“Being Stuck between Two Worlds” – Identity Configurations of Occupational and Family Identities

ABSTRACT

Most research on identity has focused on either identity processes or identity content, neglecting how individuals structure their identities. We investigated how individuals negotiated their sometimes conflicting multiple identifications of work and family into different types of workable identity configurations. We also examined differences across configurations in the gender distribution, degree of conflict, and life satisfaction. Through a mixed methods approach, a community sample of 124 young Swedish adults (50% women, Mage 33.29 years) were interviewed about work and family priorities and completed a rating scale measure of life satisfaction. The qualitative analyses showed six different types of identity configurations, the most common being Family first, followed by Everything is important, Struggling to prioritize, Now family comes first, Inability to prioritize, and Work first. The quantitative analyses showed few gender differences across configurations, but there were differences in conflict and life satisfaction between configurations. The findings from this study reflect the complex and dynamic ways identity is structured within a cultural context.

KEYWORDS:

• Identity configuration
• work
• family
• conflict
• satisfaction with life

Young adults, roughly between age 30 and 40 (Arnett, 2012), often face new experiences, contexts, and changing life circumstances that may challenge their sense of identity (Kroger, 2015; McAdams & Zapata-Gietl, 2015; Syed, 2010). Identity development during this time often involves the adoption of new roles in the identity domains of occupation, romantic relationships, and parenthood (Arnett, 2012; Gyberg & Frisén, 2017). As these domains develop and become integrated within individuals’ existing sense of identity they have the potential to conflict with one another, posing threats to their sense of sameness and identity integration (Erikson, 1968; Kroger, 2015; McAdams & Zapata-Gietl, 2015; Syed & McLean, 2016). Thus, configuring these potentially conflicting multiple identities into a workable whole becomes an important aspect of adult development (Schachter, 2004, 2005a, 2013). Nevertheless, there is relatively little research on how young adults manage and structure their multiple identities into different identity configurations to achieve a sense of sameness and integration. Accordingly, the purpose of this study is to investigate how young adults configure their potentially conflicting occupational and family identities.

Identity configurations: a cultural understanding of identity structure

Identity configurations capture the structure of identity; the different ways individuals negotiate and integrate their sometimes conflicting multiple identity domains into their broader sense of self (Erikson, 1968; Schachter, 2004, 2005a, 2005b). This process of integrating potentially conflicting identities into a workable whole captures what Erikson (1968) called identity synthesis and is one of the most important developmental tasks in adolescence and young adulthood. Importantly, how identities are configured into a coherent whole varies depending on both individual factors and cultural contexts (Schachter, 2004, 2005a, 2005b).

Numerous researchers have made a strong case for the importance of the cultural context of identity developmental in general (Galliher, McLean, & Syed, 2017; Hammack, 2008, 2011; McAdams & Zapata-Gietl, 2015; McLean et al., 2017) and for identity configurations in particular (Dahl & Galliher, 2012; Hammack, 2010; Hammack, Thompson, & Pilecki, 2009; Schachter, 2004, 2005a, 2005b, 2013). The cultural context is especially important for highlighting relevant identity domains, as well as identifying the combinations of domains that may be particularly prone to conflict. Thus, any study of identity configurations must be situated within the cultural context of the respondents.

The cultural context of Sweden is important to consider when examining how individuals configure their occupational and family identities. In Sweden there is a strong cultural discourse supporting gender equality and the importance of sharing life responsibilities (Gyberg & Frisén, 2017; Towns, 2002), especially concerning work and family practices (e.g., Almqvist, Sandberg, & Dahlgren, 2011; Haas & Hwang, 2000; Johansson & Klinth, 2008; Sommestad, 1997). There are many policies and laws that favor living as gender-equally as possible (Duvander, 2014; Johansson & Klinth, 2008). For example, there are parental leave days reserved for fathers in order to encourage both parents to take care of their children. However, even though Swedish fathers’ parental leave slowly increases each year, in practice, mothers and fathers still do not share the parental leave equally, and women more often hold part-time jobs while caring for their children (Statistics Sweden [SCB], 2018). Some have argued that there are parallel norms in Sweden: dual-breadwinner norms as well as more traditional norms of women being more family-oriented and men more work oriented (Elvin-Nowak & Thomsson, 2001; Haas & Hwang, 2019; Hagqvist, Gådin, & Nordenmark, 2017).

Of course, work-family conflict is an issue relevant for many cultures worldwide, which is reflected in a large body of research that has primarily focused on conflict related to the division of time between work and family (e.g., Byron, 2005; Cinamon & Rich, 2002; Michel, Kotrba, Mitchelson, Clark, & Baltes, 2011). However, cross-cultural comparative studies of work-family conflict have shown that people in more gender-egalitarian contexts, such as Sweden, generally experience higher level of work-family conflict, especially for woman, and a stronger negative relationship between work-family conflict and well-being than countries with more traditional gender role norms (e.g., Hagqvist et al., 2017; van der Lippe, Jager, & Kops, 2006).

Despite this large body of work, to date, work-family conflict has not been examined from an identity perspective, which may be important for our understanding of identity integration in young adults’ everyday lives. Indeed, the lack of attention to work-family conflict in the identity literature is surprising given that Erikson (1968) specifically discussed the issue when outlining his theory of identity development, where he explicitly stated that a primary task of identity development for some women is to resolve conflicts between work and family to develop an integrated identity. Moreover, he specified this to be a broader ideological issue that is relevant for all young adults in today’s society, not just those who are actively working or parenting. Thus, taking an identity perspective on work-family conflict in Sweden is especially interesting as the discourse of gender equality is not only a matter of dividing time between work and family, it is also an ideological issue (Frisén, Carlsson, & Wängqvist, 2014; Gyberg & Frisén, 2017).

Previous studies within a Swedish cultural context, many of which are based on the same dataset used in the current study, have shown that work, family, and the balance between them, are important identity domains for both women and men (Bergh & Erling, 2005; Fagerberg & Kihlgren, 2001; Frisén et al., 2014; Frisén & Wängqvist, 2011; Gyberg & Frisén, 2017; Wängqvist, Carlsson, van der Lee, & Frisén, 2016). The only previous Swedish study on gender differences in identity development among young adults, based on the same participants as in the present study, found that women had engaged in more identity exploration in the occupational and parenthood domains compared with men, but there were no differences regarding romantic relationships and work/family priorities (Gyberg & Frisén, 2017). In sum, although notions of gender equality are strong features of Swedish culture, the available data suggests the integration of occupational and family identities may be challenging, and that this may be especially the case for women.

Previous research on identity configurations

The use of the term “identity configuration” in the literature has not always been consistent, with some researchers using it to describe identity structures based on levels of exploration and commitment (e.g., Crocetti, Scrignaro, Sica, & Magrin, 2012; Luyckx, Seiffge-Krenke, Schwartz, Crocetti, & Klimstra, 2014). Our conceptualization of identity configurations is consistent with Erikson’s (1968) original conceptualization and Schachter’s (2004, 2013) elaboration on the concept, which has received relatively less attention.

Much of this work has relied on qualitative methods, focusing on drawing theoretical conclusions on the concept of identity configurations rather than generalizations to a wider population. For example, Schachter (2004, 2005a) identified different types of identity configurations of sexual identity and religious identity among Jewish Modern Orthodox young men, where some men tried to suppress or reject one identification over another whereas others tried to maintain both identifications. Schachter (2004, 2005a) suggested that identity development should be viewed as configurations of identity domains that are negotiated and co-constructed within a context that is dependent on culture, content, and development. Building on this work, two studies on sexual minority youth (i.e., Dahl & Galliher, 2012; Hammack et al., 2009) found different ways of creating workable identity configurations, highlighting how societal discourse not always matches the participants’ desires and behaviors associated with their identities. Finally, two longitudinal studies (i.e., Hammack, 2010; Syed, 2010) have shown how identity configurations evolve over time as young people face new experiences that can prompt a restructuring of their identities. Taken together, these studies demonstrate how identity configurations can provide meaning and coherence to individuals’ sense of selves when facing conflicting discourses and emphasize the importance of context for identity development.

Despite the accumulating empirical work on identity configurations, there are several limitations of this past work. First, identity configurations have been studied within limited domains, mostly focusing on aspects of minority identity, and there is a need to expand this line of research to other domains (Dahl & Galliher, 2012; Schachter, 2004; Syed & McLean, 2016). As argued previously, Sweden is likely to be a context where negotiating occupational and family identity domains may be particularly important for both women and men.

Second, although theoretically important, none of the previously mentioned studies directly measured the degree of conflict across different identity configurations (i.e., individuals’ subjective experiences of internal conflicts between identifications). Some have suggested that individuals configure their identities in order to minimize conflict (Hammack, 2010), although Schachter (2004) found that some individuals were drawn to the conflict between multiple identities. In a study of identity integration between religious identity and sexual identity, Dahl and Galliher (2009) found that although the participants did not report much integration there was some conflict experienced. However, as the study did not investigate different ways of structuring multiple identifications it is difficult to draw conclusions on conflict in terms of identity configurations. Finally, none of the abovementioned studies investigated how identity configurations were related to aspects of well-being. Doing so would contribute valuable insights on the relation between identity structure and well-being (see Schachter, 2004).

The present study

The purpose of the present study was to investigate how young adults in Sweden configure their work and family identities, addressing some of the aforementioned limitations. We used an explanatory mixed methods design (see Creswell & Plano Clark, 2007) to address the following exploratory research questions: (1a) What different types of identity configurations are evident as young adults negotiate their work and family identities, and (1b) are there any gender differences in the distribution of configurations? (2) Does the degree of conflict vary across the different identity configurations? (3) Do reports of life satisfaction vary across the different identity configurations?

Method

Participants and procedure

The present study draws from the tenth wave of data collection of the Gothenburg Longitudinal Study of Development (GoLD), which started in 1982 consisting of 144 families recruited from public childcare waiting lists. Public childcare is offered to families in Sweden from the time the child is one-year old, where parents put their child on waiting lists for pre-schools. In the latest data collection, which this study is based on, 124 individuals, 62 women and 62 men (M_age 33.29 years, SD = 0.54) of the original sample chose to participate (attrition rate 13.9%). At the time of the data collection, 77% of the participants (n = 96) were working as a main source of income, 7% (n = 9) were on parental leave, 4% (n = 5) were studying, 4% (n = 5) were unemployed or on sick leave, and 10% (n = 13) were combining part-time work, parental leave, sick leave, and/or studies. Also, 83% of the participants (n = 103) were in a romantic relationship and 66% of the participants (n = 82) were expecting or had children.

Most participants were interviewed at the Department of Psychology at University of Gothenburg or at the Department of Psychology at Stockholm University. In a few cases, where meeting at the two Universities was not possible, participants were interviewed in their homes, at public institutions (e.g., libraries), via voice-over-IP services or by telephone. All participants signed informed consent before participating. The study was reviewed by the Regional Ethical Review Board in Gothenburg (Dnr: 263–15).

Measures

Background interview

Through a structured interview, the participants were asked questions about their demographic information, such as, relationships, employment, and family and housing situation. The background information was used as descriptive information and to present case examples, where all names mentioned are fictitious.

Identity status interview

The semi-structured Identity Status Interview (ISI; Marcia, 1966; Marcia, Waterman, Matteson, Archer, & Orlofsky, 1993) is a method developed originally by Marcia (1966) on the basis of Erikson’s (1968) work as a tool to investigate and assess individuals’ identities. The interview comprises questions that encourage individuals to talk about and reflect on their lives, in the past, present, and future in order to gather information on the exploration and commitment to different identity-defining areas of life. In the present study, we analyzed the interview narratives from the part of the interview concerning the identity domain of work/family priorities. In order to encourage the participants to reflect on work/family priorities, the interviewer began by reminding the participant how important they had rated the work, romantic relationships, and parenthood domains on a 1–7-point scale where 1 was “Not important at all” and 7 was “Extremely important”. Thereafter, the interviewer explained how one can have thoughts about how to divide the time between work and family and how inner conflicts may arise as a result of having to prioritize. Then followed the semi-structured interview, where the first question asked if the participant had thought about that type of conflict. Other examples of questions for this identity domain were: “How would you like to prioritize between work and family in your life?”; “Do you have experience of work/family conflicts?”. Additional follow-up questions were asked in order to encourage the participants to elaborate on their answers.

Life satisfaction

Degree of life satisfaction was assessed using the Satisfaction With Life Scale (SWLS; Diener, Emmons, Larsen, & Griffin, 1985). The SWLS is five-item measure with statements about life satisfaction (e.g., “In most ways my life is close to my ideal” and “If I could live my life over, I would change almost nothing”) ranging from 1 (Strongly disagree) to 7 (Strongly agree) on a Likert scale. A Swedish version of the scale was used, which has been validated by Hultell and Gustavsson (2008). Cronbach’s α for the five-item scale in the present study was .85.

Narrative coding

Coding procedure

In line with thematic analysis (see Braun & Clarke, 2006), the coding procedure for identity configurations and degree of conflict began with the first author reading and re-reading 30 transcripts from the interviews in order to develop initial codes and a first draft of the coding manual. The coding was then discussed and the coding manual revised in collaboration with the second author. Then, the third author was brought in to discuss the initial coding and the coding manual was once again revised. Thereafter, the first author listened to recordings of the 94 remaining interviews and wrote summaries for each of the participant’s narratives that were read and re-read. The coding manual was once again discussed amongst the three authors in order to reach consensus and revised accordingly. Last, each interview was coded for identity configuration and conflict separately. When needed, the coder re-listened to the interviews in order to ensure the coding. When insecurities arose, the coding was discussed with the two other authors until consensus was met. In order to ensure reliability in the coding, a trained research assistant re-coded 30% of the interviews for identity configuration (Cohen’s ҡ = .79, 84% agreement) and an additional 30% of the interviews for degree of conflict (ICC = .90).

Identity configuration

Type of identity configuration was assessed through investigating how the participants described their priorities between work and family, how they wanted to prioritize, and how they negotiated between work and family. All parts of the interview in which the participants reflected on work and family were given special attention and treated as a whole in coding how they structured their occupational and family identities. The concept of family was initially presented to the participants as a combined aspect of romantic relationship and parenthood. However, what the participant then chose to include in the concept of family when responding was not part of our assessment and may thus mean different things to different people. For example, participants’ references to family could primarily mean parenthood, romantic relationships, or their family of origin.

Conflict

Degree of conflict was assessed on a 1–4-point scale, with 1 representing “No conflict” (i.e., no presence of conflict in the interview), 2 “Low conflict” (i.e., indication of there being minimal or little conflict, happening once or twice with little impact), 3 “Moderate conflict” (i.e., descriptions of conflict as something happening sometimes, or periodically and not seen as very burdensome), and 4 “High conflict” (i.e., heavy conflict that is described as an everyday struggle and/or with high intensity). Both frequency and intensity of conflict were considered to determine an overall assessment of conflict throughout the interview. For example, if a participant talked about having daily conflicts that were not severe, this was still viewed as heavy conflict due to the high frequency. Also, if a participant talked about sometimes experiencing really intense and heavy conflicts, this was coded as heavy conflict due to the high intensity.

Results

Identity configurations

We identified six identity configurations: (1) Family first, (2) Work first, (3) Now family comes first, (4) Everything is important, (5) Struggling to prioritize and (6) Inability to prioritize, which are further described below. Details about the participants’ work, relationship, and family situation are presented in Table 1.

Table 1. Descriptives of background information about the participants across identity configurations.

	Working to make a living
	10–35h/w	36–44h/w	45–60h/w	Parental leave	Studies	Sick leave	Unemployed	Combinations of part time work, parental leave, studies, sick leave, unemployment	In a romantic relationship	Have/expect children
Family first	1 (3%)	23 (64%)	1 (3%)	4 (11%)	1 (3%)	1 (3%)	–	5 (14%)	29 (82%)	26 (72%)
Everything is important	–	13 (50%)	6^a (23%)	4 (15%)	–	–	–	3 (12%)	25 (96%)	23 (88%)
Struggling to prioritize	1 (4%)	13 (57%)	8 (35%)	–	–	–	–	1 (4%)	21 (91%)	14 (61%)
Now family comes first	2 (13%)	7 (47%)	–	1 (7%)	2 (13%)	–	–	3 (20%)	14 (93%)	15 (100%)
Inability to prioritize	2 (14.3%)	6 (42.9%)	2 (14.3%)	–	1 (7.1%)	–	3 (21.4%)	–	10 (71%)	4 (29%)
Work first	–	2 (20%)	5 (50%)	–	1 (10%)	1 (10%)	–	1 (10%)	4 (40%)	–

^a Two of these participants also reporting taking parental leave alongside working more than full time.

Family first (29%, n = 36) was the most common identity configuration among the participants and included descriptions of family being the most important thing in the participants’ lives. Work, on the other hand, was viewed as something subordinate, something they did for money or “a necessary evil”, as some participants described it. Prioritizing family above all was often described as something natural or as rooted in personality characteristics. This configuration can be seen in the case of Maria, a nurse and the mother of two children, who at the time for the interview was on parental leave with her youngest child. For Maria, having children had always been more important than having a career, figuring that her more career-oriented partner could be the one to focus on having a career. For Maria, a job was just a job and something to earn a living by. When asked directly about her priority between work and family she said: “It’s so natural in my body and I think that my children come first, so that’s of course my priority.” Thus, for Maria and the other participants coded to this configuration, work was seen as an inferior part of their lives and family was the thing that really mattered for who they were.

The second most common identity configuration was Everything is important (21%, n = 26). This configuration included statements about work and family being equally important for the participants, how they wanted a 50–50 balance, or how they sought to maximize everything. Many of the participants explicitly stated that both family and work were part of their identity and that they felt a reluctance to choose one over another. A clear example of this configuration was seen in Anna, a lawyer who worked more than full time and had a child together with her husband. She said: “Don’t want to prioritize any of it, want them to be equal. For me, it’s really important to have a family, a huge part of life. But to let go of your occupational identity … To be something else that hasn’t got anything to do with children is also very important for me.” Considering both work and family to be highly important often resulted in feelings of being torn, or as Anna phrased it, “being stuck between the two worlds.” Anna explained how everything sometimes reached the limit, and how one must allow oneself to be a “bad” mother now and then, as she really did believe that her work made her a better mother in the long run. This way of thinking about the struggle between work and family was not unusual among the participants, where some talked about setting a good example for their children by caring so much about their occupation.

The third most common identity configuration was Struggling to prioritize (19%, n = 23). The Struggling to prioritize configuration included narratives where participants clearly stated that family was the most important part in their lives even though they did not always show it in practice. Many of the participants said that even though they prioritized family, work took up most of their time. In a few cases, it was the other way around, where the participants said that they would like to prioritize work more than they did. The struggles of prioritizing between work and family were evident for Johan, a father of a small child and an entrepreneur working full time with a new company that he co-founded. Johan felt that not being able to spend time with his child or wife was a risk factor in the long run, and it was something he thought a lot about: “It’s an economic issue really. I want to prioritize family, but in reality, it’s an economic issue and a matter of what I want to do. So it’s a bit hypocritical, I want to prioritize family life and then I actually prioritize the other stuff.” Johan continued with explaining that this was a consistent habit, always wanting to do one thing but then doing another. This inconsistent way of prioritizing, wanting one thing and doing another, was typical for the people coded to the Struggling to prioritize configuration.

The next configuration, Now family comes first (12%, n = 15), included participants who concluded that both work and family were important for them, but that right now the family role had taken priority. This configuration included descriptions of actively taking a step back from having a career at the moment, putting it on hold, even though it was an important part of the participants’ lives and identities. Often, the reason for doing so was because the participants had become parents, started a relationship, or simply felt that they were at another stage of their life at this time. However, many participants were open to the suggestion that this could change later on. Sara, an engineer working part time with a husband, two children, and another on the way, painted a representative case for this configuration when she said: “Yeah but right now where I’m at in life, it’s like, all [parts] are important parts but the children are of course the most important, that’s just the way it is/ … /now is not the time to focus on having a career.” Sara continued with saying that sometimes she asked herself why she should work, especially when leaving the children at daycare. On the other hand, her profession was still very important for her, so it would not work to be a stay at home parent.

For 11% (n = 14) of the participants, labeled Inability to prioritize, there seemed to be a less clear identity configuration of work and family identities. The inability to prioritize between work and family manifested in different ways: either the participants were quite brief in their statements and did not really prioritize or the participants changed their description of their priorities several times during the interview, ending up with very inconsistent narratives. For many of the participants coded to this configuration, neither work nor family seemed to be that important for their identities or something they had even really reflected on. Daniel, an engineer working more than full time, single, and without children, had not thought much about work and family priorities, as he did not see it as important. When asked directly about his thoughts on prioritizing between work and family in the future he answered: “Unclear what I’ve been thinking. Think I’ve thought that it will work itself out”.

Finally, 8% (n = 10) of the participants were coded to Work first, a configuration that included descriptions of work being the participants’ main priority. Sometimes, putting work first was viewed as being an active choice, where the participants had opted out of focusing on family, in order to focus on building a career. For others, family was viewed as something that just happened, and when it did they might change their priorities. What separated the structure of this identity configuration from the Family first configuration, besides the actual priority, is that the participants coded to Work first had prioritized away family, whereas the ones prioritizing family still worked or saw work as necessary. Andreas had devoted most of his life to his career. He said: “Much revolves around work for me. Difficult to be free a long time. It’s becoming stronger and stronger every year/ … /The more I get to know myself, the more it becomes like that. Children are totally insignificant/ … /Prioritized away children, becomes an additional job without pay, rather have that time.” Even though Andreas worked more than full time as a medical doctor, he felt that he had enough time for his wife. According to him, he could even work twenty hours more per week and still have enough time, even though they would have less physical time together. For Andreas, work was the most important part of life and it was not the quantity of time spent with his wife that mattered to him, it was the quality. Working more meant more money, and more money meant more quality.

Gender differences in the configurations

A chi-square analysis showed significant gender differences across identity configurations, χ² (5, N = 124) = 14.16, p = .015, Φ = .39 (see Table 2 for descriptive information on distributions). Following up on the analysis, we analyzed the adjusted standardized residuals (ADJR), indicating that there were only significant gender differences in the Work first configuration, which consisted of only men (ADJR = 3.3).

Table 2. Descriptive statistics (M and SD) for degree of conflict and satisfaction with life across identity configurations separated by gender.

	Total					Women					Men
		Degree of Conflict		Satisfaction With Life Scale			Degree of Conflict		Satisfaction With Life Scale			Degree of Conflict		Satisfaction With Life Scale
	n (%)	M	SD	M	SD	n (%)	M	SD	M	SD	n (%)	M	SD	M	SD
Family first	36 (29)	1.58	.91	5.44	.96	20 (32)	1.53	.72	5.40	.98	16 (26)	1.25	.58	5.49	.95
Everything is important	26 (21)	2.65	1.16	5.85	1.02	17 (27)	2.75	1.18	6.07	1.00	9 (14.5)	2.44	1.24	5.42	.96
Struggling to prioritize	23 (19)	3.39	.72	5.52	.63	12 (19)	3.45	.69	5.58	.66	11 (18)	3.36	.81	5.45	.63
Now family comes first	15 (12)	2.53	1.13	5.73	.63	8 (13)	2.13	1.13	5.95	.53	7 (11)	3.17	.98	5.49	.68
Inability to prioritize	14 (11)	1.79	1.12	4.77	1.50	5 (8)	2.20	1.30	4.68	2.25	9 (14.5)	1.56	1.01	4.82	1.06
Work first	10 (8)	2.00	1.05	4.64	1.18	-	-	-	-	-	10 (100)	2.00	1.05	4.64	1.18
Total	124 (100)	2.31	1.19	5.44	1.04	62 (100)	2.39	1.18	5.63	1.08	62 (100)	2.16	1.20	5.24	.96

Degree of conflict across identity configurations

The degree of conflict experienced by the participants was relatively evenly distributed among the participants, where 36% (n = 45) of the participants did not experience any internal conflict, 18% (n = 23) experienced low conflict, 23% (n = 28) experienced moderate conflict, and 23% (n = 28) of the participants experienced high conflict. There were no differences between women (M = 2.47, SD = 1.17) and men (M = 2.16, SD = 1.19) in degree of conflict, t(122) = −1.45, p = .151, Cohen’s d = 0.26. Level of conflict did, however, vary by identity configuration according to the Kruskal–Wallis non-parametric test, H = 38.79, p < .001 (see Table 2 and Figure 1). Bonferroni post hoc tests (see Table 3) showed that participants with a Family first configuration reported less conflict than those with a Now family comes first configuration (p = .038, Cohen’s d = .93), an Everything is important configuration (p = .001, Cohen’s d = 1.03), or a Struggling to prioritize configuration (p < .001, Cohen’s d = 2.20). Participants with a Work first configuration (p = .005, Cohen’s d = 1.54) and with an Inability to prioritize configuration (p < .001, Cohen’s d = 1.70) reported less conflict than participants with a Struggling to prioritize configuration.

Table 3. Descriptives, p-values, and effect sizes from the Bonferroni post hoc test of conflict and life satisfaction across the identity configurations.

	Degree of Conflict							Satisfaction With Life Scale
	M (SD)	1	2	3	4	5	6	M (SD)
1. Family first	1.58 (.91)	-	1.000/.41	1.000/.10	1.000/.36	.499/.53	.373/.74	5.44 (.96)
2. Everything is important	2.65 (1.16)	.001/1.03	-	1.000/.39	1.000/.14	.019/.84	.019/1.10	5.85 (1.02)
3. Struggling to prioritize	3.39 (.722)	< .001/2.20	.169/.77	-	1.000/.33	.394/.65	.294/.93	5.52 (.63)
4. Now family comes first	2.53 (1.13)	.038/.93	1.000/.10	.165/.91	-	.145/.83	.112/1.15	5.73 (.63)
5. Inability to prioritize	1.79 (1.12)	1.000/.21	.151/.75	< .001/1.70	.701/.66	-	1.000/.10	4.77 (1.50)
6. Work first	2.00 (1.05)	1.000/.43	1.000/.59	.005/1.54	1.000/.49	1.000/.19	-	4.64 (1.18)

The numbers represent p-values followed by Cohen’s d. Conflict results are presented below the diagonal, and Satisfaction With Life above.

Figure 1. Z-scores of conflict and life satisfaction across configurations in total, and separated by gender.

Satisfaction with life across identity configurations

Most participants 77% (n = 95) reported very high or high scores of life satisfaction, 14% (n = 18) reported average scores, and 9% (n = 11) reported lower scores (i.e., slightly below average, dissatisfied, or extremely dissatisfied, see Diener, 2006). Men (M = 5.24, SD = .96) reported significantly lower levels of life satisfaction than women (M = 5.63, SD = 1.08), t(122) = −2.14, p = .034, Cohen’s d = .38. Kruskal–Wallis non-parametric test indicated significant differences in life satisfaction between identity configurations (H = 16.49, p = .006; Table 2 and Figure 1). Bonferroni post hoc test (see Table 3) showed that participants with an Everything is important configuration reported higher satisfaction with life than participants with a Work first (p = .019, Cohen’s d = 1.10) configuration and an Inability to prioritize configuration (p = .019, Cohen’s d = .84).

Unfortunately, we were not able to test for gender differences in life satisfaction across the different identity configurations due to power issues, thus not knowing if it was mainly gender that explained the variance in life satisfaction across identity configurations. However, there was a small positive correlation between degree of conflict and life satisfaction (r = .18, p = .046), which varied somewhat, but not substantially, when separated by gender (women: r = .10, p = .453; men: r = .23, p = .074).

Discussion

The purpose of the present study was to investigate how young adults in Sweden configure their work and family identities. In order to get a wider understanding of these configurations, we also assessed gender differences, the degree of conflict, and life satisfaction associated with the different configurations. Taking a higher level perspective on the findings, the configurations differed conceptually on two dimensions (see Figure 2): 1) choosing (i.e., Family first, Work first, Struggling to prioritize) or not choosing (i.e., Everything is important, Inability to prioritize) one identity over another, where Now family comes first was more in the middle, and 2) level of certainty (i.e., Family first, Work first, Everything is important, Now family comes first) or ambivalence and discrepancy (i.e, Struggling to prioritize, Inability to prioritize) in their prioritization between work and family.

Figure 2. Model of the identity configurations.

Corresponding with previous research (see Hammack, 2010; Schachter, 2004), choosing one identification and suppressing another (i.e., Family first, Work first) seems to be a feasible way to structure multiple identifications in order to minimize conflict. Not being able to prioritize accordingly in everyday life, as seen in those with a Struggling to prioritize configuration, might heighten the degree of internal conflict. The inconsistency between wanting and doing might also mirror the inconsistencies seen in Swedish society, where the norms and policies do not always match everyday practices (Almqvist et al., 2011; SCB, 2018). Further, the results indicate that high conflict does not necessarily need to be associated with lower subjective well-being. As seen in the Everything is important configuration, having both a strong occupational identity and a strong family identity may come with a heightened sense of life satisfaction compared to only having a strong occupational identity or not having a strong identification at all in these domains. This suggestion corresponds with the notion that life satisfaction is highly influenced by social relationships, such as family and friends, and performance-based roles, such as occupation (Diener, 2006).

The inconsistency with previous research showing a negative relationship between work-family conflict and well-being (e.g., Hagqvist et al., 2017; van der Lippe et al., 2006), and the present finding indicating a positive relationship between the two concepts could be explained by the identity focus in the present study, something that is lacking in prior work-family conflict studies. Having integrated multiple identifications could thus be a protective factor against negative experiences compared to only having one or no identification (Barnett & Hyde, 2001). These results may also be reflective of how the men focusing on their work and the participants having neither a strong occupational nor family identity reported experiencing lower degree of life satisfaction, compared with those for whom everything is important. The lower degrees of life satisfaction may also be a result of not living up to the cultural expectations in Swedish society, both in terms of having it all and family-centered values (see Elvin-Nowak & Thomsson, 2001; Sommestad, 1997).

The identity configurations with less certainty in their choices between occupational and family identifications (i.e., Struggling to prioritize, Inability to prioritize, and to some extent Now family comes first) may highlight how identity configurations are an ongoing process that can be renegotiated over time (Dahl & Galliher, 2012; Hammack, 2010; Syed, 2010). It may be that the participants struggling with their priorities will re-negotiate and restructure their identities in order to reduce internal conflict. Similarly, the ambivalence shown by the participants coded to the Inability to prioritize configuration could indicate that this configuration is in its early stages formalizing, thus still being very loose without a clear structure. It may also be that occupation and family are not important identity domains for some people. There are, thus, similarities with the concept of identity diffusion, which has been characterized by a lack of direction in life (Kroger & Marcia, 2011). Putting their occupational identities on hold, as found among the participants coded to a Now family comes first configuration, further adds to the suggestions that these negotiations are an ongoing process. However, more longitudinal research on identity configurations is needed in order to see whether or not the participants do reconstruct their identity configurations over time.

The results of this study are likely to mirror the norms in Swedish society in the types of identity configurations that are encouraged, expected, and viable for young adults. The findings that the Family first configuration was as common among women as among men might be unexpected given that relational identity is often considered more important for women (e.g., Archer, 1989; Fadjukoff, Pulkkinen, & Kokko, 2005). However, the lack of gender differences found across all types of identity configurations, except for in the least common Work first configuration that included only men, might reflect the Swedish norms emphasizing gender equality – especially in the domains concerning work and family (Duvander, 2014; Johansson & Klinth, 2008). Many of the men coded to a Work first configuration were open to the suggestion that when “family happened,” they might change their priorities. This reasoning may be a way for some of these men to handle not living up to the norms and expectations of Swedish society, in terms of both gender role norms and family norms. Furthermore, the findings of the Family first being the most common and the Everything is important being the second most common identity configurations corresponds with the dual norms in Sweden where there are both strong family-centered norms as well as strong norms of gender equality (Elvin-Nowak & Thomsson, 2001; Sommestad, 1997). It may be that in the gender equality context of Sweden, it is more accepted for men as well to put family first compared with contexts with more traditional gender role norms. However, there might still be gender differences that we did not capture due to sample size constraints. Future research could benefit from investigating identity configurations between occupational and family identities using larger samples across cultural contexts in order to more fully understand how culture influences identity structure.

Limitations

Even though this study adds valuable knowledge there are limitations that need to be addressed. First, the sample size limited the analyses possible in this study, and it would be valuable to test for main and interaction effects of gender in the relationship between identity configurations and life satisfaction in future work. Second, the study relied on self-report measures, which may raise issues of social desirability corresponding with norms in Swedish society, and thus might not reflect the participants’ actual priorities in everyday life. However, our study focused on the participants’ own reflections of negotiating between work and family and not the specific division of time. Third, the structure and content of the Identity Status Interview may have yielded specific types of configurations and pushed aside others. It is possible that another interview protocol would lead to a different set of configurations. There may also be other important identity domains not addressed in the current study that could interact with work and family, thus potentially affecting both their well-being and experience of conflict (e.g., politics). Additionally, the configurations found may be specific to the domains of occupation and family, and that other types of identity configurations would be found in other identity domains. Future research should include more domains when trying to understand the complexity of identity configurations. In addition, this study used a cross-sectional design which inhibits us from drawing any conclusions on how the identity configurations and its processes evolve over time. Thus, more longitudinal research is needed in order to understand the temporal aspect of how identity is structured. It should also be noted that even though conflict and life satisfaction were correlated positively, there might be other negative psychological outcomes (e.g., stress) associated with high conflict that we did not assess. Another possible way of understanding this positive correlation is that participants in high conflict configurations may be more prone to social desirability. However, there are studies arguing that social desirability does not affect subjective measures of well-being (e.g., Pavot & Diener, 2009). It should also be acknowledged that we did not use a standardized measurement of internal conflict, which might have been fruitful, as our measure of conflict may suffer from issues of internal consistency, especially since we only re-coded 30% for inter-rater reliability. Furthermore, it could be useful for future research to separate intensity and frequency when coding for internal conflict as these dimensions of conflict may entail qualitative differences important for understanding identity structure. Despite these limitations, this study brings valuable information on identity configurations, a research topic that has been understudied.

Conclusions

In conclusion, these findings further add theoretically to prior research on identity structure in that there are numerous ways of configuring multiple identities and further underlines that the way young adults structure core aspects of their identities – and lives – is heterogeneous and closely connected to the cultural context in which those identities develop. The findings from this study reflect the complex and dynamic ways in which identity is structured in various types of identity configurations and further emphasizes the notion that identity configurations are a dynamic ongoing process of negotiation. Our findings advance the study of identity configurations by examining the identity domains of occupation and family, which have not been examined previously, and by actually assessing the degree of conflict associated with the configurations and links to well-being, this general approach will be useful for subsequent researchers who are interested in embracing the complexity of identity development while seeking to understand it.

Acknowledgments

The authors of this paper want to further acknowledge the valuable contribution, time and effort put in by all the participants as well as the project group involved in the data collection.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Preparation of this manuscript was supported by Grant from Swedish Research Council for Health, Working Life and Welfare (FORTE).