Digitalization has led to the development of the world of work described as the “fourth industrial revolution” (Neufeind et al., 2018; Ropponen et al., 2020). An important aspect of this development is the increasing degree of freedom in individual work design and beyond. This freedom can be used by employees to shape their work individually and proactively through crafting (De Bloom et al., 2020; Tims et al., 2022). At the same time, more flexible work has lead to collapsing work-to-nonwork interfaces (Vaziri et al., 2020) and a vulnerable work-nonwork balance¹ (WNB). Proactively crafting these interfaces and one’s work-nonwork balance constitute the core of this research.

The need for crafting WNB has mainly been fuelled by the trend to integrate work into other life domains due to: (a) the extended use of information and communication technologies (Piszczek, 2017), resulting in blurred boundaries between work and nonwork life domains (Ollier-Malaterre et al., 2019); (b) the demand for highly flexible work arrangements in a 24/7 economy (Bauer and Brauchli, 2017); and (c) organizational practices that encourage employees to expand work into nonwork life domains (Dumas and Sanchez-Burks, 2015). This trend is accompanied by increasing work density, which is the ratio of one’s workload over the resources available to perform that work (Derks and Bakker, 2014) and few opportunities for necessary recovery from work stress during and after work (De Bloom et al., 2015). In addition to the densification of work, demands in nonwork life domains are remaining or also increase (Rofcanin and Anand, 2020), leading to precarious situations for family life (Beckman and Mazmanian, 2020). Finally, the COVID-19 crisis, with mandatory home office regulations and lockdowns, has intensified this development (Cho, 2020; Vaziri et al., 2020; Rudolph et al., 2021). Proactive work designs including crafting are suggested as helpful strategies during this pandemic (Wang et al., 2020; Brauchli et al., 2022; Pijpker et al., 2022). Empirical evidence also highlights the importance of WNB for the well-being and health of employees (for reviews see Casper et al., 2018; Sirgy and Lee, 2018).

In summary, proactively balancing work and nonwork via crafting may comprise a beneficial behavioral strategy to improve WNB. To enable research and, later on, the dissemination of such proactive strategies, the present paper aimed to develop and validate the Work-Nonwork Balance Crafting (WNBC) scale.

The first critical step in developing a new scale is defining the guiding concepts. This is demanding for WNB because currently, there is no consensus regarding guiding theoretical models and conceptual definitions available (Shockley et al., 2017; for an overview and review, see Casper et al., 2018; Wayne et al., 2021). Thus, we relied on Casper et al.’s (2018) thorough definition of WNB derived from a deductive, comprehensive literature review and dictionary classifications, as well as an inductive, qualitative analysis of employees’ definitions of such a balance:

This definition incorporates the fit perspective to be satisfied with valued roles in their respective life domains (Greenhaus and Allen, 2011; Wayne et al., 2021). In addition, Casper et al. (2018) considered a current development in the WNB literature by exceeding the perspective of WNB as a balance between the domains of work and family only (Haar et al., 2019). Ideal balance is defined as a good commensuration between affective experiences and involvement and effectiveness in work and nonwork roles with value attached to these roles. To attain this fit, we proposed that individuals can engage in needs-oriented proactivity, the so-called crafting (De Bloom et al., 2020). The given WNB situation must not be passively accepted to satisfy these needs. Instead, we assumed that it can be beneficial if employees proactively adjust the WNB situation according to their own standards and respective role expectations [see Casper et al.’s (2018) WNB definition above]. For example, a father may adapt and therefore craft the balance of work and family life and, accordingly, his ideal WNB. Specifically, he might proactively craft his communication behavior at work by telling colleagues when he is unable to communicate with them during leisure time. Consequently, the development of our scale links the above WNB definition to the crafting concept considering an individual to be the effective agent of their WNB.

Thus far, we described the relevance and development of WNB conceptually and outlined its relevance in the current world of work. In this section, we linked the origin of crafting with WNB to devise a new crafting construct beyond job crafting. Before we apply the crafting concept to the new domain of WNB crafting, we first review the well-established job crafting concept. Job crafting has been referred to as the self-initiated behaviors that employees take to shape, mold, and change their jobs (Wrzesniewski and Dutton, 2001; Tims and Bakker, 2010; Tims et al., 2012; Zhang and Parker, 2019). Crafting can help satisfying psychological needs and exhibits favorable outcomes, such as employee performance and well-being (for a review see Rudolph et al., 2017; Zhang and Parker, 2019; Mäkikangas and Schaufeli, 2021). Wrzesniewski and Dutton (2001) initially described this concept as a social constructivist approach that refers to “the physical and cognitive changes individuals make in the task or relational boundaries of their work” (p. 179). Wrzesniewski and Dutton distinguished three types of job crafting: (1) the changes employees make to adjust their work tasks (task crafting); (2) the quality and frequency of the relationships they have at work (relational crafting); and (3) the subjective meaning they assign to their work (cognitive crafting). These changes may be intentional and affect “the meaning of the work and one’s work identity” (Wrzesniewski and Dutton, 2001). This perspective on job crafting has inspired a large field of research because it helps, at least partially, overcome formal job constraints and invites new opportunities for individual work redesign.

Subsequently, Tims et al. (2012) integrated the crafting concept into the job demands resource (JD-R) model (Demerouti et al., 2001; Bakker and Demerouti, 2017) as a way to balance demands and resources and establish a person-job fit (Zhang and Parker, 2019). The JD–R model is a person-centred theoretical framework (Fan et al., 2019) and characterises crafting as a “self-initiated” and “self-targeted” individual-level strategy to increase person-job fit (Tims et al., 2012). Job demands (e.g., challenging and hindering demands) refer to aspects of a job that require sustained physical, emotional, or mental effort, whereas job resources (e.g., structural and social resources) refer to job aspects that stimulate personal growth and development while being functional in achieving work goals and simultaneously reducing job demands (Demerouti et al., 2001; Bakker and Demerouti, 2007; Lesener et al., 2019). Crafting, then, is defined as “the changes that employees may make to balance their job demands and job resources with their personal abilities and needs” (Tims et al., 2012). Extending the perspective on crafting in the light of resources and demands, Costantini et al. (2021) showed that employees can also restore the fit between their demands and preferences by optimising their demands instead of only decreasing them, see also Demerouti and Peeters (2018). This insight offers a new perspective to crafting research. Importantly, such demands optimizing crafting expands the work characteristics and tailors the work process to be more efficient by eliminating obstacles and simplifying procedures (Costantini et al., 2021). In summary, crafting includes the perspective that individuals can adapt their job to improve its fit to their abilities, needs, and preferences (Wrzesniewski and Dutton, 2001; Lichtenthaler and Fischbach, 2019).

The field has recently proceeded to transfer the concept of job crafting to life domains other than work, such as home crafting (Demerouti et al., 2019), off-job crafting (De Bloom et al., 2020), life crafting (Schippers and Ziegler, 2019), and leisure crafting (Berg et al., 2010; Petrou and Bakker, 2016). Crafting may also help achieve the requirements of modern work to nonwork arrangements. Accordingly, we assumed that crafting allows an individual to create an idiosyncratic balance of work and nonwork.

Our scale builds on a pioneering qualitative study by Sturges (2012) who defined crafting for WNB “as the unofficial techniques and activities that individuals use to shape their own work-life balance” (p. 1540). There, such crafting is characterized as self-initiated (Kossek and Ozeki, 1999), and goal-oriented behavior (Parker et al., 2010) proactively taking control over one’s WNB (Clark, 2000). Overall, it is driven by preferred role configurations in the respective life domains.

Sturges (2012) identified the three following crafting strategies for WNB, on which the development of our scale is established:

1. Physical crafting includes behaviors, such as time management, selection, and alternation of work location (e.g., leaving work early to do some necessary personal chores). The qualitative interviews indicated two subcategories: (a) Temporal crafting is about managing the time spent at work to achieve a WNB. An example here is finishing work on time, that is, adhering to contracted working hours. This factor might not entirely be the employee’s decision and necessarily be proactively negotiated with a supervisor, e.g., to avoid conflicts and synchronise work schedules. Therefore, temporal crafting may also involve relational crafting, which is outlined below. However, temporal crafting also refers to after-work time, that is, committing time to an event in the evening, such as sports or theatre. (b) Locational crafting is reported as occasionally choosing to work from home instead or in addition to working at the office. Thereby, locational crafting can help in accomplishing family chores or in reducing the strains of commuting. Note that these categories did not include choosing a job and reducing travel time by moving to live near work, since these factors are substantial life changes that go beyond typical crafting strategies that can vary day by day.

2. Relational crafting involves managing and using relationships at work and at home to secure and reinforce the kind of WNB that an individual wants to achieve (Sturges, 2012). Furthermore, relational crafting is structured in two sub-types: (a) Managing out-of-work relationships refers to socializing with people working on the same working times, which helps maintain their belief that one’s concept of WNB is typical. (b) Managing work relationships occurs by reducing unnecessary interactions at work and workloads; for example, when work is extended because the individual wants to reach a goal in the work domain, the quality of relations can be ensured by communicating this proactively to relevant persons in the work context, also management.

Finally, (3) cognitive/emotional crafting involves defining and framing the perceptions of what WNB means and entails (Sturges, 2012): Conceptualization and the definition of an idiosyncratic orientation and balance toward work and nonwork (e.g., meeting social engagements during the week despite regular long working hours), prioritizing work instead of the nonwork life domain (e.g., prioritizing work and highlighting the relevance of work-related achievements), and finally, compromising the ideal WNB to reach long-and short-term goals as a compromise in balancing both life domains, (e.g., investing long working hours for a sprint to reach a work-related achievement). We renamed Sturges’ (2012) dimension of cognitive crafting into cognitive/emotional crafting to integrate affective aspects as these aspects are particularly relevant for work to nonwork conflicts or enrichment (French et al., 2018; Wayne et al., 2022). Moreover, this component seems important as (role) balance has cognitive and affective elements (Casper et al., 2018). Besides the prioritization of work as orienting principle of one’s WNB we consequently add prioritizing non-work aspects for balancing both life domains. Further below, we outlined how the items of the WNBC scale align with these crafting techniques, see also Table 1.

Considering the eroding work-nonwork boundary (Allen et al., 2020; Vaziri et al., 2020), the balance, as well as the boundary, between the work-nonwork domains must be crafted.

The concept of boundary management (Sturges, 2008; Kossek et al., 2012) refers to the active shaping of boundaries. It is defined as a “construct that reflects our mental models about the permeability of the relationship between multiple life roles, our preferences about how to manage those relationships, and our choices and constraints regarding how we enact those preferences” (Rothbard and Ollier-Malaterre, 2015).

Boundary management practices are related to both interference and enhancement processes across life domains (Bulger et al., 2007). These boundary management practices are particularly linked to the successful integration of multiple important life roles (Rothbard et al., 2005; Kossek et al., 2021). Such life roles are crucial in the WNB definition used in our study (Casper et al., 2018; Vaziri et al., 2020) and are consequently important for WNBC. Research referring to the qualities of the boundaries between life domains typically defines core characteristics of the boundaries between work and other spheres of life that are relevant for WNBC. As such, permeability refers to the extent to which psychological and behavioral aspects can diffuse through the boundaries one has set (Ashforth et al., 2000; Clark, 2000). Second, flexibility means the contraction or expansion of a domain regarding its temporal and spatial constraints and is oriented toward requirements in either life domain (Hall and Richter, 1988). For example, if family chores are plenty, this allows extending the time spent within this life domain (e.g., leaving work early or reducing daily working time for care duties [flexibility] or answering calls from family members during working time [permeability]).

Moreover, the active configuration of work-nonwork boundaries is conceptualized on a continuum from segmentation to integration (Ashforth et al., 2000; Kossek and Lautsch, 2012; Wepfer et al., 2018). Segmentation refers to strict boundary-setting and inflexible and impermeable role boundaries. Integration is characterized by flexible and permeable role boundaries. Therefore, segmentation/integration characterizes the extent to which work and nonwork roles are separated.

Specifically, the nomological net for the item development of our WNBC is built on the work-home boundary theory (Ashforth et al., 2000). Ashforth et al. (2000) refer to roles that hold expectations, rules, and norms in respective life domains and converge with the conceptualizing of roles in the WNB definition by Casper et al. (2018), which underlies our WNBC scale. These roles and role transitions between life domains are characterized as psychological, physical, and temporal constructs. Thus, this theory aligns well with the physical and cognitive-emotional dimensions of Sturges’ (2012) WNBC techniques.

In the item development of the WNBC scale, we involved the proactive boundary management behaviors outlined in detail in the item description further below.

But here, we can summarize our conceptual basis of our scale development by providing the following complete definition underlying our scale: WNBC entails the unofficial techniques and activities individuals use to shape their own work nonwork balance under consideration of their boundary preferences and their favored combination of work and nonwork roles.

Recently, Gravador and Teng-Calleja (2018) developed a work-life balance crafting behaviors survey. Like our scale, this instrument refers to Sturges (2012), but invokes a different theoretical framework. This instrument measures behaviors revolving around taking time off from work, fostering relationships with family and others, and working efficiently. The 25 items of the instrument are cumulated in eight clusters of proactive work-life balance crafting behavior themes (Gravador and Teng-Calleja, 2018).

Several shortcomings of this former approach are: First, Gravador and Teng-Calleja’s (2018) instrument contains the physical and relational crafting dimensions from Sturges’ (2012) concept, whereas the essential cognitive dimension is omitted. Cognitive crafting refers to framing and redefining WNB and shapes how employees view their WNB without engaging in specific behaviors. Therefore, we added this dimension to our scale. Second, Gravador and Teng-Calleja (2018) reported that only one of the eight scale dimensions, that is “working efficiently,” is associated with subjective well-being (satisfaction with life scale; Diener et al., 1985). In a second model, only two of the eight scale dimensions turned out to be related to work-life balance (work-life balance scale; Brough et al., 2014): “working effectively” and “saving private time.” These findings provide insights into the relative importance of crafting efforts but also shows that only few crafting dimensions of an extensive crafting scale matter. Accordingly, measuring crafting with fewer dimensions and items seems a more efficient way when examining links to well-being and WNB. Third, the instrument presented by Gravador and Teng-Calleja (2018) contains 25 items and eight clusters, and we determined the need for parsimonious instruments for, e.g., measurements in digital applications and online surveys. Besides the number of items, the complex structure and the high number of factors call for new scale development, building on the valuable results reported by Gravador and Teng-Calleja (2018). To the best of our knowledge, a WNBC scale that covers all three sub-dimensions proposed by Sturges (2012), that is, physical, relational, and cognitive/emotional crafting, has never been established. We aim to address this by developing the proposed new WNBC scale.

Moreover, three additional reasons are in favor of developing a new scale. First, the WNBC scale can potentially produce new opportunities for research by integrating two very productive and timely research streams, namely, WNB (Greenhaus and Callalan, 2020) and crafting (Hu et al., 2020). Second, our new scale can inform occupational health interventions, for which corresponding research and the development of an intervention are in progress. Third, we proposed a parsimonious two-factor structure of this scale, covering crafting in work and the nonwork life domains. Based on the presented considerations, our approach has advantages compared to previous approaches.

The participants were recruited through an online panel data service in Austria, Germany, and Switzerland. All items were presented in German. Participation was voluntary and anonymous, and the confidentiality of their data was guaranteed. Persons who declared that they were under 18 years of age, unemployed, self-employed, or worked less than 9 h a week were not included. We excluded self-employed individuals because they represented a small and divergent group and because the social context at work is a relevant factor for crafting, even if crafting is a bottom-up strategy (Kerksieck et al., 2019; Tims and Parker, 2019). A total of 330 participants completed the questionnaire in April/May 2018. We used a post hoc multivariate outlier statistic to assess data quality controlled for Mahalanobis (1936) distance, which is also recommended for online studies (Niessen et al., 2016). The participants who answered the questionnaire in less than 5 min were classified as speeders (N = 7) and were excluded with multivariate outliers (N = 3), resulting in a sample size of 320 participants. The data were analyzed with SPSS 28.

Exactly half of the resulting sample was female. The average age of the participants was 43.96 years (SD = 12.11). Participants from Germany (63.3%), Austria (19.1%), and Switzerland (17.6%). A percentage of 46.3 worked 40–44 h per week. The average organizational tenure was 12.08 years (SD = 10.66). Most participants had completed vocational education (40.1%) or had a university degree (20.1%). The largest groups were employed in healthcare/social services (15%), public administration (12.2%), and commerce (10.3%).

The WNBC scale was assessed with 16 items as developed in study 1 (Table 1). The reliability was McDonald’s ω = 0.67 for the WNBC-nonwork dimension and ω = 0.70 for the WNBC-work dimension. The WNBC scale was available in German, and other measures were translated from the published version in English to German and back-translated for accuracy.

The work-life indicator (Kossek et al., 2012) was assessed. Sample items read for the work interrupting nonwork subscale “I work during my vacations“, and for the nonwork interrupting work subscale “I do not think about my family, friends, or personal interests while working so I can focus“. Items were answered on a five-point scale ranging from strongly disagree to strongly agree. In our study, the subdimension’s reliability was ω = 0.80 for the work interrupting nonwork subscale and ω = 0.78 for the nonwork interrupting work subscale.

Life satisfaction and job satisfaction were measured using single-item measures adapted from Broeck et al. (2010): “How satisfied are you when you look at your private life as a whole?” and “How satisfied are you when you look at your professional life as a whole?” Both items were answered on a scale ranging from 1 (extremely dissatisfied) to 7 (extremely satisfied). Single-item measures are frequently used to assess general satisfaction (Lucas and Donnellan, 2012).

For the test of non-random sampling, we used multiple logistic regression (Goodman and Blum, 1996) while the dependent variable was coded dichotomously containing participants that either dropped out or participated in both study waves. The models included variables presented in the hypotheses (job/life satisfaction, work-life indicator). Nagelkerke (NK) R² indicated that the explained variance in all estimated models was not substantial; therefore, none of the assessed variables indicated presence of non-random sampling: job satisfaction T1 (B = 0.04; SE = 0.03; p = 0.21; NK R² = 0.00), life satisfaction T1 (B = 0.03; SE = 0.03; p = 0.43; NK R² = 0.000), and work-life indicator T1 (B = −0.25; SE = 0.08; p < 0.001; NK R² = 0.007).

Before hypothesis testing, we replicated the CFA of the two-factorial structure of the WNBC scale with the largest sample in this longitudinal study at wave 1 (N = 2,014): χ² = 674.996, df = 85, CFI = 0.908, IFI = 0.909, RMSEA = 0.057, and SRMR = 0.062, see Figure 1.

When testing for incremental validity, the outcomes of life/job satisfaction at T2 were simultaneously regressed on the WNBC scale, the work-life indicator, and the respective outcomes (life/job satisfaction) at baseline T1. The results yielded by hierarchical linear regression analysis are shown in Table 4.

Baseline-adjusted hierarchical linear regressions revealed that the WNBC scale explained the variance in two comprising and stable constructs above the work-life indicator (Table 4). In the first step, we included the respective variable at baseline (T1, accounting for a substantial amount of variance in the respective construct (job/life satisfaction) at T2). In the second step, the included work-life indicator did not add explained variance to the regression model. In step 3, the WNBC scale increased the amount of explained variance and improved the model significantly. In addition, such crafting showed a pattern of domain-specific predictors, confirming hypothesis 5.

To test for criterion validity, we correlated the WNBC scale with external variables (Campbell, 1960) and used established constructs to test whether such crafting yields WNB-related outcomes. The validation criteria for the work dimension of the WNBC scale are work engagement and job performance. Subjective vitality and family role performance were evaluated as validation criteria for the nonwork dimension of the WNBC scale. We selected this set of variables to evaluate concepts related to crafting efforts for work-life balance that are directly linked to employee wellbeing and performance in both domains of life. Work engagement and job performance are frequently used to assess employee’s conditions in the work domain, also in combination with work-life balance (Johari et al., 2018; Wood et al., 2020). To mirror these concepts in the nonwork life domain, we assessed family role performance, since this concept relates well to both, performance in a typical nonwork domain is associated with WLB due to, for example caring duties but respective role expectations are also relevant in the WLB definition we are working with. Subjective vitality can be an important consequence of WLB promotion, such as sustainable careers (Kossek et al., 2014) and is transferred into both life domains as a relevant and dynamic reflection of well-being (Ryan and Frederick, 1997).

Moreover, we tested whether the WNBC scale is invariant and if these validation criteria hold across different countries and work cultures. Since the scale is developed first in German we use the largest German speaking sample as a reference. Finally, the test–retest reliability and the stability of associations with respective criteria over time were tested.

To test the assumption that WNBC has substantial and time-stable effects on respective outcomes, we hypothesized the following:

Before testing the full structural equation model (SEM), we provide a CFA of the WNB scale (Wayne et al., 2021). To our knowledge, this study is the first to apply this WNB scale in the German language (χ² = 557.360, df = 164, CFI = 0.972, TLI = 0.968, SRMR = 0.030, and RMSEA = 0.051). A SEM for the association of the new WNBC scale dimensions and a WNB applying the scale for measuring such balance by Wayne et al. (2021) indicated the relevance of crafting for WNB (Table 7). For each of the WNB subdimensions, the WNBC scale delivered significant results and accounted for substantial variance in the outcome: WNBC-work was positively associated with WNB global balance (β = 0.48, p < 0.001), WNB involvement (β = 0.37, p < 0.001), WNB effectiveness (β = 0.56, p < 0.001) and WNB affective (β = 0.38, p = 0.001). Accordingly, the dimension of WNBC-nonwork was positively associated with the WNB global balance (β = 0.31, p = 0.006), with WNB involvement (β = 0.43, p < 0.001), WNB effectiveness (β = 0.37, p < 0.001) and WNB affective (β = 0.38, p < 0.001). The parameters for the model fit of the SEM indicated good fit (χ² = 1621.431, df = 566, CFI = 0.943, TLI = 0.937, RMSEA = 0.045, and SRMR = 0.050). Thus, hypotheses 10 and 10a–c were confirmed.

In study 1, we developed the scale’s items by building on earlier conceptual development and expert feedback. The following exploratory factor analysis yielded a two-factorial structure with eight items in each scale dimension—one capturing crafting on aspects of the work-life domain, and the other on aspects of the nonwork-life domain. Both scale dimensions cover physical, relational, and cognitive/emotional WNBC. We consider this parsimonious two-factor structure of the WNBC scale to be an advantage compared with earlier approaches for measuring work-life balance crafting, containing eight clusters (Gravador and Teng-Calleja, 2018). In addition, our new scale contained the previously omitted cognitive dimension identified by Sturges’ (2012) study and extended this dimension, including important emotional aspects.

Furthermore, in study 2, the confirmatory factor analysis revealed that a two-factor model reached good fit indices, outperforming the one-, three-and six-factor models. Additionally, a test for measurement invariance between the samples of studies 1 and 2 indicated that the WNBC scale was robust across several measurement occasions. Further, the WNBC scale dimensions correlated positively with proactive personality and personal initiative, displaying convergent validity to proactivity—a core crafting element (Bakker et al., 2012).

Interestingly, WNBC-work showed a higher correlation with proactive personality and personal initiative than WNBC-nonwork. Seemingly, proactivity and initiative-taking on a trait level were expressed considerably concerning the life domain of work. This concept may be explained by the fact that WNBC at work occurs in a highly formalized and hierarchical context with the informal, proactive private environment, thus requiring more proactivity to overcome the formal constraints of crafting this domain. This finding offers a new avenue for future research to extend the existing knowledge of personality traits for WNB crafting. This call was additionally fuelled by recent research on personality traits and crafting (Oprea et al., 2019) and should involve research on gender roles that may manifest in WNB decisions (Adamson et al., 2022).

The results of study 3 indicated the incremental validity of the WNBC. While WNBC was predicting job/life satisfaction above the work-life indicator, the latter was not significantly predicting these outcomes in any conducted analyses. Crafting for WNB was significantly associated with increased satisfaction in both life domains. This finding was compelling since many other contributing factors drive life/job satisfaction (Heller et al., 2004), explaining why added explained variance is not large per se. This is mainly due to the variety of stable factors (Ilies et al., 2019) relevant to life and job satisfaction.

We assumed that we could explain job and life satisfaction variance beyond the work-life indicator because we combined both crafting the WNB and boundaries in a meaningful way. This strategy was done because WNBC goes beyond merely allowing or preventing life domain transfers as suggested in the strategies of segmentation or integration (e.g., Bulger et al., 2007). The WNBC scale added crafting techniques for qualitatively shaping these transfers and respective role transitions. Because of increasingly blurred boundaries and working from home regulations before and during the COVID-19 pandemic, resulting in increased work-nonwork interference, it was important taking these boundaries more into account. Both developments resulted in less physical and time-bound boundaries (Allen et al., 2020; Vaziri et al., 2020), making WNBC an essential behavioral strategy.

Study 4 involved data from German-speaking countries, Finland, and Japan, which tested for invariance, criterion validity, test–retest reliability, and intercultural applicability of the WNBC scale. Analyses for invariance across these countries indicated that the WNBC provides metric measurement invariance. The absence of scalar invariance can be subject to different contextual factors across cultures, specifically concerning invariance testing across Japan, Austria, Germany, and Switzerland. However, strict invariance is challenging to achieve in heterogeneous groups (Clench-Aas et al., 2011) and may therefore be difficult to reach in cross-cultural research. Such measurement variance may occur due to differences in (work) cultures (Zhou et al., 2019), as has also been shown in research on proactivity and WNB (Smale et al., 2019). For example, in comparing a Chinese sample with a British or Spanish sample, Nielsen et al. (2017) showed a lack of factor loading invariance in the job crafting questionnaire potentially caused by cultural differences. Nevertheless, crafting scales may provide meaningful results in various countries, but cultural comparisons should be conducted cautiously (Schachler et al., 2019). Cultural differences may also be related to work-life balance, as discussed below and may thus affect cross-cultural invariance testing.

Analysing data from the Japanese and Finish samples showed correlations with external criteria of job performance (Kessler et al., 2003), work engagement (Schaufeli et al., 2006), family role performance (Chen et al., 2014), and subjective vitality (Ryan and Frederick, 1997) were positive at the cross-sectional level. Moreover, positive associations between WNBC and these concepts can be found in longitudinal data from Finland and Japan, indicating the relevance and stability of WNBC outcomes. Beyond this generalizability of the scale to different cultural contexts, the WNBC scale can unravel differences related to work culture. The WNBC scale indicated measurement variance for scalar measurement invariance across work cultures, which can be interpreted in a way that this scale seems sensitive to cultural differences in WNBC. Thus, the WNBC scale offers a measure for capturing such differences, but for comparing such crafting across countries, respective cultural differences need to be considered (Zhou et al., 2019). This is indicated in systematic variation of findings across countries: Individuals exerting WNBC-nonwork reported increased family role performance and subjective vitality in Finland. A result supported by the theoretical underpinning of Leslie et al. (2019) and Hofstede (2011), referring to Finland as a highly feminist work culture. In Finland, cultural norms concerning work and nonwork roles may support crafting for a WNB focusing on the nonwork domain. This finding also supported the compensation hypothesis (Beigi et al., 2019), according to which undesired states in one domain were compensated for in another life domain, as shown, for example, for leisure crafting (Petrou and Bakker, 2016). Besides, WNBC-nonwork correlated with work engagement in the sample from Finland but not in Japan. WNBC in the nonwork life domain may restore resources in Finnish participants that they were able to transfer to engagement in the work domain. This concept is an effect that has been studied in the context of sustainable careers (Kelly et al., 2020) and to which the WNBC scale can add further knowledge in future research.

Finally, in study 5, on another set of more than 900 employees, we inquired how WNBC efforts cross-sectionally contributed to the outcome of balancing work and nonwork. Both the work and nonwork dimensions of the WNBC scale were associated with the balance of both life domains on a global unidimensional and multidimensional formative construct (Wayne et al., 2021). Interestingly, employees’ appraisal of how well they combine work with nonwork roles on the global construct was positive in individuals who crafted their WNB regarding the work domain. The WNB of our European sample may be more affected by the work domain since they assign more relative importance to this life domain (see Leslie et al., 2019 for work priority beliefs). Therefore, (a) crafting work in comparison to nonwork may be more important to arrive at a positive evaluation of the “combination of work and nonwork roles” and (b) crafting the work domain can lead to satisfaction with individual standards that focus more on work-related achievements (Kelliher et al., 2019). However, crafting the nonwork life domain also substantially contributed to a positive global evaluation of WNB. In detail, crafting both life domains led to favorable judgments of items such as “Overall, my work and nonwork roles are integrated” and “My work and nonwork roles are combined in ways that are harmonious.”

This reasoning may also explain why WNBC-work contributed more to a positive evaluation of effectiveness balance. Crafting for WNB, focusing on work-related aspects, helped employees arrive at a highly proactive role balance. Performance and successfulness as key terms in these items may have spurred the link to crafting in the work domain because this domain is perceived as more performance-based (Kim, 2014). In addition, Wayne et al. (2021) suggested that work or family design factors (e.g., the significance of these life domains) can determine how performance in either life domain is judged. This factor may have contributed to the relative importance of work-related crafting aspects. However, crafting in both life domains was relevant for a positive appraisal of WNB effectiveness.

Involvement balance was strongly associated by WNBC-nonwork. Since salient work-related demands that call particularly for role involvement in the work domain for many employees may exist, WNBC-nonwork helped conclude a positive appraisal regarding the desired balance of role involvement in both life domains. This perspective on WNB would fit in with the “expandable-pie” perspective, stating that involvement in one role expands (i.e., enriches) resources for another role (Leslie et al., 2019; Rothbard et al., 2021). Further research is necessary to understand the role of WNBC in enrichments or even gain cycles, as has already been found for job crafting (Vogt et al., 2016). Nevertheless, crafting in both life domains was again relevant for balanced involvement in both life domains.

Our analysis indicated that crafting for WNB in both life domains was similarly associated with affective balance. Showing WNBC efforts led to a more positive affective balance and, therefore, more positive and fewer negative emotions in highly valued roles across life domains. This finding was important since Marks and MacDermid (1996) outlined that role balance involves affective and cognitive elements. The results indicated that the WNBC scale can capture proactive efforts that improve both the cognitive (involvement/effectiveness) and affective dimensions of balance.

The results of study 5 were in favor of the WNBC scale, as it captured crafting efforts for attaining a WNB. Here, we demonstrated that crafting for WNB, as measured with this new crafting scale, explains significant variance in employee WNB. A broad set of antecedents (e.g., work/family demands and resources) relevant for the combined study on crafting and WNB offers a variety of new research questions and outcomes (e.g., work-family interference) to be studied with the WNBC scale.

In summary, the validity and potential of the WNBC scale were displayed by extensive testing in several samples across different work cultures. Our scale is the first rigorously developed scale for measuring the construct of WNBC. As such, it has great potential to advance the scientific study of crafting the vital concept of WNB, which has elicited much attention and scholarship (Rothbard et al., 2021). The need for the study of WNBC is amplified by the increasing tendencies of blurred boundaries between life domains. This trend is exacerbated by COVID-19 measures and telework (Kniffin et al., 2020), making the balance of various life roles throughout the day an essential topic for many. The WNBC scale takes blurred boundaries into account while orienting crafting towards either life domain, as reflected in the two-factor structure of this measure. We hoped that WNBC could help employees establish a sustainable, resourceful WNB.

Given the current economic developments and work regulations imposed during the COVID-19 pandemic, employees’ WNB is increasingly under pressure (Kniffin et al., 2020). Organizations can encourage WNBC to improve their employees’ quality of life in both life domains. Our results indicated that WNB crafting is relevant for the sake of employee vitality, family role/job performance, job/life satisfaction, work engagement and self-reported WNB. This feature is important because crafting as measured with the WNBC scale illustrates behaviors that may support employee health and well-being. Crafting allows employees to purposefully balance their resources and demands of work and nonwork by proactively balancing both life domains.

For implementation in organizations, the training and education of supervisors and employees involved these crafting behaviors are relevant for employee WNB. Especially with increasing job demands, a decrease in WNB has been observed, which has been counterbalanced by supervisor support and job autonomy (Haar et al., 2019), both of which are relevant for crafting. Since WNBC can be performed at the individual level, this approach is also available to individuals without organizational support. Nevertheless, given that crafting can be trained (Gordon et al., 2018), organizations must foster opportunities for WNBC and train employees to craft their WNB. A web-based intervention (application) is currently being developed for such training purposes. Results derived from studies involving the WNBC scale will inform this application.

Besides the strength of this study, several important limitations must be acknowledged.

First, the internal consistency of WNBC subscales is relatively low in several of our validation studies. However, we cover a fairly broad spectrum of crafting efforts oriented towards WNB (physical, cognitive/emotional, relational crafting) in a compact scale with relatively few items. This implies that both sub-dimensions of the WNBC scale (work/nonwork) represent three crafting behaviours. Thus, the modest reliability coefficients seem to reflect that items were chosen to represent this conceptual breadth within the WNBC construct rather than to maximize internal consistency, likewise prominently implemented elsewhere (Ryff and Keyes, 1995). Therefore, internal consistency can be expected to be low. In fact, we would argue that internal consistency is a questionable criterion for scale quality. Adding highly similar items will lead to high internal consistency. But the additional items will add little information regarding the underlying construct (Boyle, 1991) while increasing participant burden. Our scale captures a broad spectrum of crafting efforts with a compact scale with relatively few items. To provide a more robust measure of internal consistency, we reported McDonalds ω (Zinbarg et al., 2005). Given the relatively low reliability and that items map different aspects of WNWB, future research may investigate whether items of the WNBC are formative or reflective of the construct (Coltman et al., 2008).

Second, the applied cut-off values for several CFAs model-fit assessment referring to WNBC are above standard cut-off values (Byrne, 2001) but below the.95-threshold. In this regard, we refer to an ongoing debate on the so-called “golden rules” for cut-off criteria (Niemand and Mai, 2018), while future research may apply more strict factor analytical criteria. Nevertheless, we would like to recommend our scale in the presented factor structure for use in research and practice. We base this recommendation on the convincing results of our studies, particularly study 5, which shows a substantial association between our new scale captured crafting efforts for WNB and the actual WNB measured (Wayne et al., 2021).

Third, the median age differed between the Finnish and Japanese samples, which could have biased our results. Persons identifying as female were overrepresented in the Finnish sample, and persons identifying as male were overrepresented in the Japanese sample. This imbalance in gender distribution may have aggravated the differences between countries since Finland is known to have a highly feminist work culture. Therefore, age and gender were added as control variables in the analyses. These sample characteristics may have been one reason for the lack of measurement invariance across countries for the WNBC scale. We took a first step in comparing the WNBC scale across cultures, providing evidence for metric invariance [for a comparable outline in job crafting research, see Nielsen et al. (2017)].

Fourth, we studied data from five countries with diverse cultural backgrounds that may have influenced the results (Ollier-Malaterre and Foucreault, 2017). At this point, the systematic variations and differences between the studied countries were not further analyzed. Examine the cultural variation in WNBC in further detail is beyond the scope of this validation paper. Future studies may include the interplay of cultural norms with WNBC.

In general, current perspectives on crafting need to be extended by broadening the research focus to areas of life beyond work, as done in this paper.