsu

H.-C.

(2019). Age differences in work stress, exhaustion, well-being, and related factors from an ecological perspective. International Journal of Environmental Research and Public Health, 16(1), 50. https://doi.org/10.3390/ijerph16010050

21.

Hulsegge

Eekhout

van de Ven

H. A.

Burdorf

Oude Hengel

K. M.

(2023). Educational inequalities in self-rated health and emotional exhaustion among workers during the COVID-19 pandemic: A longitudinal study. International Journal of Environmental Research and Public Health, 20(1), 1234. https://doi.org/10.1007/s00420-022-01931-y

22.

Khatun

Bharti

Tiwari

(2022). Effects of work stress on psychological well-being and job satisfaction: A review. In Gatchel

R. J.

Schultz

I. Z.

(Eds.), Handbook of occupational health and wellness (pp. 555–572). Springer. https://doi.org/10.1007/978-1-4614-4839-6_27

23.

Lahdenperä

Stenholm

Vahtera

Virtanen

Pentti

Myllyntausta

(2021). Psychological distress during the retirement transition and the role of psychosocial working conditions and social living environment. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 77(1), 135–148. https://doi.org/10.1093/geronb/gbab054

24.

Latif

(2011). The impact of retirement on psychological well-being in Canada. The Journal of Socio-Economics, 40(3), 373–380. https://doi.org/10.1016/j.socec.2010.12.011

25.

Leijten

F. R.

van den Heuvel

S. G.

van der Beek

A. J.

Ybema

J. F.

Robroek

S. J.

Burdorf

(2015). Associations of work-related factors and work engagement with mental and physical health: A 1-year follow-up study among older workers. Journal of Occupational Rehabilitation, 25(1), 86–95. https://doi.org/10.1007/s10926-014-9525-6

26.

Marchand

Blanc

M.-E.

Beauregard

(2018). Do age and gender contribute to workers’ burnout symptoms? Occupational Medicine, 68(6), 405–411. https://doi.org/10.1093/occmed/kqy088

27.

Michèle

Guillaume

Alain

Nathalie

Claude

Kamel

(2019). Social and leisure activity profiles and well-being among the older adults: A longitudinal study. Aging & Mental Health, 23(1), 77–83. https://doi.org/10.1080/13607863.2017.1394442

28.

Nivalainen

(2022). From plans to action? Retirement thoughts, intentions and actual retirement: an eight-year follow-up in Finland. Ageing and Society, 42(1), 112–142. https://doi.org/10.1017/S0144686X20000756

29.

Population Reference Bureau . (2020, March 23). Countries With the Oldest Populations in the World. Population Reference Bureau. https://www.prb.org/resources/countries-with-the-oldest-populations-in-the-world/

30.

Schultz

P. P.

Ryan

R. M.

Niemiec

C. P.

Legate

Williams

G. C.

(2012). Mindfulness, work climate, and psychological need satisfaction in employee well-being. In Gatchel

R. J.

Schultz

I. Z.

(Eds.), Handbook of occupational health and wellness (pp. 555–572). Springer. https://doi.org/10.1007/978-1-4614-4839-6_27

31.

Schütte

Chastang

J.-F.

Malard

Parent-Thirion

Vermeylen

Niedhammer

(2014). Psychosocial working conditions and psychological well-being among employees in 34 European countries. International Archives of Occupational and Environmental Health, 87(8), 897–907. https://doi.org/10.1007/s00420-014-0930-0

32.

Sousa-Ribeiro

Bernhard-Oettel

Sverke

Westerlund

(2021). Health- and age-related workplace factors as predictors of preferred, expected, and actual retirement timing: Findings from a Swedish cohort study. International Journal of Environmental Research and Public Health, 18(5), 2746. https://doi.org/10.3390/ijerph18052746

33.

Stengård

Leineweber

Virtanen

Westerlund

Wang

H.-X.

(2022). Do good psychosocial working conditions prolong working lives? Findings from a prospective study in Sweden. European Journal of Ageing, 19(4), 677–688. https://doi.org/10.1007/s10433-021-00672-0

34.

Sundstrup

Hansen

Å. M.

Mortensen

E. L.

Poulsen

O. M.

Clausen

Rugulies

Møller

Andersen

L. L.

(2018). Retrospectively assessed psychosocial working conditions as predictors of prospectively assessed sickness absence and disability pension among older workers. BMC Public Health, 18(1), 149. https://doi.org/10.1186/s12889-018-5047-z

35.

Thorsen

S. V.

Rugulies

Longaard

Borg

Thielen

Bjorner

J. B.

(2016). The association between psychosocial work environment, attitudes towards older workers (ageism), and planned retirement. International Archives of Occupational and Environmental Health, 89(6), 891–900. https://doi.org/10.1007/s00420-016-1147-4

36.

Topa

Depolo

Alcover

C. M.

(2018). Early retirement: A meta-analysis of its antecedent and subsequent correlates. Frontiers in Psychology, 8, 2157. https://doi.org/10.3389/fpsyg.2017.02157

37.

van Solinge

Henkens

(2014). Work-related factors as predictors in the retirement decision-making process of older workers in the Netherlands. Ageing and Society, 34(9), 1551–1574. https://doi.org/10.1017/S0144686X13000330

38.

Virtanen

Oksanen

Batty

G. D.

Ala-Mursula

Salo

Elovainio

Pentti

Lybäck

Vahtera

Kivimäki

(2014). Extending employment beyond the pensionable age: A cohort study of the influence of chronic diseases, health risk factors, and working conditions. PLoS One, 9(2), Article e88695. https://doi.org/10.1371/journal.pone.0088695

39.

Weich

Lewis

Jenkins

S. P.

(1998). Income inequality and the prevalence of common mental disorders in Britain. The British journal of psychiatry: The Journal of Mental Science, 178(3), 222–227. https://doi.org/10.1192/bjp.178.3.222

40.

Woźniak

Siebielec

Bojarszczuk

Gałązka

Urbaniak

(2022). Working conditions and retirement: A systematic review. Journal of Occupational Health, 64(2), 123–145. https://doi.org/10.1007/s11368-023-03632-6

41.

Zappalà

Depolo

Fraccaroli

Guglielmi

Sarchielli

(2008). The relationship between job insecurity and job performance: The mediating role of organizational commitment. Journal of Occupational and Organizational Psychology, 81(1), 101–120. https://doi.org/10.1108/PR-05-2016-0120

42.

Zhang

Yang

Pan

K. Y.

(2023). Association between psychosocial working conditions and well-being before retirement: A community-based study. Psychology, Health & Medicine, 29(3), 574–588. https://doi.org/10.1080/13548506.2023.2274316