Understanding novice school leaders’ turnover intentions: the role of leadership dynamics Open Access

School leader turnover is closely linked to turnover of teachers (Sun and Wang, 2017) and school leader colleagues (Bartanen et al., 2021), negative impacts on student outcomes (Bartanen et al., 2019) and inhibited school development (Grissom et al., 2021). With turnover rates as high as 20–30% annually (Thelin, 2020) and a median tenure of just three years (OECD, 2019), Swedish schools face a significant challenge of an overrepresentation of novice leaders, who thus make up a critical and substantial part of the education system. Novice school leaders are often placed in challenging schools due to the departure of more experienced colleagues (Béteille et al., 2012; Gates et al., 2006; Loeb et al., 2010), and early challenges in these positions can lead to premature turnover (Spillane and Lee, 2014), although some studies have found lower turnover rates among novice principals than their more experienced counterparts (Grissom and Bartanen, 2019). Despite their crucial role, it remains unclear how their leadership experiences influence their turnover intentions.

According to Hughes et al. (2019), leadership can be understood through the three domains of leader characteristics, interpersonal dynamics, and situational context. The leader characteristics domain includes individual contributions like work experience, emotional regulation, self-efficacy, age, and gender. Interpersonal dynamics focus on the shared social influence process among workplace members, emphasizing the leader follower relationship. The situational context highlights the broader environment, where factors such as organizational roles, structure and resources significantly impact leadership behaviors and effectiveness.

Much of the research on school leader turnover and turnover intentions has traditionally focused on the dimension of situational context (Snodgrass Rangel, 2018), manifested as working environment (Richard, 2024; Yan, 2020), school actors (Ni et al., 2015; Sun and Ni, 2016), and school-level factors (Béteille et al., 2012; Burkhauser et al., 2012; Fuller et al., 2015). Studies that have included leader characteristics and workplace interpersonal dynamics have enhanced our understanding of turnover (Boyce and Bowers, 2016; Håkansson et al., 2021). However, when investigating school leader turnover, the leadership experience, as described by Hughes et al. (2019), of novice school leaders has not particularly been taken into consideration. In addition, as the three leadership domains may be beneficial for understanding school leader turnover, the roles of principals and assistant principals may differ in ways that might influence how their turnover intentions can be explained (Bartanen et al., 2021). Principals typically carry the primary responsibility for the overall management and strategic direction of the school, while assistant principals often focus on specific operational aspects from a not so visible position (Bush, 2024). These differences in roles may potentially affect their turnover intentions differently. Examining these distinctions is crucial for a more nuanced understanding of leadership turnover in schools.

Even though we know that the decision to leave a job is influenced by a complex interplay of factors over time (Lee and Mowday, 1987), most of the existing research about school leader turnover has focused on turnover that has already occurred (Snodgrass Rangel, 2018; Yan, 2023). Thus neglecting the precursors to these decisions. Exploring the factors contributing to school leaders’ turnover intentions offers crucial insights for policymakers aiming to enhance leadership stability and succession (Sun and Ni, 2016; Tran, 2017).

To address the research gap, the aim of this study is therefore to investigate how leader characteristics, workplace interpersonal dynamics and situational context of novice school leaders may predict novice school leaders’ turnover intentions. The study also explores how these potential associations might differ between principals and assistant principals.

School leader characteristics may significantly shape workplace experience and turnover intention (Hughes et al., 2019). Experience has been found to have a positive relationship with turnover stability among school leaders (Baker et al., 2010). However, school leaders with both minimal and extensive experience have been shown to share similar motivations for staying, unlike those with moderate experience (Gates et al., 2006). Emotional stability, defined as inversely related to neuroticism, is a valued quality for school leaders. It enables them to manage and regulate the emotions and needs of various stakeholders as well as to themselves (Derue et al., 2011). Leaders with higher emotional stability are more likely to cope effectively with the stressors and challenges associated with their roles, resulting in increased job satisfaction and commitment to their schools (Boyland, 2011). The importance of emotional stability is supported by research findings that link successful leaders with high extraversion and low neuroticism (Do and Minbashian, 2020), while low flexibility and low emotional stability have been strongly linked to leadership derailment (Hogan et al., 2011). Similarly, self-efficacy plays a pivotal role in shaping an individual’s workplace experience, including their job satisfaction and emotional well-being (Bandura et al., 1999). In the context of school leadership, the relationship between self-efficacy and turnover intention is particularly noteworthy. School leaders with a strong sense of self-efficacy are more likely to exhibit resilience in the face of challenges and demonstrate commitment to their roles. Conversely, those with lower self-efficacy may feel overwhelmed by the demands of their positions, ultimately resulting in a higher turnover intention (Skaalvik, 2020).

School leader characteristics also include factors like age and gender, which demonstrate complex associations with turnover. Older leaders are generally more likely to leave their roles (Sun and Ni, 2016), but both younger and older leaders than middle-aged ones, also show a higher propensity to exit (Ni et al., 2015; Tekleselassie and Villarreal, 2011). Age further moderates the gender-turnover connection, with young female leaders having similar turnover rates to young men, while older men are more prone to leave than older women (Gates et al., 2006). Gender studies on turnover present mixed results, with some suggesting men are less likely to leave (Ni et al., 2015), while others find that women are less likely to leave (Sun and Ni, 2016; Tekleselassie and Villarreal, 2011).

School leaders’ interpersonal dynamics, particularly their relationships with co-workers, may significantly impact job satisfaction and turnover intentions (Prado Tuma and Spillane, 2019). Managing team frustrations and emotions can increase stress, leading to higher turnover (Arvidsson et al., 2021; Håkansson et al., 2021). The number of co-workers also plays a critical role, as smaller teams make it easier to build trust and manage relationships, potentially reducing turnover (Leithwood et al., 2008). However, as team size increases, the challenges of maintaining personal connections and trust grow, often resulting in a reliance on indirect communication, which can hinder unity (Spillane, 2012). Trust between leaders and co-workers is crucial (Tschannen-Moran and Hoy, 2000). In a professional setting, a principal may feel the need to act in ways that foster trust, while in an environment filled with conflict, they might resort to behavior that promotes distrust in an effort to gain control (Tschannen-Moran and Hoy, 1998).

The situational context addresses the conditions within the organization that may affect leadership practices and outcomes. In this context, the turnover intention of school leaders might be influenced by several factors. Swedish compulsory school leaders’ turnover rates seem to differ between those in private versus public schools (National Agency for Education, 2024). Thus, private and public schools may have different organizational structures and support systems for their school leaders, which could contribute to varying work environments and conditions (Ni et al., 2015; Sun and Ni, 2016). The school leader’s opportunity to exercise leadership might be altered by organizational aspects such as resource allocation, performance expectation and accountability, autonomy and decision-making power (Gawlik, 2008; Sun and Ni, 2016). The availability of resources, along with the organizational role demands, may shape the school leader’s experience and ability to effectively carry out their responsibilities (Wells et al., 2010). Additionally, school levels – preschool, compulsory school, high school and adult education – may present different organizational challenges, possibly affecting turnover rates (Persson et al., 2021a, b).

The Swedish National Principal Training Program plays a crucial role in preparing novice school leaders with essential skills and knowledge. This mandatory three-year in-service program, required within a year of assuming a principal role and voluntary for assistant principals, covers key areas such as legislation, administration and school leadership (Norberg, 2019). With approximately 3,500 participants, the program’s comprehensive approach helps novice leaders navigate the complexities of their roles, potentially improving retention (Jacob et al., 2015).

As pedagogical leaders, Swedish school leaders must thoroughly understand the teaching–learning environment to influence teachers’ practices and initiate educational changes (Merchant et al., 2012). This role has become increasingly complex due to enhanced state oversight, legislative changes, rigorous inspections and a focus on systematic quality assurance. Leaders must balance policy directives with improving educational outcomes (Ärlestig and Johansson, 2020), reflecting broader challenges in educational leadership globally (Hoy and Miskel, 2012). In the public sector, school leaders’ roles increasingly resemble those of first-line managers, involving strategic planning, budget management and personnel oversight (Björk et al., 2013). New public management has expanded these responsibilities, requiring leaders to navigate a dynamic work environment and maintain transparency in reporting (Wällstedt and Almqvist, 2017), which may impact turnover rates (Cregård and Corin, 2019).

This study targeted novice school leaders who were enrolled in the principal training program in Sweden during the fall of 2021, drawing on participants from cohorts enrolled in the years 2019, 2020 and 2021. Approved by the Regional Ethical Review Board (DNR, 2021/02,710), the survey reached 3,096 eligible participants after initial contact information filtering from the National Agency for Education. Out of these, 2,008 completed the survey, with 1,945 responses (63% of the target population) included in the final analysis. This sample covers a significant portion of Sweden’s novice school leaders and includes a notable share of the 5,600 principals and 6,800 assistant principals across all educational levels in the country.

Table 1 gives an overview of the participants structured by principal and assistant principal as well as female and male. The sample was predominantly female, reflecting the gender distribution of school leaders in Sweden, with most participants working as principals. The majority were employed in public schools, with female principals mainly in preschools and male principals in compulsory schools. Assistant principals, both male and female, were mostly in compulsory schools. A small proportion worked without co-workers, while most had substantial co-worker teams, which included both teachers and administrative staff. School locations were fairly balanced among urban, close to urban and rural settings. Most participants, both principals and assistant principals, held teacher or preschool teacher qualifications. Participants’ experience in the principal training program was evenly spread across the three years.

Six standardized instruments were employed to assess the study’s central constructs, as outlined in Table 2. These instruments were translated and back-translated between English and Swedish, adhering to recommended practices (Wang et al., 2006). All constructs demonstrated acceptable reliability (Tavakol and Dennick, 2011). Participants rated their responses on a seven-point Likert scale, which was later converted to a 0–10 scale for improved statistical clarity.

Turnover intention was measured using the Turnover Intention Scale (Bothma and Roodt, 2013), a unidimensional scale consisting of six items.

Leader characteristics. Gender, age and years of experience were collected as background information. Health was assessed with a single item asking participants to rate their overall health. Emotional instability was measured using the neuroticism subscale of the Mini-IPIP (Donnellan et al., 2006). Self-efficacy was measured with the Norwegian Principal Self-Efficacy Scale for instructional leadership (Skaalvik, 2020).

Interpersonal dynamics. Trust was measured with a subdimension of the Principal Trust Scale (Tschannen-Moran and Gareis, 2004). Emotional containing, group dynamics and mentoring co-workers were evaluated using subscales from the Gothenburg Manager Stress Inventory (GMSI, Håkansson et al., 2021). Emotional containing captures the extent to which leaders must manage their co-workers’ frustrations, while group dynamics assesses the perceived cohesion among staff. Co-workers refers to all school employees, including teachers and administrative staff. Mentoring co-workers refers to the energy spent on structuring and guiding their work. To assess the number of co-workers that the school leader at the time was responsible for, they were asked to fill in the number in clusters of 0, 1–10, 11–20, 21–30 and 31 and up. The categories were then rearranged to construe only three categories (0, 1–20, 21 and up), transformed into dummy variables where 0 was set as the reference value.

Situational context. School levels were transformed into dummy variables with preschool as the reference category. Organizer affiliation was similarly transformed, with private schools serving as the reference. Salary satisfaction and support from superiors were assessed using subscales from the Copenhagen Psychosocial Questionnaire (COPSOQ, Kristensen et al., 2005). Organizational role demands were measured with a subscale from the GMSI, focusing on the pressures placed on school leaders by their organizational roles. Resource deficit, also measured with the GMSI, assessed the extent to which leaders felt their schools were under-resourced for managing responsibilities effectively.

The data were analyzed using SPSS version 28. Descriptive statistics, including measures of central tendency and distribution, were calculated to provide an overview of the dataset and understand the interrelationships among the variables. ANOVAs were performed to examine differences in the study variables between principals and assistant principals. Next, a linear regression analysis was conducted, employing different models to explore the relationships between independent and dependent variables. This approach systematically introduced variables to determine their unique contributions, providing robust insights into the factors influencing turnover intention.

Table 3 provides a statistical overview of the key variables of the study, including means and standard deviations for each variable. In addition, the F statistic from a one-way ANOVA compares the results between principals and assistant principals for each of the variables. The results show that principals rated their health lower (p < 0.001) and perceived less support from their superior (p < 0.001), and at the same time not only had higher organizational role demands (p < 0.05) but also higher self-efficacy (p < 0.01).

Table 4 provides an overview of the turnover intention means and standard deviations across various categorical independent variables. Overall, the turnover intentions of principals and assistant principals appear similar and may not be attributed to any variables in the model, except for the compulsory school level, where principals experience significantly higher turnover intentions than assistant principals (p < 0.05).

The linear regression analysis, presented in Table 5, offers insights into the relationship between turnover intentions and the independent variables for the total sample but also principals and assistant principals separately. The results reveal that the explained variance in turnover intention increases as more domains are included in the models, culminating in Model 5, which incorporates all predictor variables and accounts for around 45% of the explained variance. Model 1, where only leader variables were included, could explain more of the variance than the variables included in interpersonal dynamics (Model 2) could. However, situational context (Model 4) was the group of variables that, by itself, could explain most of the variance.

Gender was a significant predictor for turnover intention among principals but less so for assistant principals, suggesting gender-related turnover factors may be more pronounced among principals, as men are more likely to leave. Trust in co-workers was a significant predictor for principals but not for assistant principals (except in model 1), highlighting the potential difference in how work relationships impact turnover intentions between these roles. Organizational role demands were a stronger predictor for principals, while support from superiors was consistently significant for both groups but showed a slightly stronger effect for assistant principals. Age showed minimal effect across models, whereas experience showed significant but weak relations for the turnover intentions of principals, while there was no significant prediction for assistant principals. Mentoring co-workers could not significantly predict turnover intention for either group. However, the experienced dynamics among co-workers was a strong predictor among assistant principals but reduced when the variables of leader characteristics and situational context were added in model 5.

One finding is the dynamic nature of the relationship between certain variables and turnover intention as new predictors were added. For instance, health, in the leader domain and emotional containing, in the domain of interpersonal dynamics, emerged as significant predictors. However, the magnitude of these relationships shifted when additional variables from other domains were introduced, suggesting that the intricate connections between factors in different domains influenced the turnover intention of both principals and assistant principals.

Another discovery is the fluctuating influence of self-efficacy and emotional instability across the models. Both variables were strong predictors in the leader domain. However, their association decreased when other variables were included, as seen when comparing model 1 and model 5. This pattern suggests that interpersonal dynamics and situational context factors played a role in confounding these relationships. In addition, the number of co-workers had no significant association to turnover intention.

Being a school leader in high school was significantly associated with increased turnover intention compared to being a school leader in preschool; however, when analyzed, each group per se showed no significant results. Affiliation to public or private school organizers did not make any significant difference for turnover intention either. Salary and resource deficit were significantly related to turnover intention but had low predictive value. Overall, the variables included in the models affected each other substantially, emphasizing the need to consider multiple domains when examining turnover intention.

The aim of this study was to investigate how leader characteristics, workplace interpersonal dynamics and situational context may predict novice school leaders’ turnover intentions. In addition, the study explored how these potential associations vary between principals and assistant principals. Health, self-efficacy, emotional instability and emotional containing emerged as strong predictors. Gender and trust were related to turnover intentions primarily among principals. Organizational role demands and support from superiors also significantly influenced turnover intentions. Overall, differences between principals and assistant principals were minor, with both groups rather similarly affected by these factors.

The results of this study cast new light on school leaders’ gender, age, experience, emotional stability and self-efficacy as predictors of intentions to remain or leave the school leader position. While previous research has shown mixed relations between age and turnover (Ni et al., 2015; Sun and Ni, 2016; Tekleselassie and Villarreal, 2011), this study found almost no relationship between age and turnover intention. This finding suggests that other factors may play a more decisive role in the turnover intentions of novice school leaders. Almost the same pattern could be seen when investigating experience. Although prior research indicates that experience generally correlates with turnover stability and better equips leaders to navigate their roles (Baker et al., 2010), this study suggests that the protective effect of experience may be diminished in contexts where novice school leaders are predominant. In contrast, gender emerged as a more pronounced factor. Despite the mixed findings in earlier studies (Ni et al., 2015; Sun and Ni, 2016; Tekleselassie and Villarreal, 2011), this study reveals a clearer trend; male principals expressed higher turnover intentions than women principals, while this relationship could not be established statistically significant among the assistant principals.

Regarding emotional instability, the results extend the existing literature by directly correlating emotional instability with increased turnover intentions, echoing the positive link between emotional stability and job commitment highlighted by Boyland (2011). The study reaffirmed the significant impact of health and self-efficacy on leaders’ turnover intentions, aligning with existing research on the role of psychological factors in occupational commitment (Bandura, 1997; Skaalvik, 2020). Other studies have shown that stress and health issues, including poor sleep, can erode principals’ self-efficacy, influencing their decision to stay in their roles (Su-Keene et al., 2024). The findings from this study suggest that while self-efficacy plays a crucial role in helping leaders face challenges and remain committed, its impact on turnover intentions may be intertwined with other factors, such as situational context and interpersonal dynamics. This indicates that the link between self-efficacy and turnover intention is complex and likely influenced by multiple factors.

The results within the domain of interpersonal dynamics indicate that trust within the team is inversely related to turnover intentions, aligning with the research that underscores the importance of trust in leadership stability (Tschannen-Moran and Hoy, 2000). This relationship underscores the importance of relational aspects in understanding turnover intention and adds complexity to models like the job demands-resources model, which considers only job-specific factors (Bakker and Demerouti, 2007). Conversely, the role of emotional containing and the management of group dynamics emerged as factors positively correlated with turnover intentions. These findings are congruent with theories emphasizing the psychological toll of emotional labor (Hochschild et al., 1983; Maslach et al., 2001), suggesting that the interpersonal dynamics within an organization can be a source of stress, thereby exacerbating turnover intention. In the Swedish context, where a significant proportion of school leaders exhibit symptoms of emotional exhaustion and burnout (Persson et al., 2021a, b), these findings gain even more relevance, offering new insights into the specific interpersonal challenges that contribute to turnover. Furthermore, the results of this study shed light on the influence of team size on turnover intentions. Contrary to Leithwood et al. (2008), who suggested the management of larger teams inherently increases turnover intention, the data of this study did not show a significant association between the number of co-workers and turnover intention. This might suggest that while the complexity of managing larger teams is acknowledged, other mediating factors, such as the group dynamics and trust, may be more influential in a leader’s decision to stay or leave.

The domain of situational context exhibited the most significant explanatory power for turnover intention, in alignment with previous research (Grissom et al., 2021; Papa, 2007). Specifically, role demands and support from superiors emerged as critical factors, thereby emphasizing the importance of organizational elements in shaping turnover behavior. More specifically, the results support the findings of Håkansson et al. (2021), highlighting the stress and reduced workability that result from such role demands, potentially leading to lower job satisfaction and increased turnover.

In combination with the other independent variables of the regression model, expected differences in turnover intentions between leaders in private versus public schools (National Agency for Education, 2024) were not observed. This challenges the notion of organizational type as a determinant of turnover and aligns with the findings of Ni et al. (2015) and Sun and Ni (2016), suggesting that the impact of structural organizational differences might be explained when dissecting the school organizer into a number of variables that theoretically construe the concept of organization (Richard, 2024). While school leaders affiliated with high schools showed elevated turnover intentions in the overall analysis, no significant effects were found when examining principals and assistant principals separately. This suggests that while school level might be associated with turnover intentions, further investigation is needed to clarify this relationship (Persson et al., 2021a, b). Further, while Gawlik (2008) and Sun and Ni (2016) argue that leadership opportunities are shaped by organizational characteristics such as autonomy and decision-making power, this study suggests that the influence of these factors on turnover intention is perhaps mediated by the level of support from superiors and role demands, a link not explicitly made in prior research. This insight builds upon Wells et al.’s (2010) recognition of the importance of support from the superintendent together with demands of the organizational role, which play a substantial role in a leader’s experience and their subsequent intentions to stay or leave.

The study reveals that while the turnover intentions of principals and assistant principals are influenced by similar factors, with health, self-efficacy and emotional instability impacting them similarly, there are nuances in their experiences that reflect their distinct roles (Bush, 2024). For principals, trust in co-workers and organizational demands are more pronounced predictors of turnover intentions. This suggests that the strategic, high-level responsibilities associated with the principal role make trust and managing organizational pressures critical areas for their retention. On the other hand, assistant principals exhibited a stronger relationship between support from superiors and turnover intentions, highlighting the importance of hierarchical support in their more operational, less visible roles. This aligns with Bartanen et al. (2021), who emphasized the significant mobility of assistant principals and their connection to principal turnover, indicating that support structures are crucial for their retention. These results emphasize the need to consider both the shared and unique aspects of principals’ and assistant principals’ roles in developing effective retention strategies, which could enhance stability in school leadership.

This study is not without limitations. The cross-sectional design restricts the ability to draw causal inferences, as it captures data at a single point in time, limiting the understanding of cause-and-effect relationships. To gain more robust insights, future research could benefit from longitudinal studies that track novice school leaders over time, offering a clearer picture of how leader characteristics and situational contexts influence turnover intentions (Richard, 2024). Additionally, incorporating mixed-method approaches that include qualitative data would enrich the analysis (Su-Keene et al., 2024), providing a deeper understanding of school leaders’ perceptions and the reasons behind their career decisions. Such qualitative insights would complement the quantitative findings and help in exploring how different contexts and interventions affect the turnover of novice leaders.

The findings of this research highlight the complex nature of turnover intentions among novice school leaders. Given the significant impact of school leader turnover on educational outcomes and school stability (Bartanen et al., 2019; Grissom et al., 2021), this study presents several key implications for policy and practice. Policymakers should prioritize developing training programs that specifically address the challenges faced by novice school leaders. Emphasizing frameworks for mentorship and ongoing professional development could serve as effective interventions to help novice leaders navigate common challenges and reduce turnover risks (Jacob et al., 2015; Richard, 2024). Furthermore, practitioners should focus on creating supportive frameworks tailored to the distinct needs of both principals and assistant principals. For principals, initiatives that foster trust within their teams and strategies to manage organizational demands could be crucial in mitigating turnover intentions. Assistant principals, on the other hand, may benefit more from robust support systems provided by their superiors. Addressing these distinct needs can lead to more effective retention strategies that accommodate the unique challenges of different school leadership roles.