Employment Status and Life Satisfaction Among Older Adults: Disentangling the Gendered Effects of Entering and Exiting Employment

Abstract

According to the “three boxes of life” concept proposed by Riley et al. (1994), individuals typically progress through three primary activities throughout their lives: education, work, and leisure. This model proposes that older people enjoy their leisure time after retiring in their later years. However, contemporary society does not follow a strict linear sequence between work and retirement among older adults (Smyer & Pitt-catsouphes, 2007). Several studies have suggested that employment status is crucial for strengthening social ties, which contributes to successful aging (Nilsen et al., 2022; Rowe & Kahn, 2015). Despite being over the age of 65, a growing number of older adults continued to remain employed. For instance, in South Korea (hereafter, Korea), 36.9% of the population over 65 years old was employed in 2020 (Y. Lee, 2020). Furthermore, 74.7% of wage earners aged 65 and older are working in new jobs post-retirement (Ji & Park, 2024). Despite these trends, little is known about how transitions into and out of employment affect life satisfaction among older adults in Korea.

The experience of older adults entering and exiting employment may differ significantly from that of younger adults, who often prioritize career growth over financial security. Entering employment provides them with critical financial benefits and opportunities to build or maintain social connections, both essential for their psychosocial well-being. This is supported by evidence indicating that employment status contributes to both economic stability and sustained social connections, which are significant for successful aging (Doan et al., 2024; Rowe & Kahn, 2015). Specifically, economic difficulties among older populations are closely linked to distress, and inadequate retirement preparation often leads to higher levels of depressive symptoms and increased feelings of loneliness (Chai, 2023; Ju et al., 2017 ). Furthermore, financial concerns can intensify feelings of profound loneliness (Drost et al., 2024; Kim & Park 2022). This process is critical for enhancing life satisfaction by fostering economic independence, especially vital in contexts with inadequate social security systems (Aquino et al., 1996; Ham & Hong, 2017).

Moreover, the Latent Deprivation model (Jahoda, 1982) lends support to the idea that work provides not only financial benefits but also latent functions. These benefits contribute to improved psychological health and social support, which in turn lead to higher levels of life satisfaction (Murayama et al., 2022). In line with this, the Activity Theory (Havighurst, 1961) emphasizes that social interaction and active engagement are central to enhancing life satisfaction during the aging process (Kim, 2023). Evidence indicates that social networks become increasingly important to individuals as they age and face challenges such as the loss of family members (Park & Kim, 2024; Yoon et al., 2022). For older adults, maintaining an active and socially connected lifestyle through employment can significantly enhance their quality of life and overall well-being (Newman & Zainal, 2020). This highlights the value of employment not just for financial reasons but as a critical element in ensuring a fulfilling and satisfying life in later years.

Exiting employment may have a lesser impact on reducing life satisfaction compared with the positive impact of entering it, as several factors can help offset the associated reductions. For older adults, leaving the workforce often brings immediate relief from work-related stressors, such as physical demands and workplace pressures, particularly in labor-intensive or high-stress occupations, leading to short-term improvements in well-being (Doan et al., 2024). Additionally, age discrimination, which fosters exclusion, undervaluation of skills, and increased pressure to justify one’s capabilities, can erode workplace satisfaction for older workers. Exiting such environments reduces stress and provides relief, partially mitigating the loss of life satisfaction (Roscigno et al., 2022). Moreover, in Korea, older adults face high poverty rates and an underdeveloped pension system, leading many to depend on financial support from their adult children (Kim & Cook, 2011). This culturally ingrained obligation for adult children to support their parents further buffers the negative effects of exiting employment (Yoon & Kim, 2021).

The asymmetric effect of entering and exiting employment may vary significantly by gender, especially in societies with strong traditional norms. In Korea, although financial necessity drives both men and women to enter employment, the impact on life satisfaction differs by gender, partly due to the internalization of societal roles (Kim, 2021). For men, even when participation is involuntary, entering employment often has a pronounced positive effect on life satisfaction. According to Role Theory (Parsons, 1942), men often view work as central to their identity, offering both a sense of purpose and a critical source of social interaction (Jang et al., 2009). This is particularly important for older men, who may have limited social networks outside of work (Um et al., 2020). Employment provides not only financial benefits but also essential social connections, which are key to maintaining psychological well-being (Kawachi et al., 2010). For these men, entering employment restores both a sense of utility and vital opportunities for social engagement, enhancing their overall life satisfaction.

The experience of older women, however, reveals a markedly different pattern, with exiting employment often mitigating potential reductions in life satisfaction. Women often enter employment out of financial necessity rather than a voluntary desire to re-engage in professional life (van Solinge et al., 2021), typically taking on lower-paying roles centered around interpersonal relationships, such as caregiving and service jobs (Vartanian & McNamara, 2002). Although these roles may offer some fulfillment, they generally do not enhance life satisfaction to the same extent as more autonomous or financially rewarding positions, which are more common for older men. Moreover, women tend to face lower job satisfaction due to work-life conflicts and challenges related to their family life course (Beaufils et al., 2023; Linehan & Walsh, 2000), making employment status less rewarding. As a result, for older women, the psychological benefits of leaving employment—such as reduced stress and a renewed focus on personal or family life—may outweigh the potential advantages of remaining employed.

The contrasting effects of employment transitions between genders can be understood within differing social and cultural contexts. Men often derive significant benefits from entering employment due to their traditional breadwinner role and limited alternative social networks (Um et al., 2020) In contrast, women frequently find greater satisfaction in exiting employment, a difference partly explained by their typically larger and more diverse social networks outside of work (McLaughlin et al., 2010; Reitzes et al., 1995). Unlike men, women tend to maintain strong social relationships and engage in informal social activities, providing alternative sources of life satisfaction beyond employment (McLaughlin et al., 2010). These findings underscore how gender-specific social resources and cultural expectations shape the relationship between employment transitions and life satisfaction in later life. Women’s well-established social networks act as a buffer against the potential negative effects of leaving employment while enhancing the benefits of exiting potentially stressful work environments.

Despite the theoretical framework supporting the potential asymmetric effects of entering and exiting employment on life satisfaction among older adults, there remains a significant gap in studies that thoroughly explore this possibility. Utilizing the nationally representative longitudinal data, this study employed an asymmetric fixed effects model to separately estimate the effects of entering and exiting employment on life satisfaction. Asymmetric fixed effects models challenge the implicit symmetry assumption of traditional regression models by allowing for distinct psychological responses to entering and exiting employment (Allison, 2019). This method is essential for understanding the complex and potentially varied ways in which transitions into and out of employment influence life satisfaction, especially among older adults who may face unique challenges in the labor market. Moreover, this study conducts gender-stratified analyses to examine heterogeneity by gender in the asymmetric effects, revealing significant gender differences that could inform the development of targeted interventions and policies aimed at enhancing life satisfaction.

Data and methods

Data

This study utilized data from the Korean Longitudinal Study of Ageing (KLoSA), a nationally representative study initiated in 2006, which includes a multistage stratified cluster sample of 10,254 individuals aged 45 and over from Korean households Lee (2020). Conducted biennially by the Korean Employment Information Service, the survey assesses socioeconomic, demographic, and health-related attributes of middle-aged and older individuals in Korea. To ensure the representativeness of the respondents to the general population in Korea, they were geographically selected based on the Korean Population and Housing Census. This study utilized longitudinal data collected over 12 years, from 2006 (Wave 1) to 2018 (Wave 7). All participants provided informed consent, and the data were anonymized before being made available in a publicly accessible database. Ethical approval was exempted because the study was a secondary analysis of publicly available data.

From an initial total of 20,137 observations over seven waves, involving participants aged 65 years or older at the baseline wave, 140 observations were excluded from the analysis due to missing values in key variables. Consequently, 19,997 observations were analyzed, including 4,122 participants in Wave 1, 3,483 in Wave 2, 3,073 in Wave 3, 2,774 in Wave 4, 2,478 in Wave 5, 2,189 in Wave 6, and 1,878 in Wave 7. Due to the unbalanced nature of the panel data structure, the final analytic sample size—defined as the number of respondents with valid data points in any of the waves—totaled 4,146. This figure exceeds the number of participants in the initial wave. The difference stems from some respondents being absent in the initial wave due to missing data, only to return in later waves with complete datasets.

Variables

Dependent variable

Our dependent variable is the level of satisfaction with one’s overall life circumstances. To measure life satisfaction, we used the survey question, “How satisfied are you with your overall life?” Respondents rated their satisfaction on a scale from 0 to 100, where higher scores indicate greater satisfaction. The use of a single-item measure for life satisfaction is validated by prior studies, which have demonstrated its comparability to multi-item scales (Cheung & Lucas, 2014).

Independent variable

Employment status is the key predictor in this study. To determine one’s involvement in employment, we used the question: “Are you currently working for income? Here, ‘work’ refers to either being employed at a job or running your own business.” Using longitudinal data, we track changes in employment status between consecutive waves. Entering employment is defined as a transition from not working to working between waves, whereas exiting employment represents a transition from working to not working.

Control variable

This study incorporated both time-constant and time-varying control variables. Time-constant variables included gender, educational attainment, and number of children. Education was categorized into four levels: elementary or lower, middle school, high school, and college or higher. Time-varying control variables included age, marital status, household size, logged household income, homeownership, region of residence, number of chronic diseases, limitations in activities of daily living (ADLs) and instrumental activities of daily living (IADLs). Marital status was determined by current marital status (single, widowed, divorced, or separated). Household income was categorized into quartiles, and homeownership was classified based on whether individuals were owner-occupiers. Region of residence was divided into three categories: large city (metropolitan cities), small city (nonmetropolitan urban cities), and rural areas. The number of chronic diseases summed up doctor-diagnosed conditions such as hypertension, diabetes, mellitus, cancer or a malignant tumor, chronic lung disease, cerebrovascular disease, arthritis or rheumatoid arthritis, psychological disease, liver disease, and/or prostatic disease. Limitations in ADLs were assessed by counting the number of seven specific activities requiring assistance: dressing oneself, washing one’s face, bathing oneself, eating, going out of the room, using a toilet, and regulating urine and bowel movements. IADL limitations were assessed by the number of tasks requiring assistance, including grooming, performing household chores, cooking, doing laundry, going out, using transportation, shopping, managing money, using the telephone, and taking medication.

Statistical Analysis

To investigate the relationship between employment status and life satisfaction among older adults, this study begins with pooled ordinary least squares models:

$y_{it}$ represents the life satisfaction of individual i at time t. The independent variable, $E{P}_{it}$⁠, denotes the employment status of individual i at time t. $X_i$ is a vector of time-constant control variables, while $Z_{it}$ is a vector of time-varying control variables. ${\epsilon }_{it}$ is the idiosyncratic error term, varying across individuals and time. Pooled ordinary least squares estimates might be biased due to unobserved individual-level heterogeneity. For instance, unobserved factors such as ability, personality traits, occupational history, and genetics could influence both employment status and life satisfaction in old age.

The application of standard fixed effects models addresses potential confounding or selection bias stemming from unobserved time-invariant factors at the individual level. This study estimates standard fixed effects models as follows:

Individual fixed effects, denoted by $v_i$⁠, account for individual-specific and time-invariant characteristics. Time-constant covariates (⁠$X_i$⁠) are omitted from the model as they are absorbed by $v_i$⁠. The coefficient $\beta$ reflects the within-person association between employment status and life satisfaction. Although standard fixed effects models provide more robust estimates than pooled ordinary least squares models, they assume symmetrical effects of the independent variable, implying that the magnitude of the effect of entering employment is the same as—but in the opposite direction to—the effect of exiting employment. If the effects of entering and exiting employment are asymmetric, traditional fixed effects models could obscure the true effects, potentially leading to the erroneous conclusion that employment status has no effect on life satisfaction among older adults.

To accommodate the possibility of differing impacts for entering and exiting employment, we adopt an asymmetric fixed effects model (Allison, 2019). In this model, we decompose the independent variable into positive and negative components:

$E{S}_{it}^{+}=E{S}_{it}-E{S}_{it-1}$ if $(E{S}_{it}-E{S}_{it-1})>0$⁠, otherwise 0.

$E{S}_{it}^{-}=-(E{S}_{it}-E{S}_{it-1})$ if $(E{S}_{it}-E{S}_{it-1})<0$⁠, otherwise 0.

$E{S}_{it}^{+}$ represents entering employment, and $E{S}_{it}^{-}$ denotes exiting employment. For $t=1$⁠, both $E{S}_{it}^{+}$ and $E{S}_{it}^{-}$ are set to be 0 due to the absence of previous wave observations. The accumulation variables are then defined as:

$D_{it}^{+}$ represents the total positive changes in employment status up to time $t$⁠, while $D_{it}^{-}$ represents the total negative changes in employment status up to time $t$⁠. This method enables the separate estimation of the effects of entering and exiting employment within fixed effects models. The preferred model is structured as follows:

In this model, ${\beta }^{+}$ and ${\beta }^{-}$ are coefficients representing the effects of entering and exiting employment, respectively. To formally assess the asymmetry of these effects, we conduct a Wald test for each model comparing ${\beta }^{+}$ and ${\beta }^{-}$⁠.

Furthermore, to investigate potential differences in these asymmetric effects between women and men, we conduct gender-stratified analyses and include gender interaction models to test the statistical significance of these differences.

Results

Table 1 provides summary statistics of a sample of 4,146 participants, stratified by gender. Approximately 58% of the respondents were female. The average age was 72.98, with a standard deviation of 6.30. Around 17% of the participants had attained at least a high school education. The mean level of life satisfaction was 57.00, with a standard deviation of 22.77. About 15% of the participants were currently employed. Significant gender differences were observed in key variables. Men reported higher life satisfaction levels (59.20) compared with women (55.42) and were more likely to be employed (28% vs. 6%).

Table 2 shows the distribution of changes in employment status across survey waves. As expected, about 93% of the respondents maintained their employment status from one wave to the next. However, nearly 7% of respondents either entered or exited employment across waves, a rate that remained stable as the sample cohort aged.

Table 3 presents the estimated association between employment status and life satisfaction, comparing results from pooled ordinary least squares, standard fixed effects, and asymmetric fixed effects models (Columns 1, 2, and 3, respectively). The pooled ordinary least squares models in Column 1 reveal a significant positive association between employment status and life satisfaction (b = 1.423, p < .001). In Column 2, accounting for unobserved individual heterogeneity reduces the association by about 65%, rendering it statistically insignificant. This suggests that much of the observed association between employment status and life satisfaction may be confounded by unobserved heterogeneity at the individual level. However, it is worth noting that the null association observed in the standard fixed effects model could be misleading due to its assumption that the effects of entering and exiting employment are symmetric.

To relax the assumption of symmetric effects of employment status, we estimated asymmetric fixed effects models (Column 3). The results from Column 3 reveal that the effects of employment status on life satisfaction are asymmetric. Specifically, entering employment is associated with an increase in life satisfaction (b = 1.719, p < .05), while exiting employment shows no significant association with life satisfaction. A Wald test (⁠${\beta }^{+}=-{\beta }^{-}$⁠) confirmed that these asymmetric patterns were statistically significant (p < .05). These results indicate that the traditional fixed effects models, which assume symmetrical effects, could have obscured the positive impact of entering employment and the null effect of existing employment.

As with most longitudinal studies involving older adults, the KLoSA faces the issue of sample attrition—i.e., the sample size decreases across waves due to mortality and study dropouts. To assess the potential impact of attrition-related selection bias on the relationship between employment status and life satisfaction, we performed a sensitivity analysis using inverse probability weighting (IPW; Metten et al., 2022). First, we estimated the predicted probability of an individual’s continued participation in the study—specifically, their chance of surviving and not withdrawing from the study. From these probabilities, we derived analytical weights that are inversely proportional to the likelihood of remaining in the study and alive. We then applied these weights to our analysis to examine the association between employment status and life satisfaction. Our results confirm that using IPW to adjust for attrition bias does not alter the main outcomes or conclusions of our study (Supplementary Table 1 in Supplementary Material).

Table 4 presents asymmetric fixed effects estimates of the associations between employment status and life satisfaction, stratified by gender (Column 1 for men and Column 2 for women). The gender-stratified analyses reveal distinct patterns in the asymmetric effects of employment status on life satisfaction. For men, entering employment is associated with increased life satisfaction (b = 3.125, p < .01), while exiting employment has no significant association with life satisfaction. In contrast, for women, exiting employment is positively associated with life satisfaction (b = 2.679, p < .01), while entering employment has no significant impact. The marginal significance of these asymmetric effects for both genders may be due to increased standard errors resulting from the reduced sample size. To assess gender differences in the impact of entering and exiting employment on life satisfaction, we present results from a gender interaction model. As shown in Column 3, the statistically significant interaction terms confirm the gender differences.

Discussion

Using a nationally representative longitudinal dataset from Korea, this study estimated the asymmetric effects of entering and exiting employment on life satisfaction among older adults aged 65 and above. Additionally, it conducted gender-stratified analyses to explore heterogeneity in these effects based on gender. Traditional fixed effects models suggested that employment status is not significantly associated with life satisfaction among older adults. However, the asymmetric fixed effects models revealed that entering employment is associated with an increase in life satisfaction, while exiting employment has no association with life satisfaction. This indicates that traditional regression models may obscure the true asymmetric relationship between employment status and life satisfaction.

This study’s findings reveal important asymmetries in the relationship between employment status and life satisfaction, with entering employment showing positive effects, while exiting shows no significant association. Entering employment in old age often represents a voluntary choice, driven by a desire to remain active and contribute to society. In contrast, exiting employment can occur for a variety of reasons, including voluntary retirement or involuntary job loss. The positive effect of employment entry on life satisfaction likely reflects the dual benefits of financial security and enhanced psychological well-being, as employment provides a sense of purpose and opportunities for social engagement (J. Kim & Yoon, 2022; Jun, 2020). On the other hand, the null effect of exiting employment may stem from the diverse experiences of retirement. While some individuals may struggle with the loss of work-related identity and structure, others might embrace retirement as an opportunity for leisure, personal growth, or family time, resulting in no overall impact on life satisfaction.

This study further investigated gender differences in the association between employment status and life satisfaction among older adults. The findings indicate that the asymmetric effects of employment transitions—i.e., statistically significant positive effects of entering employment and the null effects of exiting—are particularly driven by men. This aligns with the previous research, particularly within the context of Confucian culture, where economic activity holds different meanings across genders. In many East Asian societies, including Korea, men are traditionally viewed as the primary breadwinners. This societal expectation can make employment status more crucial for their sense of identity and life satisfaction (Jang et al., 2009). For instance, research in Korea has shown that older men are more inclined than older women to pursue economic self-sufficiency (Yeom, 2019). Moreover, older men often have less robust social networks outside of work compared with older women (Um et al., 2020). Given these psychosocial differences, entering employment can act as a protective factor for life satisfaction among older men, helping them maintain both economic stability and social identity.

The results of this study reveal a contrasting pattern in the asymmetric effects of employment transitions for older women: entering employment is not associated with increased life satisfaction, while exiting is positively linked to life satisfaction. This gender disparity highlights deeply entrenched structural inequalities in Korea’s labor market and society. Older women often face intersectional discrimination based on both age and gender, resulting in limited access to quality employment opportunities. Despite comparable educational attainment, older women are disproportionately concentrated in precarious, low-wage roles in the caregiving and service sectors (Vartanian & McNamara, 2002). These positions typically lack job security, benefits, and opportunities for advancement, especially compared with the roles typically available to men (Lu et al., 2023). Furthermore, older women’s employment status is often driven by financial necessity rather than choice, particularly among those with insufficient pension coverage due to career interruptions for family caregiving. Balancing employment with traditional gendered expectations, such as caregiving for grandchildren, can exacerbate role conflict and diminish well-being (J. H. Kim, 2018; Linehan & Walsh, 2000).

This study has a few limitations that warrant consideration. First, the KLoSA data lack detailed information on the nature of participants’ employment, such as types of employment. For instance, Lee and Kim (2017) found that individuals engaged in precarious employment, unpaid family work, self-employment, and retirement report poorer physical health compared with those in nonprecarious employment. Future research should consider these factors to provide deeper insights into the findings of this study. Second, the assessment of life satisfaction was based on a single item, which may not fully capture the complexity of this construct. Previous research has shown that women not in paid employment (i.e., retired women or family caregivers) report low levels of job satisfaction but high levels of family satisfaction (Kang et al., 2024). Future studies would benefit from employing more comprehensive measures that encompass multiple dimensions of life satisfaction. Third, this study did not consider the cumulative effect of job changes over time. Frequent job changes tend to negatively affect the well-being of employed older adults (Ng & Feldman, 2013). Including this factor in future research would help elucidate the impact of career stability or instability on life satisfaction among older populations.

Despite these limitations, this study makes several important contributions to the existing literature on employment status and life satisfaction among older adults in Korea. First, this study sheds light on the effects of employment status among the older population—a growing issue as the number of older adults in the labor market continues to rise. By focusing on the older population, the study provides insights that are particularly relevant in the context of Korea’s aging society. Second, this is among the first studies to demonstrate the asymmetric effects of entering and exiting employment on well-being, indicating that policies should be tailored to meet the unique needs of older adults during these transitions, rather than adopting a one-size-fits-all approach. Third, this study contributes to the ongoing debate on gender heterogeneity in the effects of employment status, offering evidence that such differences are pronounced, especially in the East Asian context. The stark contrast in the asymmetric patterns of the effects of entering and exiting employment underscores the importance of understanding the gender role in shaping well-being in old age.

The present study offers several insights that could guide policy development aimed at improving life satisfaction among older adults. First, it is crucial to promote employment among older men. Although reemployment has been shown to positively affect well-being (Carlier et al., 2013), transitioning back into employment often becomes more challenging with age, as age-related biases can significantly hinder older job seekers (Charni, 2022). Ageism not only limits employment opportunities but also prevents older adults from fully leveraging the benefits that employment can provide (Jin & Baumgartner, 2023). Policies should encourage employment for older men, highlighting the positive effects while ensuring access to meaningful employment and combating ageism. In addition, support mechanisms should be established for older women, who often experience improved well-being after exiting employment. Policies that enhance working conditions for those who remain employed are essential, as the gendered asymmetric effects of employment transitions underscore the importance of job satisfaction among older women (Prakash et al., 2022). Therefore, implementing policies tailored to meet the distinct needs of both men and women is crucial for protecting their well-being in later life.

Funding

None.

Conflicts of Interest

None.

Data Availability

The KLoSA data are available at https://survey.keis.or.kr/eng/klosa/klosa01.jsp with the permission of the Korea employment Information Service. Analytic methods and materials specific to the current study are available upon request from the corresponding author. The current study was not preregistered with an analysis plan in an independent, institutional registry.

Author Contributions

J. Kim and S. Park contributed equally to this research.

References