Life satisfaction and social capital in different age groups in Finland: Data from the European Values Study

The aim of the study is to explore age and language group differences in social capital and life satisfaction, and their associated factors, among Finnish and Swedish speakers in Finland. Data were derived from the European Values Study (EVS), and the sample included 2514 individuals aged 18 years and older. Various regression-based methods were used in the analyses. The results indicate that there are significant differences in social capital between Finnish and Swedish speakers but not in life satisfaction. Nevertheless, a positive association between volunteering and life satisfaction was observed for Finnish speakers. The findings corroborate that structural and cognitive social capital are associated with higher life satisfaction and highlight the potential for promoting well-being in younger and older adults through being actively engaged in and feeling connected to a community. Moreover, special attention should be given to young adults who display the lowest levels of life satisfaction and social capital.


Introduction
Life satisfaction is usually described as a cognitive appraisal of one's life, and together with positive and negative affect, it constitutes the concept of subjective well-being (SWB; Diener, 1984). During the last few decades, SWB, including life satisfaction, has received considerable interest from researchers and policymakers globally because it is an important indicator of societal progress and provides tools for promoting welfare at the individual and societal levels (Helliwell et al., 2020;Oishi & Diener, 2014;Stiglitz et al., 2018). This endeavor has, for example, been reflected in the fact that life satisfaction is included in international inquiries, such as the Better Life Index (Organization for Economic Cooperation and Development [OECD], 2020) and the World Happiness Report (Helliwell et al., 2022). Furthermore, evidence suggests that high SWB can lead to increased health and longevity (Diener et al., 2018), underpinning the importance of investing in efforts to promote perceived well-being. Diener and Fujita (1995) posited that life satisfaction is promoted by having enough resources to fulfill one's needs and desires. Thus, social, sociodemographic, and health-related resources could help explain possible differences in life satisfaction. In this study, we focus on social capital as a resource for life satisfaction across age groups. Social capital and pro-social behavior (such as volunteering) have been positively associated with SWB, as measured by life satisfaction (Calvo et al., 2012;Elgar et al., 2011;Helliwell et al., 2018;Kushlev et al., 2021;Puntscher et al., 2015), suggesting that social capital and life satisfaction are closely interlinked. Although the association between life satisfaction and social capital has been extensively investigated, less emphasis has been placed on age group differences in these phenomena.
To promote life satisfaction among the adult population, it is crucial to gain more insight into whether the same factors affect life satisfaction in different age groups. Moreover, there is a lack of research regarding the association between social capital and life satisfaction in a Finnish setting. Current literature provides some evidence of well-being differences between the Swedish-and Finnish-speaking language groups in Finland (Reini & Nyqvist, 2017;Volanen et al., 2006) as well as social capital differences (Hyyppä & Mäki, 2001;, making Finland an interesting case to examine from a language group perspective. The present study aims to investigate social capital and life satisfaction in different age groups using data from a sample that includes Finnish and Swedish speakers in Finland.
In the following section, the first subsection lays out the theoretical framework, drawing on theories on social capital and life satisfaction. The second subsection is a literature review of studies on age differences in life satisfaction and social capital, whereas the final subsection sheds light on the Finnish context.

Background Social capital as a resource for life satisfaction
Social capital has been on the research agenda since the mid-1980s and early 1990s, when social capital was first conceptualized in terms of social resources within a social network (Bourdieu, 1986;Coleman, 1988) and somewhat later as a public good based on community activities (Putnam, 1993). According to Putnam (1993), social capital can be defined as "features of social organization, such as trust, norms, and networks, that can improve the efficiency of society by facilitating coordinated actions" (p. 167). A distinction has often been made in the literature between structural components such as networks, relationships, and institutions that link people and groups together, and cognitive components that reflect the values, trust, and confidence that characterize these relationships (Putnam, 2000).
Putnam's conceptualization has been referred to as the social cohesion definition of social capital and has been commonly adopted within health and well-being research, including life satisfaction research (see Almakaeva & Wilkes, 2021). Positive associations have been found between life satisfaction and social capital as measured by social participation, social networks, and social and institutional trust (Calvo et al., 2012;Elgar et al., 2011;Puntscher et al., 2015). Several theoretical assumptions have been suggested to explain why social capital is important for health and well-being (Berkman & Kawachi, 2000). For example, trust is regarded as an important characteristic in interacting with other people and developing supportive relations, whereas social participation may strengthen an individual's self-esteem and coping strategies that can be utilized in various life situations (Berkman & Kawachi, 2000). These are seen as individual effects of social capital on well-being (Rostila, 2013b). It has also been argued that societies, regions or even welfare regimes rich in social capital reinforce cooperative attitudes and cooperative practices to promote the common good, which can indirectly influence health and well-being (Rostila, 2013b). Thus, there is a contextual explanation of the association between social capital well-being in addition to the individual well-being effect (Rostila, 2013b).

Previous research on age differences in life satisfaction and social capital
There is empirical evidence that life satisfaction is U-shaped over the life span, that is, higher in younger and older age groups compared to middle age; however, this topic is still debated (e.g., Blanchflower & Graham, 2021). The dip in life satisfaction in middle age has been attributed to factors such as high demands in relation to family, work, and community (Lansford, 2018), while the higher level of life satisfaction in older age groups has been theorized to be linked to improved emotional regulation (Charles & Carstensen, 2010). However, some previous studies have also shown that life satisfaction is relatively stable, with only minor fluctuations during the adult life span, but notable decreases in very old age (Baird et al., 2010;Fritjers & Beatton, 2012). Furthermore, the timing of the curve tends to vary across countries and individuals (Graham & Pozuelo, 2017).
In addition to the level of life satisfaction, some studies have focused on the factors associated with life satisfaction in different age groups. Many predictors of life satisfaction seem to be mainly consistent over the life span (Johansloo & Jovanović, 2021;Lansford, 2018). For example, several studies have found that the association between social support and life satisfaction is robust across age groups (Capone et al., 2021;Johansloo & Jovanović, 2021;Siedlecki et al., 2014). Jebb et al. (2020) also found a consistent association between being married and having higher life satisfaction, with only small differences in the effects between age groups. However, their results showed that although employment is significantly associated with life satisfaction across the life span, the effect is the highest at around age 50 and considerably lower at younger and older ages, indicating that there are also variations in the way different factors affect life satisfaction at different ages.
Research on social capital as a resource for life satisfaction across the adult life span is scarce (e.g., Hoogerbrugge & Burger, 2018;Yuan, 2016). Hoogerbrugge and Burger (2018) suggested that neighborhood social capital, in particular, might have a larger impact on older individuals because they spend more time in the neighborhood, while for other age groups, other types of contacts, such as those linked to employment, might be more important for life satisfaction. Similarly, for families and, in particular, for women with young children, the neighborhood may also become a more important social context for life satisfaction (Hoogerbrugge & Burger, 2018). Volunteering, seen as one aspect of structural social capital, tend to have greater benefits for older adults' life satisfaction, as they are less likely to face stress from multiple role obligations and have more time to spend on various leisure and social activities (Hansen et al., 2018). Previous research has also showed that older people tend to trust people in general more than younger people do (Bäck, 2019;Holmberg & Rothstein, 2017), and this could be related to birth cohort as well as aging.

The Finnish setting
In international comparisons, Finland, along with the other Nordic countries, has had high levels of life satisfaction (Helliwell et al., 2022;OECD, 2020) and social capital (Rostila, 2013a) among its population. Finland has even been ranked number one regarding life satisfaction for the last five years in the World Happiness Report (Helliwell et al., 2022). The high levels of life satisfaction in the Nordic countries have been attributed to factors such as welfare state benefits, low corruption, a well-functioning democracy, and state institutions, as well as high levels of social trust and social cohesion (Martela et al., 2020).
However, to our knowledge, no study has examined age group differences in the association between social capital and life satisfaction in Finland. Previous research on life satisfaction in different age groups in Finland is also scarce. In a study that included Finland, Sweden, Latvia, and Estonia (Realo & Dobewall, 2011), only small differences between age groups were noted regarding the mean level of life satisfaction in the Finnish sample. However, that study did not examine any explanatory variables for life satisfaction. Regarding social capital differences in various age groups, a study based on the Finnish Health Survey conducted among people aged 30 and older showed that social capital decreases with age (Nieminen et al., 48 2008). Younger people tend, in general, to have higher levels of social capital, an exception being trust, in which older people reported higher prevalence compared to younger age groups. Contradictory results were reported in a study conducted among Finnish individuals aged 18-79 years showing that social capital, in terms of an index encompassing, for example, associational membership, trust, and social networks, increases with age (Rinta-Kiikka et al., 2018). A similar age pattern was reported in a study conducted among people between 25 and 55 years of age, which showed that younger people are less likely to have high levels of social capital (Alanen & Niemeläinen, 2003). A study conducted among Swedish speakers in Finland revealed a relatively complex trust picture across age groups (Stolpe, 2020). For example, individuals aged 70 years and older showed the highest levels of trust in neighbors, whereas younger people aged 19-29 years reported the highest levels of trust in unknown Finnish individuals.
The present study includes two samples, one of Finnish speakers and one of Swedish speakers, who belong to the Swedish-speaking minority in Finland, which thus allows for comparisons between the two groups. There has previously been some interest in examining social capital in the two language groups. In the early 2000s, Mäki (2001, 2003) conducted studies on social capital in relation to health in the Swedish-and Finnish-speaking adult population in Ostrobothnia. The authors assessed social capital with measures of trust, social networks, and associational activities. Their results showed that the ethnolinguistic minority seemed to possess higher rates of social capital compared with the Finnish-speaking population. In addition, social capital explained part of the health advantages observed among Swedish speakers. Their initial work inspired new social capital studies conducted in a Finnish context, and several have been conducted in Ostrobothnia (e.g., Nyqvist et al., 2014) as well as other bilingual regions of Finland (e.g., Stolpe, 2020).
Previous findings have revealed relatively small, but persistent, and mixed differences between the language groups. It seems that older Swedish speakers tend to be more engaged in associational activities and report more frequent social contacts, whereas no significant differences between the language groups have been observed in social trust (Nyqvist et al., 2022). A recent study assessing changes in trust among older individuals in Ostrobothnia showed that levels of political and institutional trust decreased in both language groups between 2005 and 2016, whereas no significant change was observed for social trust (Näsman, . Moreover, negative changes were more pronounced among Swedish speakers, with the exception of trust in elder care; in 2005 and 2016, Swedish speakers reported higher trust compared with Finnish speakers. However, to our knowledge, no study has examined life satisfaction with a focus on the two language groups, including the relationship between social capital and life satisfaction, although other aspects of well-being have been assessed, such as morale (Näsman, Niklasson, et al., 2020), sense of mastery (Reini & Nyqvist, 2017), and absence of loneliness (Nyqvist et al., 2016).
To sum up, although previous research has found high levels of well-being and social capital in Finland, the picture of the features of life satisfaction or social capital in different age groups is unclear. Importantly, age-related differences in life satisfaction do not necessarily mean age-related similarities in the factors associated with life satisfaction, implying that the relationship between social capital and life satisfaction might differ depending on age. Moreover, empirical evidence utilizing data from both Finnish and Swedish speakers remains relatively rare, supporting the need to explore the Finnish context in further analyses.

Aim
The overall aim is to explore social capital and life satisfaction in different age groups in a Finnish setting. Specifically, we assess (O1) how age and language group are associated with social capital, (O2) the association between social capital and life satisfaction, and (O3) age and language group differences in the association between social capital and life satisfaction. 49 Finnish Journal of Social Research, Vol. 15 (2022)

Sample
The sample was derived from two data sets collected as parts of the European Values Study (EVS) conducted in Finland. A data collection in Finnish was conducted during 2017 to 2018 (European Values Study Group, 2020) and in Swedish during 2018 to 2019 (von Schoultz et al., 2019). The Finnish-speaker data collection was conducted through computer-assisted interviews and self-administered questionnaires (paper and online), while the Swedish-speaker data collection was conducted only with self-administered questionnaires (paper and online). A two-stage stratified random sampling process was applied for the computer-assisted interviews and simple random sampling was applied for the self-administered questionnaires. More information regarding the methodology can be found on the EVS website (www. europeanvaluesstudy.eu).
The Finnish-speaking sample consisted of 1199 individuals and the Swedish-speaking sample of 1315 individuals aged 18 years and older. No upper age limit was applied. The response rates were 23.4 percent for the Finnish-speaking sample and 36.5 percent for the Swedish-speaking sample. A discussion of issues regarding the response rates in the EVS can be found in Luijkx et al.'s (2021) study. Data on the age variable were missing for 87 individuals; therefore, they were excluded from the analysis. Additionally, the response options "don't know" and "prefer not to answer" were coded as missing for all variables included in the analysis, which accounted for 0.3-3.1 percent of the responses.

Life satisfaction
Life satisfaction was measured with a single item, which has been shown to have comparable validity to a multiple-item scale (Cheung & Lucas, 2014) and acceptable psychometric properties in different contexts (Diener et al., 2013). Life satisfaction was measured with the question "All things considered, how satisfied are you with your life as a whole these days?," where 1 represents dissatisfied, and 10 represents satisfied. The measure was used as a continuous variable. The variable was deemed to be normally distributed, with an absolute value of -1.484 for skewness and a value of 3.005 for kurtosis (Kim, 2013).

Social capital
Social capital was measured using four different variables. Participation in volunteer activities was considered a measure of structural social capital and was assessed with the question "Did you do voluntary work in the last 6 months?" ("yes" = 1, "no" = 0). Membership in associations was also used as a measure of structural social capital, and was based on the question: "For each of the following voluntary organizations, please indicate which, if any, do you belong to." including the alternatives "religious or church organizations", "education, arts, music or cultural activities", "trade unions", "political parties or groups", "conservation, the environment, ecology, animal rights", "professional associations", "sports or recreation", "humanitarian or charitable organization", "consumer organization", "self-help group, mutual aid group", and "other groups". The variable was dichotomized using median split, resulting in that being a member of two associations or more were coded as 1 and less than two associations as 0.
Social trust was considered a measure of cognitive social capital and was assessed with the question "Generally speaking, would you say that most people can be trusted or that you can't be too careful in dealing with people?" ("most people can be trusted"=1, "can't be too careful"=0). Sense of community was also seen as a measure of cognitive social capital and was measured with the question "Please indicate how close you feel to… your town or city" ("very close", "close", "not very close", "not close at all", "don't know"). The variable was dichotomized so that "very close" and "close" were considered as strong sense of community (=1) and "not very close" and "not close at all" as weak sense of community (=0).

Explanatory variables
The explanatory variables of main interest in the present study are age group and language group. The sample was divided into five age groups: 18-34 years, 35-49 years, 50-64 years, 65-74 years, and 75 years and older. Finnish speakers were coded as 0, and Swedish speakers as 1, based on which data collection they had participated in.

Control variables
Various sociodemographic variables, such as gender, educational level, civil status, and self-rated health, were expected to affect social capital and life satisfaction (e.g., Joshansloo & Jovanović, 2021;Lansford, 2018;Nyqvist et al., 2014). Therefore, they were included as controls. Women were coded as 0 and men as 1. Educational level was based on the question, "What is the highest educational level that you have attained?" The variable was dichotomized and divided into primary or secondary education (= 0) and tertiary education (= 1). The variable representing civil status was based on three questions: (1) "What is your current legal marital status?" (2) "Do you live with a partner?" (3) "Do you have a steady relationship?" An individual was considered to be in a relationship (coded as 1) if he/she was married, in a registered partnership, cohabiting, or in a steady relationship, and single if otherwise (coded as 0). Self-rated health was measured with the question, "All in all, how would you describe your state of health these days? Would you say it is" ("very good," "good," "fair," "poor," and "very poor"). The variable was dichotomized, where "very good" and "good" were considered good self-rated health (= 1), and "fair," "poor,' and "very poor" were considered poor self-rated health (= 0).

Analysis
Multivariate logistic regression was used to test the associations between age group, language group, and each social capital measure while controlling for sociodemographic variables and self-rated health (O1). Linear regression of the Ordinary Least Square type (OLS) was used to test the associations between social capital and life satisfaction (O2). Model 1 included the social capital variables. Model 2 additionally included age, language group, gender, educational level, civil status, and self-rated health.
Finally, to further examine the association between social capital and life satisfaction in different age groups and language groups (O3), joint effects were calculated and tested. The joint effects included the main effect and the interaction effect between each social capital measure and age group, and between each social capital measure and language group. Therefore, the variables computed to test joint effects included a combination of two features, such as age group and level of social trust. The joint effects were estimated within the same age group or language group by switching the reference categories; for example, 18-to 34-year-olds who had low social trust were the reference category for 18-to 34-year-olds who had high social trust. All joint effects were tested one by one in multiple linear regression models with life satisfaction as the outcome, where all other social capital measures as well as language group/age group, gender, educational level, civil status, and self-rated health were controlled for. Normalized weights were used in the regression analyses to account for the different sampling proportions of the two language groups; that is, Swedish speakers constitute about five percent of the national population. Statistical significance was set at the five percent level. SPSS 27 was used for all analyses.

51
Finnish Journal of Social Research, Vol. 15 (2022) Results  Descriptive information for the total sample and for the five age groups is presented in Table 1. Regarding the social capital variables, the proportion doing volunteer work was lowest in the 18-to 34-year-old age group (27.9%) and highest in the 65-to 74-year-old age group (40.3%). The percentages for being a member of at least two associations were similar in the different age groups, ranging from 57.9% in the 65-to 74-year-old age group to 62.9% in the 35-to 49-year-old age group. Overall, the sample had a strong sense of community (86% in the total sample), with the highest percentage in the 65-to 74-year-old age group (91.1%). Approximately three out of four (75.4%) respondents agreed that most people can be trusted, and the highest percentage was found in the 35-to 49-year-old age group. Finally, regarding life satisfaction, the mean score was lowest in the 18-to 34-year-old age group (M = 7.69, SD = 1.64) and highest in the two oldest age groups (M = 8.41, SD = 1.50; M = 8.34, SD = 1.70). Table 2 presents the results for the logistic regression models with each social capital measure and the explanatory variables. The associations between language group and each social capital measure were statistically significant, even when controlling for age, sociodemographic variables, and self-rated health.  Swedish speakers had a higher likelihood of volunteering, being a member of at least two associations, having high trust, and having a strong sense of community compared to Finnish speakers. The association with age group varied somewhat between the measures. The odds ratios (ORs) for doing volunteer work were significantly higher in all age groups compared to the reference category of 18-to 34-year-olds, with the highest OR found among the 65-to 74-year-old age group (2.40). Regarding associational participation, age group did not have a statistically significant effect. For social trust, all older age groups were associated with a higher likelihood of having high trust compared to being 18 to 34 years old, although the effect of being 65 to 74 years old was not statistically significant. The likelihood of having high trust was the highest among the 50-to 64-year-old age group (OR = 1.75). Finally, being in the three oldest age groups was associated with a statistically significant higher likelihood of feeling a strong sense of community. The OR was notably higher in the 65-to 74-year-old age group compared to the 50-to 65-year-old age group and the ≥75-year-old age group (4.33 vs. 1.90 and 2.48). The results for the linear regression models with life satisfaction as the outcome are shown in Table 3. Model 1 included the social capital variables, where positive responses to all measured aspects of social capital were significantly associated with higher life satisfaction. In Model 2, having done volunteer work, having high trust, and having a strong sense of community were associated with higher life satisfaction when age, language group, sociodemographic variables, and self-rated health were controlled for, while

55
Finnish Journal of Social Research, Vol. 15 (2022) the positive association between associational membership and life satisfaction was no longer statistically significant. Being in the 50-to 64-year-old, 65-to 74-year-old, and ≥75-year-old age groups was associated with higher life satisfaction, while the effect of language group on life satisfaction was not statistically significant. Table 5. Estimated joint effects of language group and social capital on life satisfaction (n=2162 a ).
The results of the joint effect analyses are presented according to age group in Table 4 and according to language group in Table 5. Regarding the age groups, volunteering had a statistically significant positive effect on life satisfaction in the 50-to 64-year-old age group. In addition, being a member of at least two associations had a statistically significant positive effect on life satisfaction in the youngest (18-34 years) and the oldest (≥75 years) age groups. Further, high social trust had a statistically significant positive effect on life satisfaction in the 18-to 34-year-old, 35-to 49-year-old, and ≥75-year-old age groups. Having a strong sense of community also had a statistically significant positive effect on life satisfaction in the 18-to 34-year-old, 50-to 64-year-old, and ≥75-year-old age groups. Regarding the language groups, volunteering had a statistically significant effect on life satisfaction in Finnish speakers, but not in Swedish speakers. In addition, having high social trust and a strong sense of community had statistically significant positive effects on life satisfaction in both language groups.

Discussion
In this study, we examined social capital and life satisfaction in different age groups of Finnish and Swedish speakers in Finland. Language group was significantly associated with all social capital measures in the logistic regression analysis, which showed that Swedish speakers have a higher likelihood of volunteering, being a member of at least two associations, feeling a strong sense of community, and having high trust. Older age generally seemed to have a positive association with social capital, except for associational membership, where the role of age group seemed to be less relevant. Older age and positive responses regarding all social capital measures except associational membership were significantly associated with higher life satisfaction in the multivariable analyses. However, in a closer examination of the association between social capital and life satisfaction in the different age groups, no clear age pattern was observed, as the associations were significant within the youngest, middle-aged, and oldest age groups. Similarly, the joint effect analyses of social capital and language group showed that social capital had a similar effect on life satisfaction in the two language groups, except for volunteering, where a positive effect was found for Finnish speakers.
We confirmed previous findings (e.g., Hyyppä & Mäki, 2001) showing that being a Swedish speaker is associated with higher structural and cognitive social capital. It has been discussed in the literature that Swedish speakers, due to less geographic mobility and strong institutional support, experience a higher degree of social capital than Finnish speakers (Hyyppä & Mäki, 2001). However, previous work has mainly focused on the relationship between social capital and various health measures (Hyyppä & Mäki, 2001;Nyqvist et al., 2014), and to a lesser extent on exploring social capital differences between the two language groups. Therefore, this study adds to the literature on social capital resources in the two language groups in Finland. Furthermore, overall, we found no statistically significant association between being a Swedish speaker and having higher life satisfaction, corresponding to the results of Näsman, Niklasson, et al. (2020), who found no association between language group and morale (another measure of SWB) in a sample of older adults. The joint effect analysis displayed similar effects for social trust, sense of community, and life satisfaction in both language groups. These results imply that social capital, in general, is important for well-being in both language groups. An exception was volunteer work, which benefits life satisfaction for Finnish speakers, whereas no such association was noticed in the Swedish speaking group. This finding could be related to factors such as the type and intensity of volunteering, which we were not able to explore with this data set and thus, should be investigated further to gain more insights into possible explanations.
In general, the present results imply that older adult age groups in Finland tend to be more resourceful when it comes to social capital, which is in line with the studies by Rinta-Kiikka et al. (2018) and Alanen and Niemeläinen (2003) but partly deviates from the results of Nieminen et al. (2008). For example, the present results indicate that the likelihood of volunteering and having a strong sense of community is higher in the older age groups than in the youngest age group. However, an exception is associational membership, for which no significant age differences were found. As for the role of social capital in life satisfaction, both structural and cognitive social capital were positively associated with life satisfaction, which mirrors previous research (e.g., Calvo et al., 2012;Helliwell et al., 2018;Kushlev et al., 2021). One aspect of structural social capital, associational membership, was not associated with life satisfaction when age group, language group, sociodemographic variables, and self-rated health were controlled for. This result might be explained by the crude measure of membership that did not discriminate between active and passive membership (Nyqvist et al., 2012) or by the fact that not all associational activities contribute to health and well-being (Nyqvist et al., 2008), such as political engagement, which might originate from political dissatisfaction (Portela et al., 2013).
No clear age pattern was observed regarding the association between social capital and life satisfaction. For example, associational membership was positively related to life satisfaction within the 18-to 34-year-old and ≥75-year-old age groups, whereas volunteer work was important to life satisfaction in the 50-to 64-year-old age group. Further, having high trust and feeling a strong sense of community had a positive joint effect on life satisfaction in several of the age groups, from the youngest to the oldest, reflecting the importance of structural and cognitive social capital in various life stages. Thus, the present results corroborate previous research highlighting that social capital is essential for life satisfaction in old age (Hansen et al., 2018;Hoogerbrugge & Burger, 2018), but also denotes an influence within younger age 57 Finnish Journal of Social Research, Vol. 15 (2022) groups.
Finally, the present results showed that mean life satisfaction was lowest in the youngest age group (18-34 years). Thus, these results differ somewhat from the results of Realo and Dobewall (2011), who noted that the level of life satisfaction was lowest among people aged 41 to 60 years in their Finnish sample. Even though their data were also partly derived from the EVS, differences in the inclusion of cohorts and in the use of control variables could possibly explain some of the differences in the results. Considering that higher life satisfaction among young adults, for example, has been associated with engaging in post-comprehensive studies, higher education, and work, which, in turn, reflect positive development (Upadyaya & Salmela-Aro, 2017), the relatively lower level of life satisfaction among younger adults is worth recognizing and should be investigated further. This concern has become even more evident in light of the COVID-19 pandemic, when young adults in Finland seemed to be particularly affected (Ranta et al., 2020). Based on the present findings regarding the positive joint effects of associational membership and social trust on life satisfaction in the 18-to 34-year-old age group, investing in efforts to strengthen social capital could be pivotal in the promotion of life satisfaction among young adults.

Limitations
This study was based on EVS data, which enabled the exploration of age group differences in social capital and life satisfaction in a Finnish setting. Thus, the present study has contributed new insights into these phenomena and their associated factors. Nevertheless, when interpreting the results, several limitations should be considered.
First, the sample consisted of one data collection conducted in Finnish, and one collection conducted among Swedish speakers in Finland. The sample for the data collection conducted in Swedish included Swedish-speaking Finns, while the sample for the data collection in Finnish did not target a specific language group. Information regarding the respondent's mother tongue was not available for the Finnish sample. Therefore, it cannot be ruled out that Swedish speakers were included in the data collection among Finnish speakers. Future research would also benefit from including other linguistic groups, such as the indigenous Sami in Finland and linguistic minority groups with a migrant background, in order to get a more comprehensive picture of life satisfaction in Finland. Furthermore, the data collection was conducted using mixed modes (Luijkx et al., 2021), meaning that there is a possibility that there were some differences in responses depending on whether the data collection was conducted face-to-face or not. The sample sizes of the age groups also varied somewhat, which could be partly explained by the overrepresentation of respondents in young-old age (Luijkx et al., 2021).
Second, the analyses were based on cross-sectional data, implying that the results reflect differences between the age groups in the sample and not changes over the life span. It is also important to consider that differences between age groups could be affected by cohort effects. On the same note, it is not possible to determine any causality between social capital and life satisfaction. However, it could be assumed that the relationship is reciprocal (Kushlev et al., 2021); that is, social capital promotes life satisfaction but also vice versa.
Third, it should be noted that the results concerning life satisfaction might not necessarily apply to the broader concept of SWB, which is more multidimensional (Diener, 1984). Similarly, the concept and measures of social capital vary across studies, which affects the possibility of comparing the results. In this study, we focused on structural and cognitive social capital using indicators similar to those in previous research (Moore & Kawachi, 2017).

Conclusions
The results of the present study support the notion that social capital can be seen as an important resource Näsman and Nyqvist 58 for life satisfaction, even when age, language group, sociodemographic characteristics, and self-rated health are considered. Although Swedish speakers reported higher levels of social capital than Finnish speakers, there was no association between language group and life satisfaction. In addition, the effect of social capital generally seemed to be similar in the two language groups, but volunteering was positively associated with life satisfaction only for Finnish speakers. The results imply that there is the potential to promote life satisfaction by strengthening social capital across various ages. Nevertheless, this study identifies 18-to 34-year-olds as having lower levels of life satisfaction and social capital, calling for the need to pay attention to young adults in policy development and future research on this matter.

Funding
The work by Marina Näsman was supported by the Society of Swedish Literature in Finland.