Two Faces of Leadership: Transformational and Abusive Pathways to Job Satisfaction,Affective Organizational Commitment,and Turnover Intentions

Leadership Styles and Turnover Intentions

Transformational leadership is commonly defined as a positive form of leader-employee interaction that drives constructive change at individual and organizational levels (Hermanto et al., 2024). By fostering self-motivation, transformational leaders encourage employees to enhance their skills, reframe their beliefs, and exceed expectations (Chua & Ayoko, 2021; Yuwono et al., 2023). This style is associated with organizational citizenship behaviors, stronger work engagement, and a culture of empowerment that promotes trust, involvement, and productivity (Ahmad et al., 2023; Jacobsen et al., 2021; Scholl & Schermuly, 2020). Transformational leaders act as role models by articulating and sharing a long-term organizational vision, supporting employee development, granting decision-making authority, and strengthening morale and loyalty (Bastari et al., 2020; Jensen et al., 2019). Bass and Avolio (1994) conceptualized transformational leadership as comprising four core components: idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration. Idealized influence refers to the leader’s behaviors, such as acting as a role model, that engender followers’ respect, trust, and admiration. Inspirational motivation reflects the leader’s ability to clearly articulate expectations, goals, and a compelling vision that inspires and energizes followers. Intellectual stimulation involves encouraging followers’ creativity and critical thinking when addressing problems or developing new ideas and approaches to work. Finally, individualized consideration encompasses the leader’s attentiveness to followers’ professional needs, supporting their personal development and helping them realize their full potential (Bass & Avolio, 1994).

Unlike transformational leadership, abusive leadership is a destructive form of leader behavior characterized by persistent hostile verbal and nonverbal actions that undermine subordinates’ dignity, psychological well-being, and professional functioning (Labrague et al., 2020; Mrayyan, 2025; Tepper et al., 2017; Webster et al., 2016). This leadership style has gained increasing scholarly attention due to growing concern about its prevalence and detrimental consequences across organizational contexts (Jantjies & Botha, 2024; Morris, 2019). Abusive leadership encompasses behaviors such as bullying, intimidation, manipulation, and unfair treatment, which are often sustained over time and may be reinforced by organizational environments that tolerate or normalize such conduct (Bakkal et al., 2019; Marchant-Pérez et al., 2024; Padilla et al., 2007). Consistent evidence indicates that abusive leadership is associated with adverse attitudinal, behavioral, and well-being outcomes for employees (Abdallah & Mostafa, 2021; Saleem et al., 2021).

Transformational and abusive leadership thus represent two opposing poles of leader behavior, reflecting fundamentally different relational dynamics between leaders and followers (Diko & Saxena, 2023; Jantjies & Botha, 2024). Given these contrasting dynamics, scholars have increasingly examined how such leadership styles are associated with employees’ work attitudes and withdrawal-related outcomes, particularly turnover intentions, which remain a critical concern for both theory and practice in organizational behavior and human resource management (Jantjies & Botha, 2024).

Transformational leadership has been associated with reduced turnover intentions, yet the mechanisms underlying this relationship remain underexplored. For instance, Saeed and Jun (2022) demonstrated a robust, negative association between transformational leadership and turnover intentions in SMEs, mediated by affective organizational commitment and job embeddedness. Additionally, Diko and Saxena (2023) found that the relationship between transformational leadership and employees’ intention to leave is partially mediated by employee engagement. In contrast, abusive leadership has been linked to higher turnover intentions, yet empirical work quantifying this association remains more limited (Jantjies & Botha, 2024). Nunes and Palma-Moreira (2024) revealed a significant positive association between abusive leadership and turnover intentions in Portuguese and Angolan organizations, with burnout (particularly disengagement) as a critical mediator. Additionally, Lopes et al. (2025) reported that abusive leadership is positively and significantly associated with turnover intentions.

Building on the previous literature, the following hypotheses were formulated.

Hypothesis 1

Transformational leadership is negatively associated with turnover intentions.

Hypothesis 2

Abusive leadership is positively associated with turnover intentions.

The Mediating Role of Job Satisfaction and Affective Organizational Commitment

Leadership styles directly and indirectly affect turnover intentions through key attitudinal variables, such as job satisfaction and affective organizational commitment. Job satisfaction reflects employees’ evaluative judgements of their work experiences (Weiss, 2002). Affective organizational commitment captures individuals’ emotional attachment and identification with the organization (Meyer & Allen, 1991). Both constructs are robustly associated with lower turnover intentions (Falatah & Conway, 2019; Meyer & Allen, 1997; Meyer et al., 2002; Pathardikar et al., 2025), highlighting their potential as mediating mechanisms in the leadership-turnover relationship.

Positioning job satisfaction and affective organizational commitment as mediators in the leadership-turnover relationship is well supported by established organizational psychology theories. According to the three-component model of organizational commitment (Meyer & Allen, 1991, 1997), job satisfaction is viewed as an immediate evaluative response that leads to stronger emotional attachment to the organization. Supporting this, meta-analyses have shown that job satisfaction reliably predicts affective organizational commitment (Mathieu & Zajac, 1990; Meyer et al., 2002). Additionally, classic turnover models (Hom & Griffeth, 1991; Mobley, 1977) suggest that changes in job satisfaction occur before the development of withdrawal-related attitudes. Moreover, the study by Luz et al. (2018) found that employees who feel involved in organizational processes and perceive adequate benefits and rewards report higher job satisfaction and affective organizational commitment, which, in turn, support their continued tenure and, consequently, lower turnover intentions. Together, these perspectives indicate a sequence in which leadership styles influence employees’ job satisfaction, which, in turn, promotes affective organizational commitment and ultimately influences turnover intentions.

In the context of a constructive leadership style, transformational leadership has been positively associated with job satisfaction and affective organizational commitment, thereby fostering a supportive, motivating, and empowering work environment. Previous studies suggested that transformational leadership enhances job satisfaction and affective organizational commitment, reducing turnover cognitions (Breevaart & Zacher, 2019; Saira et al., 2021). These studies support the proposition that job satisfaction and affective organizational commitment represent critical pathways through which transformational leadership indirectly reduces employees’ turnover intentions.

In contrast, abusive leadership undermines job satisfaction and affective organizational commitment by fostering destructive work environments characterized by intimidation, unfairness, and lack of support. Previous studies suggest that abusive leadership is associated with reduced satisfaction and weakened affective organizational commitment, which in turn increases employees’ likelihood of turnover intentions (e.g., Alsadaan & Alqahtani, 2024; Bou Reslan et al., 2025; Ofei et al., 2023). This aligns with previous studies showing that destructive leadership behaviors erode psychological resources and organizational attachment, thereby increasing turnover intentions (Labrague et al., 2020; Ofei et al., 2023). Recent studies further highlight that the adverse effects of abusive leadership on job satisfaction and affective organizational commitment may be particularly severe in high-demand contexts, where employees have limited coping mechanisms (Nunes & Palma-Moreira, 2024).

This evidence underscores the importance of examining positive and negative leadership styles in relation to turnover intentions, with job satisfaction and affective organizational commitment serving as mediators. Accordingly, in this study, the following hypotheses are proposed:

Hypothesis 3

Job satisfaction and affective organizational commitment mediate the relationship between transformational leadership and turnover intentions.

Hypothesis 4

Job satisfaction and affective organizational commitment mediate the relationship between abusive leadership and turnover intentions.

Procedure and Sample

The data were collected through Google Forms. A convenience sampling approach was employed, whereby several organizations were contacted and invited to facilitate the dissemination of the study among their employees. Following their voluntary agreement to disseminate the study among their employees, the organizations were provided with a survey link. Subsequently, the participating organizations issued an email to their employees, inviting them to participate in the study. The link initially directed participants to an informed consent form, which, upon completion and signature, permitted access to the questionnaire.

The questionnaire responses were anonymous, and no identifying information was requested, thereby ensuring confidentiality and anonymity.

A total of 304 participants employed in various organizations across Portugal and across different sectors were included. The participants were, on average, approximately 34 years old (SD = 11.48). Most participants were women (58.9%), whereas the remaining were men (41.1%). Regarding education, 7.2% of participants had less than 9 years of schooling, 23.0% had twelve years of schooling, 36.8% had graduated, 31.6% held a master’s degree, and 1.3% held a PhD. 66.8% of the participants held permanent contracts, and most participants (40.1%) had between one and 5 years of service in their current organization. 33.2% of participants reported teleworking, while 66.8% reported working exclusively on-site.

Instruments

Measurement Model and Common Method Variance

A test of the measurement model was conducted to assess the potential impact of common method variance and to establish discriminant validity (Podsakoff et al., 2003, 2012). As an initial diagnostic, Harman’s single-factor test was performed by entering all measurement items into an unrotated principal component analysis. The results indicated that five factors with eigenvalues greater than 1.0 emerged, accounting for 65.11% of the total variance, whereas the first factor accounted for only 35.70%. These findings suggest that no single factor accounted for the majority of the covariance among the measures, indicating that common method variance was unlikely to be a serious concern. Consistent with this conclusion, a one-factor confirmatory factor analysis (CFA) model, in which all items were specified to load onto a single latent factor, exhibited poor fit to the data [χ² (299) = 2193.25, p < .01, SRMR = .13; CFI = .58; IFI = .58; RMSEA = .15]. In contrast, the hypothesized five-factor model demonstrated an acceptable fit [χ² (289) = 618.17, p < .01, SRMR = .05; CFI = .93; IFI = .93; RMSEA = .06] and fit the data significantly better than the one-factor model [Δχ² (10) = 1575.08, p < .01]. Overall, these analyses indicate that the factor structure of the study variables was consistent with the proposed conceptual model and that the observed indicators loaded on their intended latent constructs.

Descriptive Statistics

As shown in Table 1, participants reported moderate levels of transformational leadership (M = 3.78, SD = 0.90) and relatively low levels of abusive leadership (M = 1.13, SD = 0.29). Mean scores indicated moderately high job satisfaction (M = 3.82, SD = 0.75) and affective organizational commitment (M = 4.76, SD = 1.34), as well as relatively low turnover intentions (M = 2.42, SD = 1.09).

OPEN IN VIEWER

Table 1.

Descriptive Statistics and Correlation Matrix

	Mean	SD	1.	2.	3.	4.	5.	6.	7.	8.	9.
1. Sex a	0.59	0.49
2. Age	33.70	11.48	−.03
3. Education a	4.93	1.04	.03	−.32**
4. Tenure	2.63	1.58	−.07	.66**	−.35**
5. Telework a	0.33	0.47	−.08	−.09	.40**	−.14*
6. Transformational leadership	3.78	0.90	−.01	−.04	.14*	−.06	.22**
7. Abusive leadership	1.13	0.29	−.03	.02	−.17**	.03	−.08	−.33**
8. Job satisfaction	3.82	0.75	−.07	.07	.12*	.11	.15*	.49**	−.39**
9. Affective organizational commitment	4.76	1.34	−.06	.26**	−.10	.28**	.05	.44**	−.25**	.63**
10. Turnover intentions	2.42	1.09	.08	−.27**	.11	−.25**	−.06	−.31**	.14*	−.53**	−.55**

The correlations among the studied variables were consistent with expectations (Table 1). Transformational leadership was positively correlated with job satisfaction (r = .49, p < .01) and affective organizational commitment (r = .44, p < .01) and negatively correlated with turnover intentions (r = −.31, p < .01). Thus, Hypothesis 1 was supported. Abusive leadership was negatively associated with job satisfaction (r = −.39, p < .01) and affective organizational commitment (r = −.25, p < .01) and positively associated with turnover intentions (r = .14, p < .05). Thus, Hypothesis 2 was supported. Job satisfaction was positively correlated with affective organizational commitment (r = .63, p < .01) and negatively correlated with turnover intentions (r = −.53, p < .01). Affective organizational commitment was also negatively associated with turnover intentions (r = −.55, p < .01).

Several demographic variables were significantly correlated with the study variables (Table 1). Age was positively correlated with affective organizational commitment (r = .26, p < .01) and negatively correlated with turnover intentions (r = −.27, p < .01). Education was positively associated with transformational leadership (r = .14, p < .05) and job satisfaction (r = .12, p < .05) and negatively associated with abusive leadership (r = −.17, p < .01). Tenure was positively correlated with affective organizational commitment (r = .28, p < .01) and negatively correlated with turnover intentions (r = −.25, p < .01). Telework was positively linked with transformational leadership (r = .22, p < .01) and job satisfaction (r = .15, p < .05), indicating that those who work remotely tend to perceive a more transformational leader and experience higher job satisfaction than onsite workers.

Mediating Models Testing

The PROCESS macro (Model 6), with 5,000 bootstrap resamples and a 95% confidence interval, was used to test serial multiple mediation models. Two separate models were tested: (1) Transformational leadership as the independent variable (X), and (2) abusive leadership as the independent variable (X). In both models, job satisfaction (M1) and affective organizational commitment (M2) were entered as mediators sequentially, with turnover intention (Y) as the dependent variable.

Limitations and Future Research

Although this study offers valuable insights, several limitations must be recognized. First, the cross-sectional design prevents causal inference, as associations cannot establish temporal precedence. Methodological work has consistently shown that cross-sectional mediation analyses may yield biased or misleading estimates of longitudinal mediation processes and cannot empirically verify temporal precedence among variables (Cole & Maxwell, 2003; Maxwell & Cole, 2007; Maxwell et al., 2011). As a result, the proposed ordering of leadership styles, job satisfaction, affective organizational commitment, and turnover intentions should be understood as theoretically derived rather than empirically confirmed. Thus, future research employing longitudinal or experimental designs would help more robustly test the mediating role of job satisfaction and affective organizational commitment in the leadership-turnover relationship. Second, all data were collected via employee self-reports at a single point in time, an inherent limitation that raises the possibility of common method variance. Although self-reported measures were appropriate, given that the focal constructs capture employees’ perceptions of leadership behaviors and job-related attitudes grounded in personal experiences, such perceptual assessments may nonetheless be influenced by shared affective or cognitive biases. For instance, individuals’ generalized evaluative tendencies or transient affective states may shape concurrent appraisals of leadership and work-related attitudes, potentially inflating observed associations. To mitigate these concerns, several procedural and statistical remedies were implemented in line with the methodological recommendations of Podsakoff et al. (2003, 2012). Specifically, respondents were assured of anonymity and confidentiality, informed that there were no right or wrong answers, and survey items were presented in a counterbalanced order. In addition, we conducted Harman’s single-factor test and confirmatory factor analyses, whose results suggest that common method variance is unlikely to fully account for the observed relationships. Nevertheless, the possibility of residual method bias cannot be ruled out entirely. Accordingly, future research should employ multi-source and/or multi-wave designs, such as combining employee self-reports with supervisor ratings or objective indicators, to further distinguish perceptual biases from substantive relationships and to strengthen causal inferences (Gallegos et al., 2022). Third, although key demographic characteristics of the sample were reported, finer-grained information on participants’ job categories, industries, and organizational contexts was unavailable. The data collection protocol prioritized respondent anonymity and confidentiality and sought to minimize response burden, which constrained the amount of organizationally identifiable information that could be collected. As a result, caution is warranted when interpreting the findings across specific occupational or industry contexts. Future studies would benefit from collecting more detailed contextual descriptors to examine whether the observed relationships vary across job roles, industries, or organizational settings. Fourth, the study was conducted in Portugal, which may limit generalizability to other cultural settings. As such, the current findings should be interpreted in light of their cultural context. Conducted within a Portuguese sample, the findings reflect leadership perceptions embedded in a societal context characterized by moderate collectivism and relatively high power distance, a cultural configuration shared with several Southern European countries, such as Spain, Italy, and Greece, which may influence employees’ responses to both transformational and abusive supervisory behaviors (Akaliyski et al., 2026). Comparative research suggests that leadership effects and their underlying mechanisms vary across cultures, limiting the direct generalization of the present pattern of associations to contexts with different cultural profiles (Den Hartog & De Hoogh, 2024; Rockstuhl et al., 2023). Future research should replicate the proposed model across diverse cultural settings using longitudinal or multi-wave designs to examine whether these relationships hold under different societal norms and leadership expectations. Finally, future research could extend the model by examining additional variables that may help explain or condition the observed associations. Variables such as psychological safety, resilience, and perceived organizational support may be relevant correlates that further illuminate the relationships between leadership and turnover intentions (Djourova et al., 2020; Siyal, 2023). Similarly, individual differences, such as emotional intelligence and self-efficacy, could serve as boundary conditions, potentially altering the strength or direction of these associations (Bui et al., 2024; Tian & Guo, 2024). Investigating such contingencies, particularly using longitudinal or multi-method designs, would help clarify the temporal ordering and robustness of these relationships, thereby refining our understanding of when and for whom leadership styles are most consequential for retention outcomes.

Theoretical and Practical Contributions

This study advances research on leadership and turnover by exploring the links between transformational and abusive leadership and turnover intentions. While transformational leadership has been studied in relation to positive organizational outcomes, abusive leadership has more recently gained attention as a distinct concept, often examined separately. By considering both leadership styles, this study extends theoretical frameworks that account for the dual nature of leadership (i.e., constructive and destructive) and their distinct pathways to turnover intentions. Furthermore, the results contribute to the literature by showing that job satisfaction and affective organizational commitment mediate these associations, thereby reinforcing the central role of work attitudes in the leadership-turnover relationship.

From a practical perspective, the findings underscore the importance of actively cultivating transformational leadership behaviors while systematically preventing and correcting destructive leadership practices. Beyond general leadership development, organizations may benefit from implementing evidence-based interventions that have been shown to shape leader behavior. For instance, leadership coaching and structured 360-degree feedback processes can enhance leaders’ self-awareness, reinforce transformational behaviors such as inspirational motivation and individualized consideration, and reduce harmful conduct (Aasland et al., 2010; Bass & Riggio, 2006). Such interventions enable leaders to align their behavior more closely with employees’ expectations for support, fairness, and empowerment, thereby strengthening job satisfaction and affective organizational commitment. In parallel, organizations may consider preventive initiatives to identify and mitigate destructive leadership patterns early. Civility training and behavioral assessment tools have been shown to reduce abusive supervision and interpersonal mistreatment by clarifying behavioral norms and increasing accountability (Tepper et al., 2017). Embedding these practices in leadership development programs, performance appraisal systems, and promotion criteria may help organizations discourage harmful leadership behaviors that erode trust and morale. Collectively, these targeted interventions provide actionable pathways for HR practitioners and organizational leaders to translate the study’s findings into retention-focused strategies by fostering healthier leadership climates and reducing employees’ intentions to leave.