Social resources, work demands and workplace commitment among Finnish school principals and teachers Open Access

In the context of school, changes in working life have affected teaching and school leadership demands. In addition, wide-ranging educational reforms, such as curriculum development, have set new standards for the teaching staff working in Finnish schools in recent years (Toyama et al., 2022; Nissinen et al., 2024; Räsänen et al., 2020). The increasing job demands placed on educators are a well-recognized problem in many countries such as China, Finland, France and Norway (Chan, 2002; Räsänen et al., 2020; Desrumaux et al., 2015; Skaalvik and Skaalvik, 2018; Elomaa et al., 2023). In Finland, such demands have led to increased levels of turnover intentions, and declining workplace commitment has become a concern (Räsänen et al., 2020). At the same time, studies have shown that social factors, for example school climate, are associated with workplace commitment (Collie et al., 2011). Such social resources can function as a key resource for occupational wellbeing and workplace commitment and mitigate the negative effects of increasing job demands. Studies have highlighted the importance of diverse forms of social resources (e.g. supervisory support, collegial support) in maintaining the occupational and general wellbeing of school principals'- and teachers (Beausaert et al., 2023; Edinger and Edinger, 2018). However, previous studies have typically examined teachers and school principals separately, and social resources are often treated as generic constructs. Less is known about whether associations between different aspects of social resources and work demands, and workplace commitment differ between teachers and school principals working within the same organizational and policy context.

Building on job demands-resources (JD-R) theory (Bakker and Demerouti, 2007), the aim of the study is to examine whether the theory's core assumptions hold equally for different occupational roles embedded in the same organizational and policy context, which is characterized by high professional autonomy and increasing demands. The study intends to inform school leadership, municipalities, and policymakers on how to support workplace commitment in educational work environments and secure the quality of teaching.

In Finland, the educational system is characterized by a high level of decentralization and autonomy at the municipality and school levels (OECD, 2013). Following this, teachers and school principals are highly educated professionals holding master's degree in education who have autonomy to pedagogically design and apply their own school and classroom practices (Haapaniemi et al., 2021). School principals in Finland serve as pedagogical and administrative leaders of the school (Fang et al., 2025). In addition to master's degree, principals have accomplished mandatory studies in educational leadership and administration and possess sufficient work experience as teachers to receive qualifications to serve as school principals. Despite the shared educational background, the dual leadership role of school principals may cause different sets of demands and resources, and their associations with workplace commitment compared to teachers.

The Finnish educational system is publicly funded, noncompetitive framework emphasizing equality and trust rather than accountability and standardized testing (Pyhältö et al., 2011). This context provides an interesting setting for examining differences as high levels of autonomy and social trust coexist with increased demands. In addition, the same educational and policy context, combined with a similar educational background, offers a unique setting to reveal possible nuanced differences between the two occupational groups.

The theoretical framework of this study is built on the JD-R model (Bakker and Demerouti, 2007). The core idea of the JD-R model is that every occupation contains potential risk factors (job demands) and assets (job resources) that can affect psychological, physiological, social, or organizational costs or goals (Bakker and Demerouti, 2007). According to JD-R theory, demands and resources may form the foundation of employee wellbeing, behaviors and performance, such as workplace commitment (Bakker et al., 2023). Another premise of the JD-R model is its dual processes, where job demands predict the health impairment process and job resources predict the motivational process, which predicts organizational outcomes, such as occupational wellbeing (Bakker and Demerouti, 2007).

Previous studies using the JD-R model have shown that various work demands (e.g. workload, discipline problems, classroom disturbances) and resources (e.g. self-efficacy, autonomy, collegial and administrative support) are associated with occupational wellbeing among school principals and teachers (Collie et al., 2020; Dicke et al., 2018; Skaalvik and Skaalvik, 2018). Moreover, previous studies has shown that social resources are associated with occupational wellbeing among several occupational groups (Upadyaya and Salmela-Aro, 2020; Beausaert et al., 2023; Edinger and Edinger, 2018). The current study is utilizing JD-R theory by examining associations of various aspects of social resources (collegial support, supervisory support, sense of community), as job resources, with educators' occupational wellbeing. In addition, different forms of work demands (cognitive, emotional, social) associations with occupational wellbeing, used as job demands, are examined utilizing JD-R theory. Next, these key constructs of the study are presented further.

Workplace Commitment. One way to operationalize the occupational wellbeing of teachers and school principals is organizational commitment (Collie et al., 2020; Collie and Perry, 2019). Workplace commitment refers to psychological state that characterizes employees' relationship with organization and influences their decision to stay or leave (Meyer and Allen, 1991). The concept of organizational commitment includes three different dimensions: normative, continuance, and affective commitment (Meyer and Allen, 1991). These dimensions vary in nature and quality of commitment. Normative commitment is seen to be formed from obligations, loyalty and duty (Meyer and Parfyonova, 2010). Continuance commitment, in turn, is based on employees' need to stay in the current organization and considerations of the possible costs of leaving (Karakus and Aslan, 2009). Affective commitment is defined as an employee's emotional attachment to, identification with and involvement in their work (Meredith et al., 2023).

Building on previous workplace commitment studies (Liu and Bellibas, 2018), current research also examines organizational commitment through the lens of affective commitment, which encompasses emotional aspects of commitment, pride in one's work, and a sense of belonging in one's job, and it is often considered as a key indicator of occupational wellbeing. In addition, the emotional aspects of workplace commitment emphasize engagement and healthy psychological functioning at work (Collie et al., 2020). For example, affective commitment has been found to be negatively associated with absenteeism, and positively associated with self-efficacy (Bakker et al., 2003; Chesnut and Burley, 2015).

According to previous studies, lack of commitment, workload and problems in the school system are the main reasons for teacher turnover (Räsänen et al., 2020). In addition, studies have shown that among school principals, affective commitment is positively associated with occupational wellbeing and negatively associated with turnover intentions and burnout (Houle et al., 2022). Consequently, workplace commitment is a crucial factor in retaining educational professionals and maintaining the quality of teaching, which highlights the need to explore the workplace commitment of both teachers and school principals within the same study.

Social Resources. Social job resources are understood as the social relationships, networks and community bonds available to individuals, which can support their occupational wellbeing and work-related goals. The concept of social resources is related with the broader and more theoretically developed term social capital, which is a complex concept, and researchers define the term in different ways (Ahn and Davis, 2020). Nonetheless, even though different theories and conceptualizations differ, and depend on the discipline of the researcher, they are considered complementary, each contributing to the understanding of the complex concept of social capital (Villalonga-Olives and Kawachi, 2015). Social resources refer to the structural and cognitive aspects embedded in workplace relationships and are built on the broad theoretical work on social capital.

The main theories of social capital have been presented by Bourdieu (1986), Coleman (1988) and Putnam (2000). Common to all these theories is the recognition that the concept of social capital includes the social network as a structural core component operating as a resource bank for the individual, based on cognitive factors such as trust, reciprocity, shared norms and values and used to achieve personal or collective goals. These are also the aspects we refer to here as social resources. Within the JD-R model, social relationships are typically considered social job resources, covering both structural and cognitive dimensions.

The structural perspective on social resources can be examined horizontally and vertically. Horizontal social resources occur at the same hierarchical level, for example between peers, while vertical social resources are found in relationships between individuals occupying different levels of formal power, for example an employee and supervisor (Beausaert et al., 2023). The social community at work has been considered an indicator of cognitive social resources. The social community has been seen to reflect an individual's attachment and belonging to the community, and dense networks promote the preservation and production of resources. Moreover, both a sense of community and cognitive social resources highlight the importance of participation, trust, values and norms (Lin et al., 2001; Ahn and Davis, 2020; Pooley et al., 2005).

Work Demands. One of the main assumptions of the JD-R model is that job demands vary depending on the occupation and organization (Bakker et al., 2023). In the current study, work demands were examined through emotional demands, cognitive demands and role conflicts that school principals and teachers experience at work. Role conflicts can resemble a social job demand that occurs between people or groups of people within the context of work. Role conflicts arise when expectations about interaction with different people clash (Papastylianou et al., 2009). Recently, teachers' work has become more complex, and teachers face diverse expectations caused by high-pressure work and lack of resources, which is one of the reasons behind increasing turnover intention (Räsänen et al., 2020).

In addition, the roles of school principals and teachers include demanding emotional work through interpersonal interactions with colleagues, students, parents and other stakeholders, which can further increase the emotional demands on the profession. Maslach et al. (2001) argue that emotional demands can affect an individual's capability to be involved with, and responsive to, the needs of service recipients, which, in the context of school, may lower the quality of teaching. Previous studies have found that emotional demands are negatively associated with workplace commitment (Bozionelos and Kiamou, 2008; Collie et al., 2018). Moreover, as healthy psychological functioning is often an essential part of workplace commitment, the role of cognitive demands (e.g. multiple tasks, working memory, creativity, decision making) in workplace commitment has also been examined. A constant flow of developmental tasks, school reforms and educational innovations has increased the complexity of teachers work, and teachers must tolerate more insecurity and incompleteness and a higher workload, which can cognitively burden them (Räsänen et al., 2020). Nonetheless, though cognitive demands are likely to be related to workplace commitment, there is a gap in the literature regarding these potential associations.

In addition to direct effects, the JD-R model offers a functional platform to examine interaction effects between job demands and job resources and their association with organizational outcomes. The buffering effect suggests that high resources reduce the negative effects of job demands (Bakker and Demerouti, 2007). By contrast, the boosting effect suggests that job resources play a stronger role in boosting organizational outcomes when demands are high (Bakker and Demerouti, 2007; Dicke et al., 2018). For example, social capital as a form of participatory organizational culture can reduce emotional demands (Kowalski et al., 2010), and servant leadership can buffer the positive association between workload and depressive symptoms (Upadyaya et al., 2016). However, in the context of workplace commitment, less is known about how social resources may buffer against work demands among educators. Thus, the present study focuses on examining such effects further. All the direct and interaction effects between constructs are presented in Figure 1.

1. Do the associations between social resources (social support from supervisor, social support from colleagues, social community at work) and workplace commitment differ among Finnish school principals and teachers?

2. Do the associations between work demands (emotional demands, cognitive demands, role conflicts) and workplace commitment differ between Finnish school principals and teachers?

Hypotheses: Building on JD-R theory, and its core assumption that every occupation contains its own potential psychological, physical, social or organizational assets and risks, which can affect organizational costs or goals, study hypothesizes that aspects of social resources are positively associated with workplace commitment among school principals (H1) and teachers (H2), while aspects of work demands are negatively associated with workplace commitment among school principals (H3) and teachers (H4), and there is not any differences between the two groups (Bakker and Demerouti, 2007).

• (3)

  Do the possible buffering effects of social resources on work demands regarding workplace commitment differ between Finnish school principals and teachers?

Data collection considering school principals occurred through the Finnish Principal Barometer (University of Helsinki, 2025b), which is collected annually. The data were collected between May and August 2023 through an online survey. The data for the present study were drawn from the fifth data collection round of a cohort study that began in 2019. Altogether 498 (63% men) school principals participated in the study. The age range of the participants was 30–67, and the average age of the school principals was 50 years.

The data considering teachers were part of the Finnish Teachers' Occupational Wellbeing Study (University of Helsinki, 2025a). Data has been collected every half a year, starting in 2020. In the current study, we used data from spring 2023 (May–June). A total of 1855 (80% women) teachers answered the online questionnaire. The participants' age ranged from 23 to 68, while the mean age was 50 years.

The link to the questionnaire was distributed to school principals via email, along with a description of the study and its aims. The data collection considering teachers was carried out in collaboration with the Trade Union of Education in Finland and considering school principals with the Finnish Principals Association. The link to the questionnaire was delivered to all the members of the associations in question. The members of associations considered teachers and school principals from all the levels of education (early childhood education to university level) and no distinctions or clarifications were made considering educational context. Participation in the study was voluntary, and the research project was approved by the university's Ethical Review Board.

The study utilized part of the Copenhagen Psychosocial Questionnaire (COPSOQ II). The COPSOQ is a widely used and validated self-report instrument that includes various psychosocial dimensions of the workplace (Pejtersen et al., 2010). Key figures for the instruments are presented in Table 1.

Social Resources. Both structural (horizontal and vertical) and cognitive aspects of social resources were assessed using three scales from the COPSOQ interpersonal relations and leadership dimension. The horizontal aspect of social resources was measured by social support from colleagues (e.g. “How often do you get help and support from your colleagues?”) and the vertical aspect of social resources was measured by social support from supervisor (e.g. “How often is your nearest supervisor willing to listen to your problems?”). Both scales contained three items. Cognitive social resources, social community at work, also included three items (e.g. “Do you feel part of the community at your place of work?”). All the questions considering social resources were answered using a Likert scale from 1 - Always – 5 - Never/hardly ever, and the scales were reversed before analysis.

Work demands. Work demands included emotional demands, cognitive demands, and role conflicts. Role conflicts were part of the COPSOQ domain “Interpersonal relations and leadership” and included four items (e.g. “Are contradictory demands placed on you at work?”). Questions were answered with a Likert scale from 1 - To a very large extent – 5 - To a very small extent. Cognitive and emotional demands were from the COPSOQ domain “Demands at work” (e.g. cognitive demands: “Do you have to keep your eyes on lots of things while you work?”, emotional demands: “Is your work emotionally demanding?”. The Likert-type 5-point scale for cognitive demands ranged from 1 - always – 5 - never/hardly ever. The first two items of the emotional demands scale were measured from 1 - always – 5 - never/hardly ever, while the last two were measured from 1 - to a very large extent – 5 - to a very small extent.

The original COPSOQ scale for emotional demands contained four items, but one item was excluded from the present study. Removing item 4 (“Do you get emotionally involved in your work?”) increased the scale's Cronbach's alpha from 0.67 to 0.75 for school principals, and from 0.69 to 0.76 for teachers. Further examination revealed that the reason behind this might be poor translation from English to Finnish. All the work demand scales were reversed before analysis with school principals and items for role conflicts with teachers.

Workplace commitment. Workplace commitment was measured using items from the COPSOQ domain “Work organization and job contents”. Workplace commitment included four items (e.g. “Do you feel that your place of work is of great importance to you?”). Three items were answered on Likert scale ranging from 1 - To a very large extent – 5 - To a very small extent and one item on a scale from (item 4) 1 - Always – 5 - Never/hardly ever. Items 1, 2 and 3 were reversed before analysis with school principals, and item 4 was reversed with teachers.

Demographics. We controlled age and gender as covariates to account for potential sociodemographic differences among educators. Years of experience have previously been used as a covariate in recent studies (Collie et al., 2020), but because of limitations of the data, it was not possible to include in this study. In both questionnaires, age was reported by birth year, which we transformed to age by subtracting the birth year from the year the questionnaire was conducted (2023). Gender was recoded as a binary variable (0 = female, 1 = male) for a more simplified interpretation.

Latent correlations of latent constructs representing all the variables are presented in Table 2. For a further examination of the correlations between all the variables, please see supplementary materials Table A.

The normality of the data was tested with the Shapiro–Wilk test. In the Shapiro–Wilk test, all the p-values from both datasets were <0.05, which confirms the null-hypothesis that the data are non-normally distributed. In addition, multivariate normality of the data was tested with Mardia's test. In Mardia's test the skewness and kurtosis tests for both datasets returned significant p-values, which means that the data violated multivariate normality (Mardia, 1970). Consequently, because of the nature of data collection, there was no reason to assume that the data were not missing at random. Therefore, in this study, missing data were treated as MAR. MAR allows missing data to be handled using the full information maximum likelihood (FIML) approach (Allison, 2003). To avoid issues with multicollinearity, correlations between variables were examined. According to the methodological literature on structural equation modeling, correlations higher than 0.85 for any pair of variables indicate possible issues with multicollinearity (Teo et al., 2013). When testing the correlations between variables within both datasets, none exceeded that limit.

Multigroup confirmatory factor analysis (MG-CFA) and structural equation modeling (MG-SEM) were the main analyses used in the study, and they were conducted with R version 4.4.1 (R Core Team, 2024) in R Studio (R Studio Team, 2024). The Lavaan package (Yves Rosseel, 2012) was used to define the measurement model and conducting multigroup CFA, and main effects SEM model, while the Modsem package (Slupphaug et al., 2024) was used to test the multigroup interaction SEM model. Model fits were evaluated with the robust comparative fit index (CFI), robust root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR). In addition, with CFI and RMSEA, Bayesian Information Criteria (BIC) were used as additional evaluation criteria when testing measurement and structural invariance.

For the CFI, values higher than 0.90 were considered an acceptable fit and values higher than 0.95 an excellent fit, for the RMSEA, values smaller than 0.05 were considered to reflect a good fit and values between 0.05 and 0.08 an acceptable fit, and for the SRMR values smaller than 0.05 were considered as good fit and values smaller than 0.10 an acceptable fit (Teo et al., 2013). Models were estimated using the maximum likelihood robust (MLR) method to gain standard errors adjusted for the non-normality of the data.

Measurement invariance strategy included three steps covering configural, metric and scalar invariance. Configural invariance was examined by allowing all parameters to be freely estimated between groups, metric model by constraining factor loadings as equal, and scalar invariance by constraining factor loadings and item intercepts as equal between groups (Xu and Tracey, 2017). After establishing an acceptable level of measurement invariance (metric), structural invariance was tested in multigroup SEM by constraining also path regressions. To compare differences in structural regressions in multigroup SEM, achieving metric invariance in measurement invariance is needed (Davidov et al., 2014). During invariance testing, not only was the overall model fit examined, but changes in selected fit indices. To evaluate the level of invariance thresholds of CFI ≥0.01, RMSEA ≥0.015, SRMR ≤0.03, and the lowest value of BIC showed the best fit (Cheung and Rensvold, 2002). If the constrained model fits similarly (differences between the thresholds) or better, the level of invariance can be accepted.

First multigroup confirmatory factor analysis was conducted. Configural model showed acceptable fit. After that, metric model was determined and metric model showed an acceptable fit. Model fits of each model are presented in Table 3.

After metric measurement invariance was achieved, multigroup structural equation modeling was conducted to examine if associations between social resources, work demands, and workplace commitment differ between school principals and teachers. In the first model, only main effects were included. Both baseline model and constrained models showed acceptable fit. Fit of the constrained model was better, indicating that there are no differences between groups in regressions. From the social resources, social support from supervisors and social community at work were significantly associated with workplace commitment, and from the work demands, only role conflicts were significantly associated with workplace commitment. From demographics, women and older educators showed a significant but weak association with higher workplace commitment. Standardized coefficients of the main effects model are visualized in Figure 2, and model fit indices are presented in Table 4.

Finally, interaction model was defined. Only interactions between significant main effects were added in the model. In the study, interaction effects were exploratory in nature, and when testing interaction model, non-significant main effects were not added because of model complexity, stability, and reliability (Little et al., 2006). Crossover interaction effects are possible when testing SEM models, but they were not expected within the context of the present study. To determine interaction effects, residual centering approach (RCA) was selected because it provides conceptually straightforward and computationally stable solutions.

Testing invariance of structural regressions with interaction models, baseline and constrained models showed acceptable to excellent fit. Furthermore, constrained model showed a better fit, pointing out that associations of social resources, work demands and workplace commitment do not differ between school principals and teachers. Model fit indices are presented in Table 5. The interaction model showed that associations of social support from supervisors, social community at work, role conflicts, and workplace commitment remained significant. In addition, associations between demographics and workplace commitment remained weak but significant. None of the interaction effects were significant. Standardized coefficients of interaction model are presented in Figure 3.

The purpose of the study was to examine whether the associations of social resources and work demands with workplace commitment differ among Finnish school principals and teachers. Further, we examined differences in associations between the interaction effects of social resources and work demands in workplace commitment. Building on JD-R theory (Bakker and Demerouti, 2007), we hypothesized that aspects of social resources are associated positively with workplace commitment, while aspects of work demands are related negatively, regardless of the role of educator. The interaction effects were exploratory in nature, and hypotheses were not set.

The results of the study partially supported our hypotheses. The study shows that there were no differences in the structural regression paths between Finnish school principals and teachers. From the social resources, social community at work and supervisory support were positively associated with workplace commitment among educators (H1&2). Role conflicts were negatively associated with workplace commitment among educators (H3&4). However, cognitive and emotional demands, and collegial support were not associated with workplace commitment. In addition, none of the interaction effects were significant.

The results showed that the associations between workplace social resources and workplace commitment were similar between Finnish teachers and school principals. Social support from supervisors and social community at work was associated with workplace commitment among both school principals and teachers. These results highlight the significance of the quality of leadership and dense networks in the work community in schools.

Previous studies have also highlighted the importance of supervisory support and sense of community among teachers and school principals (Collie et al., 2018; Beausaert et al., 2023). In Finland, the hierarchical structure of schools has traditionally positioned principals as school leaders. Nonetheless, school principals are also teachers by education, and it is natural that they view teachers as their close social community at work. Previous studies have noted that the fragmentation caused by a high workload limits school principals' ability to focus on strategic and pedagogical leadership, which may lead to a lack of resources and hinder professional development (Lantela et al., 2024).

In Finland, the teaching profession is highly appreciated, and teachers enjoy strong autonomy and decision-making power over their work (Salmela-Aro et al., 2019). School principals are crucial facilitators of teachers' social capital, and as school leaders, they form the key aspects of school organizational culture, such as school mission, vision, values and norms (Minckler, 2014). As the social community at work represents trust, participation, safety and the density of the social network, a well-functioning organizational culture is an essential resource for enhancing teachers' workplace commitment. Therefore, the fragmentation of school principals' job may not only lead to school principals decreasing resources but also distancing them from teachers' day-to-day work and affect the perceived social climate.

In the present study, collegial support was not associated with workplace commitment, which contradicts the findings of earlier research (Kaihoi et al., 2022; Beausaert et al., 2023). One possible explanation among teachers is that teaching in Finland, due to the high level of autonomy, is often seen as an individualistic profession, and it is possible that teachers rely on colleagues more for practical aspects of the profession (e.g. sharing materials, co-teaching), and these interactions do not translate into a stronger emotional attachment to the workplace. Among school principals, this finding can be explained by the complexity of the work and hierarchical structure, which may cause a lonely leadership role as in Finland, not all the school principals have colleagues within the same school.

The second aim of the present study was to examine whether the associations between work demands and workplace commitment differ between Finnish school principals and teachers. Again, results showed that associations were similar between the groups. Of the aspects of work demands, role conflicts were negatively associated with workplace commitment among both school principals and teachers, while emotional demands and cognitive demands were not associated with workplace commitment.

There is a gap in the research on the effect of cognitive demands on workplace commitment among educators. The study's results on emotional demands contradict previous research among educators (Skaalvik and Skaalvik, 2011). Teaching and school leadership are professions where emotional demands are commonplace, and thus, they may not directly affect workplace commitment. For example, in previous studies, teachers' intention to remain in the profession has been mostly explained by individual factors (Casely-Hayford et al., 2022).

Role conflicts reflect the diverse expectations educators increasingly face at work and within the school community. Decreasing workplace commitment is one of the main reasons behind teachers' increasing turnover intentions, and especially younger teachers' turnover intentions have been linked to the increasing number of tasks, problems of interaction and a perceived mismatch between expectations and reality (Räsänen et al., 2020). Moreover, school principals operate with high levels of autonomy, and there is no standardized job description, which renders the work more fragmented and complex (Salo and Saarukka, 2023), which may cause an increasing number of role conflicts. The results of this study support earlier findings and highlight the importance of effective interaction and clear, well-defined job descriptions.

Next, the interaction effects between social resources and work demands were examined. As the associations were similar between work demands, social resources and workplace commitment, the same interaction terms were set with both. None of the interaction effects was associated with workplace commitment among educators. As role conflicts may violate personal values (Papastylianou et al., 2009), it is also possible that role conflicts are connected to more fundamental issues concerning a person's job (e.g. values), and no aspect of workplace social resources affects teachers' personal opinions. Moreover, as educators enjoy considerable autonomy in their work, they may manage role conflicts independently rather than relying on external support.

Earlier studies have suggested that future research should clarify the concept of social capital (Beausaert et al., 2023). In addition to social capital, the concept of social resources also needs more clarified theorizing. Thus, there is space for research on domain-specific definitions and instruments measuring social resources in educational settings. Educational settings are often more specific, and, for example the Finnish school system differs from that of many countries. Moreover, there is a need for more research to identify the most relevant and demanding interactions that educators face, their antecedents, and consequences.

To better understand social resources in educational settings, both qualitative and quantitative research is required. This also occurs with work demands, such as role conflicts. Qualitative research settings or different types of measurement reports (e.g. observations, network analysis) may deepen our knowledge of the field and help us better understand the increasing demands faced by educators, and the way social resources are formed within the schools. This study was conducted in a cross-sectional setting, and a longitudinal approach is necessary to examine the evolution of levels of social resources through time. In addition, this research focused on the differences in associations between social resources, work demands, and workplace commitment, but in the future, more descriptive research could be important to compare the two groups of educators. For example, the context of school and school-level factors (e.g. size of the school) were excluded from this research, while these aspects may have an effect on social structures of the school. These factors may reveal completely new views and nuances about educators' occupational wellbeing.

For the instruments used in the school principals' study, Cronbach's alpha and MacDonald's omega were relatively low for cognitive demands, suggesting that the scale's reliability may be suboptimal.

In the context of social capital, social resources are seen as shared and joint features of a social community, which can also be unevenly distributed within the community (Bourdieu, 1986; Coleman, 1988; Putnam, 2000). However, this study points out that in the Finnish educational system, these features are relatively equally distributed despite the hierarchical structure. The functioning school community appears to be a more important factor than the role of educator. The findings of this study suggest that interventions considering cohesion of the workplace could be targeted similarly for both school principals and teachers in Finland. Trust, collaboration and social resources develop in the same way for both and should be considered as collective resources. This highlights the importance of well-designed structures of mentoring and supporting educators' jobs as well as dialogue between teachers, school principals and management.

This study examined how associations of social resources, work demands and workplace commitment differ between Finnish teachers and school principals. The findings highlight that between teachers and school principals, there are no differences in these associations and the professional community works similarly for both teachers and school principals in Finland. Moreover, they demonstrate that role conflicts are negatively associated, while supervisory support and social community at work are positively associated with workplace commitment among educators.

The resilience of the school as an organization can also be examined through the lens of social resources, as schools that foster strong support networks are likely to be better positioned to adjust to challenges and crises. In addition, a well-functioning organizational culture that highlights the importance of effective social networks can also connect social capital to school-wide effectiveness. A deeper knowledge of social networks enhances our understanding of human resources and places the development of educator's workplace commitment on more sustainable ground.

The author wishes to thank the Finnish teachers and school principals who participated in the study. Sincere thanks also to the Finnish Principals’ Association (Suomen Rehtoritry) and the Trade Union of Education in Finland (OAJ) for their support in facilitating the data collection.

The supplementary material for this article can be found online.