Nordic Yearbook of Population Research, Vol 58 (2025): 59–76 Kinship and Friendship Networks among Swedish and Finnish Speaking Finns ANTTI O. TANSKANEN Population Research Institute, Väestöliitto, Finland Department of Social Research, University of Turku, Finland MIRKKA DANIELSBACKA INVEST Research Flagship Centre, University of Turku, Finland Population Research Institute, Väestöliitto, Finland ANNA ROTKIRCH Population Research Institute, Väestöliitto, Finland Abstract Prior research has indicated that in Finland Swedish speakers have more social capital than Finnish speakers in terms of social participation, social support, friendship ties and trust. However, no com- prehensive study has assessed the size of close kinship and friendship networks among both linguistic groups. We compare the size of kinship and friendship networks in these language groups using popu- lation-based surveys of baby boomers, aged 68–73 years, and their adult children, aged 19–56 years at the time of data collection. Overall, baby boomers reported larger close networks than did their adult children. Swedish speakers reported slightly more close relatives than Finnish speakers did in the older (6.9 vs. 6.1) and the younger generations (4.7 vs. 4.6) but the differences were not statistically significant. However, Swedish speakers reported on average significantly larger close friendship net- works than Finnish speakers in both the older (7.4 vs. 5.3) and younger (5.6 vs. 4.2) generations. The differences in close friendship networks remained significant after controlling for sex, age, and other socio-demographic factors. Our results suggest that having more close friends is among the social cap- ital factors that characterize Swedish speaking Finns. Keywords: Finland, friendship, kinship, linguistic groups, social networks 60 Introduction Social networks are a fundamental aspect of everyday life. They can be defined to include all informal so- cial ties (i.e., excluding formal ties providing services, support or cash transfers) between actors and com- prise both familial and non-familial ties. Prior research has shown that the scope of an individual’s close social networks varies with age, sex, socioeconomic status, and income (Light & Moody, 2020). Stronger social networks can, in turn, promote both higher wellbeing and beneficial demographic behaviours relat- ed to health, partnership formation, childbearing, and active aging (e.g., Artamonova et al., 2024; Balbo & Barban, 2014; Deindl et al., 2016; Mathews & Sear, 2013). Social networks consist of varying levels of relationships, ranging from very close ties to more distant ones (Antonucci, 1986). In this study, we concentrate specifically on close social networks with regards to kinship and friendship. There are two key reasons why it is important to focus on close social ties with relatives and friends rather than more distant relationships. First, greater closeness is associated with increased investment given and received as well as more frequent interaction (e.g., Danielsbacka et al., 2015; Pollet et al., 2013). Second, close social ties have a stronger influence on health and well-being than more distant ties (e.g., Dunbar, 2018; Patterson & Margolis, 2023). Here, we are interested in possible differences in close networks between Swedish- and Finnish-speaking Finns. Finland is a bilingual country with two official languages: Finnish and Swedish. Most of the pop- ulation speaks Finnish, while about 5% belong to a historically significant Swedish-speaking minority (Saarela, 2021). The country’s language policy promotes linguistic rights and cultural preservation, ensuring that Swedish speakers also have access to education and public services in their own lan- guage. Swedish-speaking Finns have better well-being than the Finnish-speaking majority population in several respects. Swedish speakers have higher levels of education and income compared to Finnish speakers (Härtull & Saarela, 2021, 2024; Saarela & Finnäs, 2003; Saarela & Finnäs, 2004), better self-perceived health (Nyqvist & Martelin, 2007; Suominen et al., 2024), a higher life expectancy at birth (Reini & Saarela, 2021), lower divorce rates (Andersson et al., 2024; Finnäs, 1997), somewhat higher fertility (Rotkirch & Berg, 2018), a lower rate of premature mortality (Saarela & Finnäs, 2005; Saarela & Finnäs, 2016; Sipilä & Martikainen, 2009), and a longer active life (e.g., Hyyppä & Mäki, 2001). Differences in socioeconomic status partially explain the linguistic group differences in well-be- ing and demographics (e.g., Hyyppä & Mäki, 2001; Saarela & Finnäs, 2016), but cultural characteristics and social relationships also appear to play a role (e.g., Hyyppä & Mäki, 2003; Nygvist et al., 2008). It is often argued that the Swedish speakers benefit from greater social capital compared to Finnish speakers in Finland (e.g., Hyyppä & Mäki, 2001a). The classic sociological definition of social capital views it as an individual resource that encompasses social networks, social support, and trust in other individuals (Bourdieu, 1986; Coleman, 1994). The few empirical studies on the topic have provided support for the prediction that Swedish speakers tend to have more social capital than Finnish speakers. First, Hyyppä and Mäki (2001b) used survey data gathered from Ostrobothnia, which is a bilingual area in Finland, and investigated six measures of social capital, namely auxiliary friends (i.e., friends who are there to help if needed), trust in other people, singing in a choir, participation in community ac- tivities, participation in religious activities, and participation in associations. They found that Swedish speakers showed higher levels of trust, participated more often in community events, and sang more fre- quently in a choir than Finnish speakers. No significant difference was detected in auxiliary friends, par- ticipation in associations or participation in religious activities between Swedish and Finnish speakers. Nordic Yearbook of Population Research, Vol 58 (2025): 59–76 61 Next, using the same Ostrobothnia data, Hyyppä and Mäki (2003) compared social capital between Swedish and Finnish speakers by analyzing differences in four factor variables, which indicated associ- ational activity (measured by cooperation with neighbours and participation in a volunteer organization, a cultural hobby club, a school related group, a sports organization, a community meeting, or a hunting association), “friendship network” (reciprocal trust, friendship connections, and number of friends), re- ligious involvement (church attendance, visiting religious meetings, and being a member of a religious association), and participation in hobby groups (being a member of a cultural association, participating in music, writing, and drama groups, or singing in a choir). Results showed Swedish speakers had a larger friendship network and they participated in hobby groups more frequently than Finnish speakers. In contrast, Finnish speakers engaged in associational activities more than Swedish speakers, while no differences in religious involvement were found between the two linguistic groups. Finally, Nyqvist and colleagues (2008) used a nationwide survey gathered from five university hospital regions in Finland (i.e., Helsinki, Turku, Tampere, Kuopio and Oulu) to examine whether differences in social capital exist between Swedish and Finnish speakers. In this study, social capital was measured by social participation (involvement in society or club activities), social contacts with family, friends, or neighbors with all three types of ties grouped together, and trust in other people. They found that Swedish speakers had more social capital than Finnish speakers across all these measures, that is, they engaged in more social participation, had more social contacts, and exhibited greater trust in others. Although the differences in social capital between the linguistic groups were partly explained by socioeconomic factors, for instance, Swedish speakers had on average higher education and income than Finnish speakers, cultural characteristics such as stronger community participation also appear to play an important role in explaining these differences. There is evidence indicating that Swedish speakers have more social capital than Finnish speakers when social capital is measured in terms of social participation, social support, friendship ties and trust. However, no comprehensive studies have been conducted on the size of both close kinship and close friendship networks among Swedish- and Finnish-speaking Finns. To fill this gap, we used popula- tion-based data of two adult generations to examine the kinship and friendship networks of Swedish and Finnish speakers. It is important to study different generations, as prior studies have shown that strong social ties may benefit individuals differently at different stages of the life course (Alwin et al., 2018). For instance, the role of relatives in an individual’s social network tends to increase with age. We also analyzed several sociodemographic factors (e.g., sex, education, parental status and health), which are likely to independently associate with social network size. Data and methods In this study we use population-based survey data from the Generational Transmissions in Finland (Gentrans) project, which gathered information on Finnish baby boomers, here also called the ‘older generation’, born between 1945 and 1950 and their adult children, or the ‘younger generation’, born between 1964 and 1999. We used data from the third survey wave, which included questions about the size of the kinship and friendship network. The surveys were collected by Statistics Finland in 2018 and 2019, when the baby boomer generation’s respondents were between 68–73 years old and the respon- dents from the generation of their adult children were between 19–56 years old. In the older generation, 62 participants were fairly evenly distributed by age: 16% were 68 years old, 17% were 69, 70, 71, or 72 years old, and 16% were 73 years old. In the younger generation, 2% of respondents were under 30, 28% were aged 30–39, 60% were 40–49, and 10% were 50 or older. The older generation’s survey reached 2,663 respondents with a response rate of 66% and the younger generation’s survey 1,945 respondents with a response rate of 56%. With the participants’ permission, variables from population registers was merged with the surveys, which provided more background information on the respon- dents (see Hämäläinen et al., 2021 for more information of the Gentrans surveys). The dependent variables in the present study measured reported social network size. The question- naires asked the respondents to report the size of their close kinship and friendship networks, respec- tively. The questions were “How many relatives do you have that you feel close to?” and “How many close friends do you have?” and asked respondents to estimate the number of close kin and friends and write it down in the questionnaire. In the present analyses the size of the close kinship and friendship networks ranged from 0 to 20 or more. The main independent variable measured the participants’ linguistic background, i.e., whether the respondents were Swedish or Finnish speakers. This information came from population registers and was based on the mother tongue registered. When parents register their child’s name, they also register the child’s mother tongue. The language can be changed through a separate application but in practice changing the mother tongue is rare. Everyday language use and linguistic identity in Finland are often more complex than the officially registered main language, but registered language nevertheless rep- resents a good and often best proxy for linguistic identity (Saarela, 2021). In the Gentrans surveys about five per cent of respondents in both family generations were Swedish speakers which corresponds to their proportion of the population (Hämäläinen et al., 2021). Other independent variables considered the respondents sociodemographic factors and included sex (1 = female, 2 = male), age in years, partnership status (1= no partner, 2= has a partner through cohabitation or marriage), parenthood status (1 = childless, 2 = parent), education (1 = primary or lower secondary level, 2 = upper secondary level, 3 = lower degree level tertiary education, 4 = higher degree level tertiary education), self-perceived financial condition (1 = low-income, 2 = middle-income, 3 = comfortable off or wealthy), self-perceived health (1 = poor or very poor, 2 = fair, 3 = good, 4 = very good) and type of residence (1 = urban, 2 = semi-urban, 3 = rural). These variables were derived from the survey questionnaires, except for the information about the type of residence which was acquired from population registers. Descriptive statistics of older and younger generation respondents’ back- grounds are available in Table 1. We will first provide descriptive results of the social networks of Swedish and Finnish speakers by calculating the mean number of individuals in the close kinship and friendship networks, respec- tively. Next, we present multivariate results by running linear regression models in which, in addition to language, other sociodemographic factors were included. We conduct the analyses using a two-step approach, where the first step includes covariates that are unlikely to be mediators (sex and age), and the second step also includes covariates that could potentially act as mediators (partnership status, parenthood status, education, financial condition, health, and type of residence). The regression models show the associations between the dependent and independent variables while controlling for all other independent variables. Since our dependent variables, the size of kinship and friendship networks, are not normally distributed, we applied square root transformations to them for the regression analyses. Nordic Yearbook of Population Research, Vol 58 (2025): 59–76 63 Table 1. Descriptive statistics (%/mean) Older generation Younger generation %/mean SD %/mean SD Language Finnish speaker 95.3 95.0 Swedish speaker 4.7 5.0 Gender Female 60.2 65.0 Male 39.8 35.0 Age 70.5 1.67 2.3 5.81 Partnership status No partner 25.3 23.6 Has a partner 74.7 76.4 Parenthood status Childless 17.6 33.6 Parent 82.4 66.4 Education Primary or lower secondary level 27.2 7.8 Upper secondary level 49.1 59.2 Lower degree level tertiary education 10.1 29.1 Higher degree level tertiary education or doctorate education 13.7 3.9 Perceived financial condition Low income 41.2 27.5 Middle income 38.5 47.8 Comfortable off or wealthy 0.3 24.7 Self-perceived health Poor or very poor 7.4 23.4 Fair 42.8 58.0 Good 44.3 15.5 Very good 5.5 3.2 Type of residence Urban 69.1 76.9 Semi-urban 16.1 3.5 Rural 14.8 9.7 Size of close kinship network 6.2 5.01 4.6 3.93 Size of close friendship network 5.4 4.01 4.3 3.31 n 1,937 1,854 64 We also conduct sensitivity analyses to achieve more robust results. First, we calculated the size of close kinship and friendship networks for Swedish and Finnish speakers using variables where net- work size ranges from 0 to 10 or more, ensuring that the results are not driven by the specific choice of network size cutoff. In addition, we provide the distribution of the number of close individuals in kinship and friendship networks by linguistic group to capture potential differences between Swedish and Finnish speakers. Results Older generation First, we provide information about the size of the older generation’s kinship and friendship networks among Swedish and Finnish speakers (Figure 1). There was no statistically significant difference in the number of close relatives between Swedish speakers (mean (M) = 6.9; 95% confidence intervals (CIs) = 5.7–8.1) and Finnish speakers (M = 6.1; 95% CIs = 5.9–6.4). Swedish speakers (M = 7.4; 95% CIs = 6.2–8.5) had more close friends than Finnish speakers (M = 5.3; 95% CIs = 5.1–5.4) in both generations. We further examined the reported network maxima and minima. Only around 1.1 per cent of Swed- ish speakers reported having no close kin and similarly 1.1 per cent reported having no close friends, while 4.3 per cent of Finnish speakers reported having no close kin, and 2.2 per cent reported having no close friends (Figures 2 and 3). Altogether only 0.6 per cent of Finnish speakers had neither close kin nor friends, whereas all Swedish speakers in our survey data had at least one close relative or friend (Supplementary Table 1). In addition, 7.8 per cent of Swedish speakers reported having the maximum number of close kin (i.e., 20 or more), and 10.0 per cent reported having the maximum number of close friends (20 or more), while 5.2 per cent of Finnish speakers had the maximum number of close kin, and 3.0 per cent reported the maximum number of close friends in our questionnaire (Figures 2 and 3). Figure 1. Older generation: Mean size of close kinship and friendship networks by language with 95% confidence intervals Nordic Yearbook of Population Research, Vol 58 (2025): 59–76 65 Table 2 shows the results from the regression models considering factors correlating with the size of the baby boomer generation’s kinship networks. There were no significant differences in kinship network size according to language. This was the case in both the first step model, where only sex and age were included, and in the second step model, which also included other sociodemographic factors. Women had larger kinship networks than men, and partnered individuals had more close kin in networks than do individuals without partners. Parents had larger close kinship networks than childless individuals. Improved self-reported health was associated with larger close kinship networks. Individuals from rural areas had larger close kinship networks than individuals in urban areas. Finally, there were no signifi- cant differences in kinship network size according to age, education, or financial condition. Figure 2. Older generation: Distribution of the number of close kin by language, shown as percentages Figure 3. Older generation: Distribution of the number of close friends by language, shown as percentages 66 Table 2. Older generation: Stepwise regression of kinship network size on language group, controlling for sociodemographic covariates (n = 1,937) Coeff. SE 95% CI Coeff. SE 95% CI Lower Upper Lower Upper Language Finnish speaker Ref. Ref. Swedish speaker 0.17 0.11 -0.05 0.38 0.13 0.11 -0.08 0.35 Sex Female Ref. Ref. Male -0.11 0.05 -0.20 -0.01 -0.13 0.05 -0.22 -0.03 Age -0.01 0.01 -0.04 0.02 -0.004 0.01 -0.03 0.02 Partnership status No partner Ref. Has a partner 0.13 0.06 0.02 0.24 Parenthood status Childless Ref. Parent 0.12 0.06 0.002 0.25 Education Primary or lower secondary level Ref. Upper secondary level 0.02 0.06 -0.09 0.13 Lower degree level tertiary education 0.07 0.09 -0.10 0.24 Higher degree level tertiary education or doctorate education 0.04 0.08 -0.13 0.20 Perceived financial condition Low income Ref. Middle income -0.005 0.05 -0.11 0.10 Comfortable off or wealthy -0.02 0.07 -0.16 0.13 Self-perceived health Poor or very poor Ref. Fair 0.28 0.10 0.08 0.49 Good 0.35 0.11 0.14 0.56 Very good 0.41 0.13 0.15 0.67 Type of residence Urban Ref. Semi-urban 0.10 0.06 -0.03 0.22 Rural 0.17 0.07 0.03 0.30 Nordic Yearbook of Population Research, Vol 58 (2025): 59–76 67 Table 3. Older generation: Stepwise regression of friendship network size on language group, controlling for sociodemographic covariates (n = 1,937) Coeff. SE 95% CI Coeff. SE 95% CI Lower Upper Lower Upper Language Finnish speaker Ref. Ref. Swedish speaker 0.36 0.08 0.20 0.52 0.35 0.09 0.18 0.52 Sex Female Ref. Ref. Male 0.15 0.04 0.08 0.22 0.15 0.04 0.08 0.23 Age 0.02 0.01 0.004 0.04 0.03 0.01 0.01 0.05 Partnership status No partner Ref. Has a partner 0.03 0.04 -0.06 0.11 Parenthood status Childless Ref. Parent -0.02 0.05 -0.12 0.08 Education Primary or lower secondary level Ref. Upper secondary level -0.06 0.04 -0.15 0.02 Lower degree level tertiary education -0.02 0.07 -0.16 0.11 Higher degree level tertiary education or doctorate education -0.05 0.07 -0.18 0.08 Perceived financial condition Low income Ref. Middle income 0.05 0.04 -0.03 0.14 Comfortable off or wealthy 0.04 0.06 -0.07 0.15 Self-perceived health Poor or very poor Ref. Fair 0.11 0.08 -0.05 0.27 Good 0.25 0.08 0.09 0.42 Very good 0.27 0.11 0.06 0.47 Type of residence Urban Ref. Semi-urban 0.06 0.05 -0.04 0.16 Rural 0.12 0.05 0.02 0.22 68 Based on Table 3 Swedish speakers had larger friendship networks than Finnish speakers even after var- ious sociodemographic factors are controlled for. The findings were similar in both first and second step models. Regarding close friendship networks, men reported larger friendship networks than women. Respondents with good or very good health had larger friendship networks than people who reported having poor or very poor health. Those from rural areas had larger friendship networks than those in urban areas. There were no significant differences in friendship network size according to partnership status, parenthood status, education, or financial condition. In the sensitivity analyses we investigated the size of the kinship and friendship networks using variables where the network size varied from 0 to 10 or more (results not shown in tables or figures). These results exhibit a similar trend to our main analyses, which employed network size ranging from 0 to 20+, although the mean network size was obviously smaller. In the older generation, there was no significant difference in the number of close relatives between Swedish speakers (M = 5.6, 95%, CIs = 4.9–6.2) and Finnish speakers (M = 5.3, 95% CIs = 5.1–5.4) but Swedish speakers (M = 6.0, 95% CIs = 5.4–6.6) had more close friends than Finnish speakers (M = 4.9, 95% CIs = 4.7–5.0), again in line with results from the main analysis. Younger generation Figure 4 presents the numbers of the size of the kinship and friendship networks of the younger gener- ation, the adult children of Finnish baby boomers. No statistically significant difference was detected in the number of close relatives between Swedish speakers (M = 4.7; 95% CIs = 3.8–5.6) and Finnish speakers (M = 4.6; 95% CIs = 4.4–4.7), but Swedish speakers (M = 5.6; 95% CIs = 4.7–6.5) had a statis- tically significantly higher number of close friends than Finnish speakers (M = 4.2; 95% CIs = 4.0–4.4). Among Swedish speakers, 7.6 per cent had no close kin, but no one reported having no close friends. In comparison, 8.5 per cent of Finnish speakers had no close kin, and 3.8 per cent had no close friends (Figures 5 and 6). When combining friends and kin, 1.1 per cent of Finnish speakers had neither close kin nor friends, whereas all Swedish speaking respondents had at least one close relative or friend Figure 4. Younger generation: Mean size of close kinship and friendship networks by language with 95% confidence intervals 69 (Supplementary Table 1). In addition, 2.2 per cent of Swedish speakers reported the maximum number of close kin (i.e., 20 or more), and 5.4 per cent reported the maximum number of close friends, com- pared to 2.0 per cent of Finnish speakers with the maximum number of close kin, and 1.4 per cent with the maximum number of close friends (Figures 5 and 6). Overall, results from the younger generation are in line with those of the older generation. Table 4 shows no significant differences in kinship network size in the younger generation accord- ing to language. In the first-step model, the coefficient was positive (indicating that Swedish speakers had more close relatives than Finnish speakers), while in the second-step model, the coefficient was negative (indicating the opposite). In both steps, however, the associations were non-significant, mean- ing that no differences in kin network size were detected. It was found that women had larger close kin networks than men, and respondents with middle income had larger networks than people with low income. Finally, there were no significant differences in kinship network size according to age, partner- ship status, parenthood status, education, health, or type of residence. Figure 5. Younger generation: Distribution of the number of close kin by language, shown as percentages Figure 6. Younger generation: Distribution of the number of close friends by language, shown as percentages 70 Table 4. Younger generation: Stepwise regression of kinship network size on language group, controlling for sociodemographic covariates (1,854) Coeff. SE 95% CI Coeff. SE 95% CI Lower Upper Lower Upper Language Finnish speaker Ref. Ref. Swedish speaker 0.04 0.10 -0.16 0.23 -0.03 0.10 -0.22 0.17 Sex Female Ref. Ref. Male -0.09 0.05 -0.18 -0.004 -0.10 0.05 -0.19 -0.01 Age 0.001 0.004 -0.01 0.01 -0.001 0.004 -0.01 0.01 Partnership status No partner Ref. Has a partner 0.08 0.06 -0.04 0.19 Parenthood status Childless Ref. Parent 0.09 0.05 -0.01 0.20 Education Primary or lower secondary level Ref. Upper secondary level -0.11 0.08 -0.27 0.05 Lower degree level tertiary education -0.06 0.09 -0.24 0.12 Higher degree level tertiary education or doctorate education -0.08 0.14 -0.35 0.19 Perceived financial condition Low income Ref. Middle income 0.19 0.05 0.08 0.30 Comfortable off or wealthy 0.11 0.07 -0.03 0.24 Self-perceived health Poor or very poor Ref. Fair 0.06 0.13 -0.20 0.33 Good 0.04 0.13 -0.21 0.29 Very good 0.14 0.13 -0.12 0.40 Type of residence Urban Ref. Semi-urban -0.13 0.09 -0.31 0.05 Rural -0.14 0.08 -0.28 0.01 Nordic Yearbook of Population Research, Vol 58 (2025): 59–76 71 Table 5. Younger generation: Stepwise regression of friendship network size on language group, controlling for sociodemographic covariates (1,854) Coeff. SE 95% CI Coeff. SE 95% CI Lower Upper Lower Upper Language Finnish speaker Ref. Ref. Swedish speaker 0.32 0.08 0.16 0.47 0.34 0.08 0.18 0.50 Sex Female Ref. Ref. Male 0.03 0.04 -0.05 0.10 0.03 0.04 -0.04 0.10 Age -0.01 0.003 -0.01 -0.003 -0.01 0.003 -0.01 -0.001 Partnership status No partner Ref. Has a partner -0.11 0.05 -0.20 -0.02 Parenthood status Childless Ref. Parent -0.03 0.04 -0.11 0.05 Education Primary or lower secondary level Ref. Upper secondary level 0.09 0.07 -0.04 0.22 Lower degree level tertiary education 0.12 0.07 -0.02 0.26 Higher degree level tertiary education or doctorate education 0.12 0.11 -0.09 0.34 Perceived financial condition Low income Ref. Middle income 0.09 0.04 0.01 0.18 Comfortable off or wealthy 0.04 0.05 -0.07 0.14 Self-perceived health Poor or very poor Ref. Fair 0.24 0.11 0.03 0.45 Good 0.33 0.10 0.13 0.53 Very good 0.49 0.11 0.28 0.69 Type of residence Urban Ref. Semi-urban -0.03 0.07 -0.18 0.11 Rural 0.06 0.06 -0.06 0.18 72 Table 5 illustrates that Swedish speakers had larger friendship networks than Finnish speakers even af- ter several potentially confounding factors were controlled for. The results were similar in both first and second step models. The size of the close friendship network was larger among the younger respondents in this generation. Partnered respondents had smaller networks compared to those without a partner and middle-income individuals have larger networks than low income individuals. Worse self-rated health was associated with fewer close friends. There were no significant differences in friendship network size according to sex, parenthood status, education, or type of residence. Finally, for sensitivity analysis purposes, we examined the size of kinship and friendship networks using variables where network size varied from 0 to 10 or more (results not shown in tables or figures) also for the younger family generation studied here. No significant difference in kinship network size was found between Swedish speakers (M = 4.3, 95% CI = 3.6–4.9) and Finnish speakers (M = 4.2, 95% CI = 4.1–4.3) in the younger generation, but Swedish speakers (M = 5.0, 95% CI = 4.4–5.6) reported having larger close friendship networks than Finnish speakers (M = 4.0, 95% CI = 3.9–4.1). Discussion The present study investigated for the first time whether Swedish-speaking Finns have larger social net- works than Finnish-speaking Finns, taking into account both close friends and close relatives in a sur- vey of two family generations from a nationally representative sample. Results indicated no statistically significant difference in the size of close kinship networks of these linguistic groups. However, Swedish speaking respondents in both generations reported significantly larger friendship networks than Finnish speaking respondents. In the older generation the Swedish speakers had, on average, 2.1 more friends and in the younger generation 1.4 more friends than Finnish speakers did. Prior studies have shown that Swedish speakers have more social capital than Finnish speakers when social capital is defined as social participation, social support, friendship ties, and trust (Hyyppä & Mäki, 2001b, 2003, Nygvist et al., 2008). Our study confirms and expands on these results by showing that Swedish speakers also have larger friendship networks than Finnish speakers, which is an important aspect of social capital (Bourdieu, 1986; Coleman, 1994). Perhaps the study closest to ours is the one by Hyyppä and Mäki (2003), who compared Swedish and Finnish speakers using a factor variable that included a measure of the number of friends as well as two other variables: reciprocal trust and friend- ship connections. Although they found stronger friendship ties among Swedish speakers than Finnish speakers using this three-item factor variable, they were not able to detect the actual size of participants’ friendship networks or whether this varied between linguistic groups. We did not find statistically significantly larger close kinship networks among the Swedish speak- ers. This may be somewhat surprising, given that Swedish speakers have higher fertility rates and lower premature mortality rates than Finnish speakers, which suggests they should, in general, have more liv- ing relatives, and could hence be expected to also have more close relatives. One possible explanation for our findings is that cultural background may have a greater influence on the structure of non-kin relationships, while the closeness of family ties tends to remain more stable across different groups, and be concentrated to one’s closest relatives. Nordic Yearbook of Population Research, Vol 58 (2025): 59–76 73 We also detected some sex differences in the size of social networks. In both generations women had larger close kin networks than men. This may be related to the fact that women are, on average, more family oriented than men, which may lead them to form and maintain close kin relationships (Hrdy, 2024). However, we also found that in the older generation, men have larger friendship networks than women. This finding may be related to the differences in the friendship ties between women and men. Men typically have larger but less intimate groups of friends, while women have smaller but more intimate groups of friends (Dunbar et al., 2024). The present study has several strengths. First, we were able to use population-based surveys gath- ered from two adult generations, which increases the generalizability of our findings. Second, we were able to investigate kinship and friendship networks separately, providing a more nuanced understanding of linguistic group differences in the size of social networks. Third, in our analyses, we controlled for several potentially confounding variables, making the results more robust. Our study also has its limitations. Although we had accurate information on participants’ linguistic background from registers, we lacked precise data on their self-perceived linguistic identity. Neither did we have information on the mother tongue of the spouses of the respondents. Over 30 per cent of Swed- ish-speaking women and about 40 per cent of Swedish-speaking men are married to a Finnish speaker, and when these couples have children, almost 70 per cent of them are registered as Swedish speakers (Saarela, 2021). These issues highlight the complexity of linguistic identity, which is likely to shape the everyday social networks of families. Our findings showing that the Swedish speakers have larger friendship networks than Finnish speakers highlight the importance of future studies. Do the larger networks in Swedish speakers lead to better health and wellbeing or improved socioeconomic success? Are larger networks also more supportive, i.e., do Swedish speakers receive more financial and practical support from their close ones than Finnish speakers? How do bilingual individuals’ social networks compare to those of Swedish or Finnish speakers? Finally, it would be interesting to study the potential intergenerational transmission of social network size and whether this differs between the language groups. Funding statement This study was funded by the Society of Swedish Literature in Finland for the FinNät project, the Stra- tegic Research Council at the Academy of Finland for the NetResilience consortium (grant numbers 364371 and 364382), and the Academy of Finland for the KinExposure project (grant number 338869). 74 References Alwin, D. F., Felmlee, D. H., & Kreager, D. A. (Eds.). (2018). Social networks and the life course: Integrating the development of human lives and social relational networks. Springer. Artamonova, A., Sorsa, T., Berg, V., Hägglund, A. E., & Rotkirch, A. (2024). Social resources are associated with higher fertility intentions in contemporary Finland. Comparative Population Stud- ies, 49, 81–116. Andersson, L., Saarela, J., & Uggla, C. (2024). Divorce among more and less divorce-prone popula- tions following unilateral divorce laws. Journal of Marriage and Family. Antonucci, T. C. (1986). Measuring social support networks: Hierarchical mapping technique. Gener- ations: Journal of the American Society on Aging, 10, 10–12. Balbo, N., & Barban, N. (2014). Does fertility behavior spread among friends? American Sociological Review, 79, 412–431. Bourdieu, P. (1986). The forms of capital. In J. E. Richardson (Ed.), Education, Handbook of Theory for Sociology of Education. Greenwood Press. Coleman, J. S. (1994). Foundations of social theory. Harvard University Press. Danielsbacka, M., Tanskanen, A. O., & Rotkirch, A. (2015). Impact of genetic relatedness and emo- tional closeness on intergenerational relations. Journal of Marriage and Family, 77, 889–907. Deindl, C., Brandt, M., & Hank, K. (2016). Social networks, social cohesion, and later-life health. Social Indicators Research, 126, 1175–1187. Dunbar, R. I.M. (2018). The anatomy of friendship. Trends in cognitive sciences, 22, 32–51. Dunbar, R. I. M., Pearce, E., Wlodarski, R., & Machin, A. (2024). Sex differences in close friendships and social style. Evolution and Human Behavior, 45, 106631. Finnäs, F. (1997). Social integration, heterogeneity, and divorce: The case of the Swedish-speaking population in Finland. Acta Sociologica, 40, 263–277. Hyyppä, M. T., & Mäki, J. (2001a). Why do Swedish-speaking Finns have longer active life? An area for social capital research. Health Promotion International, 16, 55–64. Hyyppä, M. T., & Mäki, J. (2001b). Individual-level relationships between social capital and self-rat- ed health in a bilingual community. Preventive medicine, 32, 148-155. Hyyppä, M. T., & Mäki, J. (2003). Social participation and health in a community rich in stock of social capital. Health Education Research, 18, 770–779. Hrdy, S. B. (2024). Father time: A natural history of men and babies. Princeton University Press. Hämäläinen, H., Danielsbacka, M., Hägglund, A. E., Rotkirch, A., & Tanskanen, A. O. (2021). Suku- polvien suhteet. Väestöliitto. Härtull, C., & Saarela, J. (2021). Educational mobility across three generations of Finnish-speaking and Swedish-speaking Finns. Research in Social Stratification and Mobility, 75, 100649. Härtull, C., & Saarela, J. (2024). Earnings mobility across three generations of natives in Finland: A comparison of Finnish and Swedish speakers. Research in Social Stratification and Mobility, 94, 100990. Light, R., & Moody, J. (Eds.). (2020). The Oxford handbook of social networks. Oxford University Press. Mathews, P., & Sear, R. (2013). Does the kin orientation of a British woman’s social network influ- ence her entry into motherhood? Demographic Research, 28, 313–340. Nordic Yearbook of Population Research, Vol 58 (2025): 59–76 75 Nyqvist, F., Finnäs, F., Jakobsson, G., & Koskinen, S. (2008). The effect of social capital on health: The case of two language groups in Finland. Health & Place, 14, 347–360. Nyqvist, F., & Martelin, T. (2007). Skillnader i självskattad hälsa mellan svensk- och finskspråkiga i Finland. Sosiaalilääketieteellinen aikakauslehti, 44, 78–86. Patterson, S. E., & Margolis, R. (2023). Family ties and older adult well-being: Incorporating social networks and proximity. The Journals of Gerontology: Series B, 78, 2080–2089. Pollet, T. V., Roberts, S. G. B., & Dunbar, R. I. M. (2013). Going that extra mile: Individualstravel further to maintain face-to-face contact with highly related kin than with less related kin. PLoS ONE, 8, e53929. Reini, K., & Saarela, J. (2021). Life expectancy of the ethnically mixed: Register-based evidence from native Finns. International Journal of Environmental Research and Public Health, 18, 3415. Rotkirch, A., & Berg, V. (2018). Svenskspråkigas fruktsamhet i Finland. Vilka är skillnaderna mellan språkgrupperna? Tankesmedjan Agenda. Saarela, J. (2021). Finlandssvenskarna 2021. En statistisk rapport. Svenska Finlands Folkting. Saarela, J., & Finnäs, F. (2003). Social background and education of Swedish- and Finnish-speakers in Finland. European Journal of Education, 38, 445–456. Saarela, J., & Finnäs, F. (2004). Interethnic wage variation in the Helsinki area. Finnish Economic Papers, 17, 35–48. Saarela, J., & Finnäs, F. (2005). Mortality inequality in two native population groups. Population Studies, 59, 313–320. Saarela, J., & Finnäs, F. (2016). The ethnolinguistic community and premature death: A register-based study of working-aged men in Finland. Journal of Racial and Ethnic Health Disparities, 3, 373–380. Sipilä, P., & Martikainen, P. (2019). Language-group mortality differentials in Finland in 1988–2004: Assessment of the contribution of cause of death, sex, and age. European Journal of Public Health, 19, 492–498. Suominen, S., Stark Ekman, D., Saarela, J., Volanen, S.-M., Stenlund, S., Sillanmäki, L., & Sumanen, M. (2024). Better perceived health among the Swedish-speaking minority as compared with the Finnish-speaking majority in Finland: A cross-sectional study with an intergenerational perspec- tive. Scandinavian Journal of Public Health, 52, 1–7. 76 Supplementary Table 1. Percentage of respondents with no close kin or friends Older generation Younger generation No close kin Finnish speaker 4.3 8.5 Swedish speaker 1.1 7.6 No close friends Finnish speaker 2.2 3.8 Swedish speaker 1.1 0.0 Neither close kin nor friends Finnish speaker 0.6 1.1 Swedish speaker 0.0 0.0 n 1,937 1,854