Introduction
Labor-force participation over the life course is gendered and influences social relationships of adults by creating the opportunity to establish social networks outside an individual’s household (Schmitz et al., Reference Schmitz, Naegele, Frerichs and Ellwardt2023). Over time, these work-related ties might grow in value, turning colleagues into friends and confidants and a source of support (Cozijnsen et al., Reference Cozijnsen, Stevens and Theo2010; Loe and Johnston, Reference Loe and Johnston2016). Labor market inactivity, on the contrary, may lead to smaller networks that are characterized by a stronger orientation toward family ties (Patacchini and Engelhardt, Reference Patacchini and Engelhardt2016). Adults who are embedded in large and diverse networks in later life show better mental health (Santini et al., Reference Santini, Koyanagi, Tyrovolas, Mason and Haro2015), less cognitive decline (Cohn-Schwartz et al., Reference Cohn-Schwartz, Levinsky and Litwin2021; Ellwardt et al., Reference Ellwardt, Van Tilburg and Aartsen2015), and reduced mortality risk (Litwin and Shiovitz-Ezra, Reference Litwin and Shiovitz-Ezra2006). Whereas multiple studies have highlighted how social networks influence labor market participation (Denaeghel et al., Reference Denaeghel, Mortelmans and Borghgraef2011; Granovetter, Reference Granovetter1995; Granovetter, Reference Granovetter1977), little work has been done on the reverse direction. Thus, a research gap exists on how labor market participation shapes social networks in later life and especially following retirement.
As labor market participation over the life course differs between social groups, those being less active in the labor market might be disadvantaged when trying to establish supportive and meaningful social networks. Taking on a gendered perspective seems especially fruitful since networks differ between men and women, and the prevalent gendered life courses have proven to result in women being less involved in the labor market, especially in older birth cohorts (Moen, Reference Moen, Shanahan, Mortimer and Kirkpatrick2016; Schmitz et al., Reference Schmitz, Naegele, Frerichs and Ellwardt2023). In addition, there are differences in regard to access to and characteristics of work-related networks (e.g. emotional closeness, friendship ties) between men and women (Campbell, Reference Campbell1988; Ibarra, Reference Ibarra1993). Therefore, chances of “turning a colleague into a friend” might be gendered, resulting in varying network compositions that can act as a source of support and help in older age. Thus, the current study aims to add to research looking at the effect of one’s work history on social ties and relationships of older retired people. Here we are especially interested in differences in regard to gender and levels of labor market integration.
Personal relationships and gender
Social relationships are a fundamental aspect of life for older adults. Personal social networks are conceptualized as the collection of social relations that affect an individual’s life and from which one might receive support. They are often composed of close family and friends (Litwin et al., Reference Litwin, Levinsky and Schwartz2020) but might also include contacts made in the context of one’s work history (Dahlin et al., Reference Dahlin, Kelly and Moen2008). The convoy model of social relationships states that close relationships are based on emotional attachment and move with individuals throughout the life course (Kahn and Antonucci, Reference Kahn, Antonucci, Baltes and Brim1980). More peripheral relationships are based on role requirements associated with participation in specific social fields and are present in people’s lives out of convenience (Antonucci et al., Reference Antonucci, Ajrouch and Birditt2014). The model divides social relationships into structural and functional aspects. Functional aspects, such as emotional closeness, are more strongly associated with well-being (Santini et al., Reference Santini, Koyanagi, Tyrovolas, Mason and Haro2015; Schwartz and Litwin, Reference Schwartz and Litwin2017). Structural aspects, such as network size and composition, are also meaningful for well-being and also for cognitive health and the attainment of resources (Cornwell, Reference Cornwell2011; Rafnsson, et al., Reference Rafnsson, Shankar and Steptoe2015; Sharifian et al., Reference Sharifian, Kraal, Zaheed, Sol and Laura2019).
The convoy model identifies gender as a central aspect affecting social relationships (Antonucci et al., Reference Antonucci, Ajrouch and Birditt2014). Concordantly, a wide body of literature indicates that men and women approach social relationships differently. Women typically maintain larger, more emotionally supportive and diverse personal networks compared to men (Ajrouch et al., Reference Ajrouch, Fuller, Akiyama and Antonucci2018; Cornwell, et al., Reference Cornwell, Schumm and Laumann2008; Stevens and Van Tilburg, Reference Stevens and Van Tilburg2011). This gap continues and even widens with age (Fischer and Beresford, Reference Fischer and Beresford2015; Schwartz and Litwin, Reference Schwartz and Litwin2018). For example, a study focusing on older Europeans found that men had smaller networks, which included a higher proportion of family members over time (Schwartz and Litwin, Reference Schwartz and Litwin2018). Men also tended to consider their spouse as an emotionally meaningful tie, more than women (Dahlin et al., Reference Dahlin, Kelly and Moen2008).
Employment histories and social networks
Discussions of the effects of work experiences on social relationships in later life should be placed within the context of lifelong work histories (Price and Dean, Reference Price and Dean2009). Nevertheless, simply knowing whether an individual was employed at a particular point in time is less useful than understanding the duration and pattern of their labor-force participation throughout life. The life course perspective draws attention to the life trajectories and transitions into and out of social roles that condition opportunities to form and maintain relationships (Settersten, Reference Settersten2015). This perspective would emphasize that the experiences during later life are related to earlier life course events, such as gendered employment patterns (Elder Jr, Reference Elder1994; Schmitz et al., Reference Schmitz, Naegele, Frerichs and Ellwardt2023). Additionally, individual life courses are interwoven as people’s lives are linked with those of others (“linked-lives”) (Bengtson et al., Reference Bengtson, Elder, Putney and Johnson2012; Settersten, Reference Settersten2015). These “links” may derive from one’s immediate family, but lives can also be linked with professional contacts such as colleagues or other business contacts (Henning and Lieberg, Reference Henning and Lieberg1996).
The opportunity to link one’s life with people from the workplace might be structured by overall labor market integration – the duration of years spent working and the number of jobs one obtained over the life course. Whereas labor market inactivity limits the exposure, having many different jobs leads to meeting a wide array of individuals. In addition, working for more years overall could also lead to meeting more new people and to form meaningful relationships that last over time (Zarankin and Kunkel, Reference Zarankin and Kunkel2019). For example, adults who were employed at age 33 reported more support compared to those who were unemployed (Matthews et al., Reference Matthews, Stansfeld and Power1999). These experiences can also be gendered, for example, men are more likely to incorporate work-related contacts into their networks (Dahlin et al., Reference Dahlin, Kelly and Moen2008), whereas women discuss important work-related matters more often with family members (Renzulli et al., Reference Renzulli, Aldrich and Moody2000).
Regarding social connectedness, retirement itself, as a major life course event, has proven to have a lasting impact, although studies find mixed effects. Some show that retirement leads to a reduction in the share of colleagues and friends and an increase in the share of family members (Comi et al., Reference Comi, Cottini and Lucifora2022), although not all retirees are found to disengage from work-related ties (Cozijnsen et al., Reference Cozijnsen, Stevens and Theo2010). These studies, however, focus on the effects of retirement itself. Albeit, the creation and maintenance of social ties is a lifelong process, as the convoy model stresses. Thus, it is imperative to examine how they can be shaped by employment throughout the life course.
Currently, limited research considers the long-term effect that employment histories of adults have in relation to their late-life relationships. Using the convoy model, the current study stipulates that adults who worked for more years and in more jobs had more opportunities to “collect” convoy members from different workplaces, although the quality in regard to emotional depth of these relationships may vary. Furthermore, people who are more involved in the workforce may be used to frequent social interactions, therefore designing their social lives as more diverse and active to mimic their work lives (Price and Nesteruk, Reference Price and Nesteruk2015).
Employment histories may also affect the relationship with family members. It is possible that people who were involved to a larger extent in the labor market have had less time and opportunities to invest in their close family. The idea of work interfering with family has been widely recognized and studies have highlighted the effect labor market involvement can have on parental time, family involvement, and work–family conflict (Byron, Reference Byron2005; Cho and Allen, Reference Cho and Allen2012; Michel et al., Reference Michel, Kotrba, Mitchelson, Clark and Baltes2011). Such a pattern could manifest in less close kin relationships in later life.
With reference to the life course perspective, gender differences should be taken into account since men and women differ both in their social relationships (as described above) and in their involvement in the workforce. Despite an increasing convergence of gender roles in recent times, life courses can follow a “gendered path,” when women assume more household and care responsibilities and showcase a lower and more interrupted labor market activity as well as an earlier retirement timing than men (Moen, Reference Moen2001; Moen, Reference Moen, Shanahan, Mortimer and Kirkpatrick2016; Williams and Umberson, Reference Williams and Umberson2004). Men tend to work for more years as they are less likely to take breaks for childcare and family care and to change their place of employment more frequently (Steiber and Haas, Reference Steiber and Haas2012). Feminist scholars have pointed out that these persistent gendered work courses are resulting in “new social risks” for women while working for longer years can benefit women, for example, in terms of their financial situation (Geyer et al., Reference Geyer, Welteke, Resources, Geyer and Welteke2021) or health (Cai, Reference Cai2010).
However, less is known about the effects of life course employment on personal networks in later life. In this regard, women who were continuously working until their retirement were found to engage more in volunteer activities and less in solitary activities compared to women whose involvement in the workforce was discontinued during their life (Price and Dean, Reference Price and Dean2009). This may stem from continuously employed women maintaining relationships with former coworkers (Loe and Johnston, Reference Loe and Johnston2016). However, these studies focused only on women, and thus research is lacking that compares men and women. Gender differences could emerge not only due to different levels of involvement in the labor force, but also men and women who were similarly involved could show gender differences due to gendered ways of “collecting” and (emotionally) investing in work ties, resulting in ties transferring differently into older age. While women who are involved in the labor market could collect more friends over their life, women might also be impacted to a larger extent by their work histories in relation to their family ties. Due to role expectations, they could be expected to a larger degree to spend time and provide care to their families. Violating such expectations, due to working more, could have stronger consequences for their relationships with close family. Men might not face similar consequences, due to lower expectations from them to spend time and care for their families.
To summarize, this study will examine whether the employment histories of men and women, namely years in employment and number of jobs, are associated with their social relationships in later life. This advances existing research in several ways: first, by utilizing retrospective data, we are able to apply a life course perspective to social networks in retirement, linking social connectedness in older age to earlier life course events and individual employment histories. Secondly, by employing a gender perspective, we are able to pay attention to the effects of gendered work courses on social relationships.
We will test the following hypotheses:
-
1. Working for longer years and in more jobs will be associated with larger social networks, particularly with more friendship ties in the social networks.
-
2. Working for longer years and in more jobs will be associated with worse relationships with family members, namely one’s spouse and children, and with lower emotional closeness with the network.
-
3. Women will have larger social networks, which will be more likely to include their friends and children and less likely to include their spouse, compared to men.
-
4. The associations between employment histories and personal networks (hypotheses 1 and 2) will be stronger among women.
Data and methods
Participants and procedure
The study is based on the Survey of Health, Aging, and Retirement in Europe (SHARE). The survey collects information on a variety of domains in a representative sample of community-dwelling adults aged 50 and above and their spouses of any age (Börsch-Supan et al., Reference Börsch-Supan2013). Questionnaires are administered in the respondents' house by trained interviewers, using a computer-assisted personal interviewing program, following the participants' informed consent. The current study combines data collected in 2015 and 2017. The 2015 data collection wave was used since it collects information on social networks. The 2017 wave, called SHARELIFE, gathered retrospective information on respondents’ life course, including questions about employment history. Eighteen countries participated in both waves and were included in the study: Austria, Germany, Sweden, Spain, Italy, France, Denmark, Greece, Switzerland, Belgium, Israel, Czech Republic, Poland, Luxembourg, Portugal, Slovenia, Estonia, and Croatia.
The analytical sample consisted of adults aged 50+. We included respondents who participated in both the 2015 and 2017 waves and had information on the study variables. The final analytical sample numbered 18,478 participants. Attrition analysis indicated that of the participants in the 2015 wave, those who had full information were (p < .05) younger, more likely to be women, more educated, were in better financial, mental and physical health, and were more likely to have worked for more years and in a larger number of jobs. They had a larger social network, reported higher emotional closeness, and were more likely to mention friends as confidants. They did not differ in terms of mentioning their spouse or children as confidants.
Measures
Social networks
Social networks were assessed using a name-generator approach, which maps individuals' social milieu from their perspective. Respondents are asked to name up to seven persons with whom they discuss important matters and are subsequently asked to provide more information about these “confidants” and about the relationship with them (Schwartz and Litwin, Reference Schwartz and Litwin2018). The study devised five outcome variables based on this measure. The size of the network summed the number of persons named (0–7). The types of persons in the network were assessed using three dummy variables indicating the citation (yes/no) of a spouse, child, and friend in the network. The spouse and child variables included only those who had a spouse or children, respectively. Emotional closeness was probed by asking respondents “How close do you feel to (confidant’s name)?” with answers ranging from “Not very close” (1) to “Extremely close” (4). The score on this measure was the average rate of emotional closeness with confidants.
Employment history
The information was taken from the job episodes panel, a dataset generated from SHARELIFE (Brugiavini et al., Reference Brugiavini, Orso, Genie and Naci2019). The job episodes panel is a longitudinal dataset, including information regarding respondents’ employment situation for every year from the time they entered the labor market. It has been shown that the share of missing observations in the job episodes panel is quite low, with missing data on employment events being retrieved from other information provided at SHARELIFE (Brugiavini et al., Reference Brugiavini, Orso, Genie and Naci2019). Since the outcome social networks data used in this study were collected in 2015, the study only included employment history up to 2015. This information was used to create summary measures of years in employment and number of jobs that lasted for 6 months or longer.
Gender was measured as a binary variable of men and women.
Covariates
The covariates included sociodemographic and health variables that were previously found to be associated with late-life social networks (Cornwell and Laumann, Reference Cornwell and Laumann2015; Litwin and Stoeckel, Reference Litwin and Stoeckel2014; Schwartz and Litwin, Reference Schwartz and Litwin2017). Age was used as a continuous variable, and education was measured as a binary variable (below secondary education/secondary education and above). We also measured the number of children (continuous variable) and whether the respondent provided care to anyone within the household (yes/no). Financial adequacy was measured using a question about the extent to which the respondent’s household is able to make ends meet, with response options ranging from “with great difficulty” (1) to “easily” (4). Self-assessed health was measured on a 5-point Likert scale. Physical health was also measured by the number of mobility, arm function, and fine motor limitations (range: 0–10). Depressive symptoms were examined via the Euro-D scale for late-life depressive symptoms (Prince et al., Reference Prince1999), composed of 12 yes/no questions about symptoms experienced in the past month (range: 0–12).
Data analysis
Data analysis showed descriptive data of the sample characteristics and a bivariate comparison between men and women, using chi-square or t-tests. The main analysis consisted of regression analyses, with the different aspects of the networks as the outcome measures. Ordinary least squares (OLS) regressions were used for social network size and mean emotional closeness with the network (Lumley et al., Reference Lumley, Diehr, Emerson and Chen2002). Logistic regressions were used for the three binary measures of citing a spouse, child, and friend within the network. Each regression predicted the outcome variables using employment history and gender in the first model and in the second model added the covariates. The second stage of analysis in each regression was an interaction of employment history with gender. The final models present the interactions that emerged as significant.
Results
Table 1 shows the sample characteristics and a bivariate comparison of men and women. The sample consisted of 53% women. Women reported larger social networks, compared to men, and were more likely to cite a child and friend as confidants. Men were more likely to cite their spouse as a confidant. The participants reported having been employed for an average of 37 years and having worked in over two jobs. Women reported fewer years in employment and were employed in fewer jobs. Women were less likely to have secondary or higher education, had less children on average, reported worse financial status, worse self-rated health, and more mobility limitations compared to men.
Table 1. Sample characteristic and bivariate comparison between men and women
*p < .05, **p < .01, ***p < .001.
Table 2 presents OLS regression models predicting social network size with employment histories and gender. The first model examined the main effects of employment histories and gender. It showed that women reported larger social networks, and adults who worked for more years and in more jobs also reported larger networks. The second model added the covariates and found similar trends. The third model examined the interaction of gender with employment history, with a significant interaction between gender and number of jobs. A simple slope analysis of the interaction indicated that among women there was a positive association between number of jobs and social network size (β = 0.09, p < .001). This association between number of jobs and social network size was seen among men as well, albeit it was smaller (β = 0.04, p < .001). Figure 1a shows a graphical representation of the interaction. While we follow the guidelines of Lumley et al. (Reference Lumley, Diehr, Emerson and Chen2002), which indicated that linear regression analyses are robust for any distribution of data in large samples (such as our sample), we conducted a supplementary Poisson regression analysis (available upon request) to predict social network size, since it could be considered as a count variable. This additional analysis also found the interaction of gender with number of jobs to be significant.
Figure 1. (a) and (b) The interaction between gender and number of jobs over the life course in relation to social network size and to the odds of mentioning a friend as a social network (SN) member.
Table 2. OLS regression models predicting social network size with employment histories and gender
B = coefficient, β = standardized coefficient; Models 2 and 3 control for age, education, number of children, providing care, financial status, self-rated health, mobility limitations, depressive symptoms, and country effects (not shown). *p < .05, **p < .01, ***p < .001.
Table 3 shows logistic regression models that predict the presence of three categories of social network members in the networks – a spouse, children, and friends. The first and second models showed that women were less likely, compared to men, to mention a spouse among their social network members. Adults who worked in more jobs were also less likely to mention their spouse as a confidant. These results were also seen after adding the covariates. The interactions of employment history with gender were not significant. The next outcome variable was having children in the social networks. Women were twice as likely, compared to men, to mention their children in the network. Adults who worked for more years were also more likely to cite children. This was seen both without and with the covariates. The interaction terms were not significant.
Table 3. Logistic regression models predicting the presence of a spouse, children, and friends in the social networks, with employment histories and gender
SN = social networks. OR = odds ratio; Models 2 and 3 control for age, education, number of children, providing care, financial status, self-rated health, mobility limitations, depressive symptoms, and country effects (not shown). *p < .05, **p < .01, ***p < .001.
The next model examined mentioning of a friend in the network as the outcome. It shows that women had over twice the likelihood of mentioning friends, as was the case for adults who worked in more jobs over their life course. The interactions of gender with number of jobs were significant. The association between number of jobs and citing friends in the social network was positive for both men and women, but it was stronger among women. Figure 1b shows a graphical presentation of this interaction effect. We carried out an additional interaction analysis with the R package modglm (Mccabe et al., Reference Mccabe2022) since nonlinear models require special considerations that have to be taken into account when interpreting interaction effects in generalized linear models. This package defines interactions as the change in a marginal effect of one variable as a function of change in another variable with the use of partial derivatives and discrete differences to quantify these effects. We found that the average interaction effect across observations was significant (B = 0.014, SE = 0.003, 95% CI [0.007, 0.020]), and it was significant for 98.6% of the sample.
Table 4 presents OLS regression models predicting emotional closeness with one’s social network. The first model predicted emotional closeness with one’s social network members. The results found no gender difference in emotional closeness, and this result remained after adding the covariates. Adults who had a larger number of jobs over their life course reported lower emotional closeness with their social networks, and this association remained after adding the covariates. Those who worked for more years reported higher emotional closeness, although this became nonsignificant after adding the covariates. The interaction of gender with years of employment was significant. A simple slope analysis of the interaction indicated that among men there was a positive association between years in employment and emotional closeness with the social network (β = 0.04, p < .01). This association was not significant among women (β = 0.00, p > .05).
Table 4. OLS regression models predicting emotional closeness, with employment histories and gender
B = coefficient, β = standardized coefficient; Models 2 and 3 control for age, education, number of children, providing care, financial status, self-rated health, mobility limitations, depressive symptoms, and country effects (not shown).*p < .05,**p < .01,***p < .001.
Discussion
Drawing on a life course perspective, the current study sets out to explore the associations of employment history with social networks in later life, and whether these associations differ by gender. By doing so, the study helps to better understand the composition and coming to be of networks in older age that play an important role in older people’s well-being and are an important pillar of social support. Our results show a complex, gendered, pattern. Adults who worked in more jobs had overall better structural characteristics of their networks – reported larger social networks in later life and were more likely to include their children and friends within those close networks, although they were less likely to include their spouse. On the other hand, working in more jobs was related to lower emotional closeness with the network. These results varied between men and women; women who were involved to a larger degree in the labor market over their life course had larger social networks in later life and were more likely to include friends in these networks. Among men, on the other hand, working for more years was related to higher emotional closeness with the network.
Our results show that adults who worked in a larger number of jobs had larger social networks and were more likely to mention their children and friends as close confidants, in accordance with our first hypothesis. Thus, their networks had better structural characteristics and included more diverse sources of support (Antonucci et al., Reference Antonucci, Ajrouch and Birditt2014). A potential explanation is that employment could be related to a more active life (Barnett and Hyde, Reference Barnett and Hyde2001), and these adults could have maintained such high levels of activity also in their social lives. The inclusion of both friends and family members indicates the diversity of ties, which can offer more varied types of interactions and activities, with potentially beneficial health implications (Cohn-Schwartz et al., Reference Cohn-Schwartz, Levinsky and Litwin2021; Fiori et al., Reference Fiori, Antonucci and Cortina2006; Park et al., Reference Park2015). Adults who worked in more workplaces may have been in more contexts in which they could meet new friends and could have collected more members to their “convoy” over the years (Antonucci et al., Reference Antonucci, Ajrouch and Birditt2014). Some of these nominated friends might have started as colleagues and became friends over time. Although the percentage of network members who are defined as a colleague or ex-colleague is low (2%), it is possible that these colleagues are currently considered to be more friends than colleagues. Alternatively, since the participants are retired, such ex-colleagues might not be considered as colleagues anymore.
These results partly contradict our hypothesis that extensive employment histories will be related to worse family relationships. However, our findings could be explained by indications that multiple roles due to involvement in the labor market and with one’s family could be beneficial to family ties (Barnett and Hyde, Reference Barnett and Hyde2001). Although previous studies focused on the effects of retirement itself on network composition in later life (Comi et al., Reference Comi, Cottini and Lucifora2022), this study applies a life course perspective and shows the effects of lifelong work events instead of focusing on a single event.
On the other hand, extensive employment histories were related to worse functional aspects of the networks, which were deemed less emotionally close and were also less likely to include one’s spouse. It is possible that the expansion of one’s network over the life course and the inclusion of less close confidants such as friends came at the expense of the emotional depth of the network, as such ties are often less close compared to one’s spouse (Litwin and Stoeckel, Reference Litwin and Stoeckel2014). Thus, it is possible that the expansion of the network over the work course could have some disadvantages in terms of the network quality. Moreover, it is possible that working in more jobs over time meant having a less meaningful role as a partner, with harmful implications for the spousal relationship (Barnett and Hyde, Reference Barnett and Hyde2001).
Gender differences indicate that, in accordance with our hypothesis, women had larger social networks and were more likely to name children and friends as confidants, while men were more likely to mention their partners (Fischer and Beresford, Reference Fischer and Beresford2015; Schwartz and Litwin, Reference Schwartz and Litwin2018). Our study adds to the literature by pointing to gender differences in the associations of employment history with later life social relationships, in accordance with our fourth hypothesis. These differences indicate that women may have “benefited” from working in more jobs, especially in terms of the structural aspects of their relationships. That is, among women, working in more jobs was more strongly related to larger social networks, which include friends. Women who are more involved in the labor market could be used to more social interactions and this tendency might have been reflected in their social lives (Price and Dean, Reference Price and Dean2009). Women may also have a stronger social motive for working (Thrasher et al., Reference Thrasher, Zabel, Wynne and Baltes2016) and could focus more on creating social ties in the workplace and maintaining these ties as friendships (Loe and Johnston, Reference Loe and Johnston2016). Women who are involved in the labor market can experience better mental health (Wan et al., Reference Wan, Antonucci, Birditt and Smith2018), which may also lead to greater social involvement. Working may help women form multiple roles and interests, which can translate into richer social lives that include varied confidants (Barnett and Hyde, Reference Barnett and Hyde2001). On the other hand, women who were less involved in the labor market might have adhered to more traditional gender roles of being active within the household (Eggers et al., Reference Eggers, Grages, Pfau-Effinger and Och2020) and were less inclined to form relationships with friends who are outside of their family.
Men, on the other hand, showed advantages of their work histories in terms of the emotional aspects of their social networks; men who were employed for more years reported better emotional closeness with their network compared to women. These results can attest to the different social needs of men and women, with men potentially perceiving their social milieu as better when going through more varied work environments. Future research should better understand the underlying mechanisms of these trends.
The current study has some limitations. The outcome measure of close social networks disregards weaker ties. While close ties are meaningful to later life health (Cornwell and Laumann, Reference Cornwell and Laumann2015; Schwartz and Litwin, Reference Schwartz and Litwin2017), weaker ties are also beneficial (Huxhold et al., Reference Huxhold, Antonucci, Fiori, Webster and Arbor2020), especially since adults may have kept in touch with colleagues who were not defined as close friends. An additional limitation is that we had no information on the social networks of participants when they entered the workforce. Thus, our results could also reflect the other direction of influence that sociable adults participate more in the workforce. Social networks can impact labor-force participation, for example, networks can help gain employment or advance in the labor market (Hess et al., Reference Hess, Naegele and Mäcken2020; Woehler et al., Reference Woehler, Cullen-Lester, Porter and Frear2021). Furthermore, the study does not go into much detail regarding workplace-related characteristics such as company size and type, as these were not available in the retrospective data. It can be assumed that, for example, work-related networks in small- and medium-sized companies differ from those in large enterprises or family-run businesses. Additionally, it should be noted that some respondents were not included in the final analysis due to missing information and they differed in several attributes from those who were included. We also note that the effects found in this study are small, but they are of potential theoretical interest. It would be desirable for our study results to be repeated by other researchers with different data and analytical procedures, such as qualitative methods, which could analyze underlying mechanisms in more detail.
To sum up, this study contributes to the literature in several ways. It adopts a life course gendered perspective which can contribute to a greater understanding of social development in later life. Our study adds to the understanding of what impacts social networks in later life, which were recently shown to be related to mental health resilience (Hopper et al., Reference Hopper, Best, Wister and Cosco2023), living conditions (van der Velpen et al., Reference van der Velpen2022), and time perspective (Soylu and Ozekes, Reference Soylu and Ozekes2022). We add an examination of life course factors, in addition to the more prevalent focus on later life factors. Our gendered outlook shows that men and women differ in the effects of work histories, suggesting that women are more advantageous in regard to the structural aspects of their networks and men could “benefit” in terms of the functional aspects. These results extend previous research on the employment histories and social milieu of women by examining such trends among both genders and looking at structural as well as functional aspects of networks. Policymakers and practitioners should be made aware that labor market inclusion over the life course has implication on social networks in later life. It might be especially worthwhile to support the involvement of women in the labor market to enhance their social ties in later life. In increasingly aging societies, social embeddedness may be worth promoting to establish sources for support of aging individuals.
Conflict of interest
None.
Description of authors’ roles
E. Cohn-Schwartz was responsible for the statistical design of the study and for carrying out the statistical analysis and wrote the paper. L. Naegele assisted the design of the study, and with writing and editing the article.
Acknowledgments
This paper uses data from the generated Job Episodes Panel (DOI: 10.6103/SHARE.jep.710). The Job Episodes Panel release 7.1.0 is based on SHARE Waves 3 and 7 (DOIs: 10.6103/SHARE.w3.710, 10.6103/SHARE.w7.710). This paper uses data from SHARE Waves 6 and 7 (DOIs: 10.6103/SHARE.w6.710, 10.6103/SHARE.w7.710). The SHARE data collection has been funded by the European Commission through FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3: RII-CT-2006-062193, COMPARE: CIT5-CT-2005-028857, SHARELIFE: CIT4-CT-2006-028812), FP7 (SHARE-PREP: GA N°211909, SHARE-LEAP: GA N°227822, SHARE M4: GA N°261982, DASISH: GA N°283646), and Horizon 2020 (SHARE-DEV3: GA N°676536, SHARE-COHESION: GA N°870628, SERISS: GA N°654221, SSHOC: GA N°823782), and by DG Employment, Social Affairs & Inclusion. Additional funding from the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, the U.S. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, OGHA_04-064, HHSN271201300071C), and from various national funding sources is gratefully acknowledged (see www.share-project.org).
This project was carried out within the research group: “Feminine Aging, Masculine Aging: Aging and Gender in Israeli Society,” which worked during 2019–2021 under the supervision of Dr. Gabriela-Spector and Prof. Galit Nimrod. The group was part of a nationwide project of the Feminist Forum in Sapir College, Israel, and funded by the Israeli Council for Higher Education.
Open access
