While the benefits of servant leadership are well established, its effectiveness during major organizational change remains unclear. Grounded in conservation of resources theory, this study examines how servant leadership functions in such transitions, particularly when employees experience job insecurity.
Survey data (two waves) and administrative sickness absence records were collected from 215 employees during the anticipation phase of a merger.
Servant leadership predicts reduced sickness absence and turnover intentions through the mediating effect of emotional exhaustion. However, these positive effects weaken significantly when employees experience job insecurity.
The study provides insights for organizations managing mergers. Managing employee job un-certainty seems important for tailoring leadership, including the efficacy of servant leadership.
Our results, while surprising, expand our understanding of the boundary conditions of servant leadership and contribute to a realistic discussion of its effectiveness during times of uncertainty and change.
Introduction
Servant leadership, characterized by leader humility and altruism, stands out in the field of positive leadership for its unique benefits on employee well-being and organizational performance (Eva et al., 2019). Despite its prominence in leadership practice and scholarship, there is ongoing debate about its applicability and effectiveness across various contexts. One context where servant leaders' influence is unclear is the realm of significant organizational changes, such as mergers and acquisitions (M&As), where leaders are tasked with navigating highly uncertain environments. A merger is often a highly disruptive socio-psychological event that can trigger a range of negative reactions from those affected, including mistrust, stress, and health concerns (Seo and Hill, 2005; Sung et al., 2017). Despite theoretical contentions that servant leadership may represent an “ideal leadership approach for introducing and implementing change within the context of organizations” (Baldomir and Hood, 2016, p. 27), empirical evidence does not fully align with this optimistic view (Eva et al., 2019). The few existing studies on servant leadership in change settings show varied limitations and offer a fragmented view of its effectiveness. Some research relies on hypothetical scenarios (Van Dierendonck et al., 2014), which can limit ecological validity (i.e. the extent to which findings generalize to real-world settings), while others examine perceived uncertainty in organizational contexts not actively undergoing change (Kool and Van Dierendonck, 2012). This disconnect has led even strong advocates of servant leadership to concede that, while there is compelling evidence supporting its overall effectiveness, it often falls short in demonstrating a direct connection to “the specific outcomes of clearly defined and bounded organizational change efforts” (Roberts, 2020b, p. 50).
Few studies have explored servant leadership during actual change processes (Heine et al., 2022; Sousa and Van Dierendonck, 2014), and those that have often overlook the subjective experiences of those affected by the change, particularly the felt uncertainty. Change research has consistently emphasized that individuals are not inherently resistant to change or naturally stressed by it. Instead, their reactions are best understood as a function of their subjective appraisals of the change, specifically the perceived threats or benefits (Oreg et al., 2018). Consistent with this perspective, perceptions of uncertainty and threats to job continuity are known to vary among individuals (Rafferty and Griffin, 2006). Thus, understanding servant leadership’s influence on employee reactions to change requires a closer examination of these personal experiences.
With these issues in mind, our main purpose in this research is to scrutinize servant leadership’s role in the context of significant organizational change. Specifically, we draw on conservation of resources (COR) theory (Hobfoll et al., 2018) to investigate the interplay of servant leadership and change recipients' individually perceived job insecurity in the context of an ongoing merger process. Building on the established notion that servant leadership helps provide, preserve, and build essential resources for employee well-being (Bentein and Panaccio, 2020; Eva et al., 2019), we propose that it can mitigate negative responses often associated with mergers and acquisitions (M&As), specifically emotional exhaustion, turnover intentions, and sickness absence among change recipients. However, a critical question remains: whether and to what extent these positive effects of servant leadership persist when accounting for employees’ subjectively perceived job insecurity.
Addressing these issues offers important contributions to theory and research. By identifying potential boundary conditions of servant leadership, we help refine and calibrate its underlying assumptions. This scrutiny is important, especially given critiques of servant leadership for its “excessive positivity” and a “generally simplistic and out of date view of corporate life” (Alvesson and Einola, 2019, p. 383). This analysis not only helps clarify the strengths and limitations of servant leadership but also has practical implications for leader selection, development, and the application of leadership strategies during organizational change.
Theoretical background and hypothesis development
In what follows, we present our theoretical model (Figure 1), outlining the interaction between servant leadership and perceived job insecurity during change.
The flowchart begins with a text box on the left labeled “Servant Leadership.” A rightward arrow emerges from “Servant Leadership” and connects to a central text box labeled “Emotional Exhaustion.” Two diagonal rightward arrows emerge from “Emotional Exhaustion” and connect to two text boxes positioned on the right and arranged vertically; the top text box is labeled “Sickness Absence,” and the bottom text box is labeled “Turnover Intentions.” A diagonal rightward arrow emerges from “Servant Leadership” and connects to “Sickness Absence.” A bottom rightward arrow emerges from “Servant Leadership” and connects to “Turnover Intentions.” Three upward arrows emerge from a text box placed below “Emotional Exhaustion,” labeled “Perceived Job Insecurity” and connect to the arrows that link “Servant Leadership” to “Sickness Absence,” to “Emotional Exhaustion,” and to “Turnover Intentions,” respectively.Theoretical model. Source: Authors’ own creation
The flowchart begins with a text box on the left labeled “Servant Leadership.” A rightward arrow emerges from “Servant Leadership” and connects to a central text box labeled “Emotional Exhaustion.” Two diagonal rightward arrows emerge from “Emotional Exhaustion” and connect to two text boxes positioned on the right and arranged vertically; the top text box is labeled “Sickness Absence,” and the bottom text box is labeled “Turnover Intentions.” A diagonal rightward arrow emerges from “Servant Leadership” and connects to “Sickness Absence.” A bottom rightward arrow emerges from “Servant Leadership” and connects to “Turnover Intentions.” Three upward arrows emerge from a text box placed below “Emotional Exhaustion,” labeled “Perceived Job Insecurity” and connect to the arrows that link “Servant Leadership” to “Sickness Absence,” to “Emotional Exhaustion,” and to “Turnover Intentions,” respectively.Theoretical model. Source: Authors’ own creation
Servant leadership, sickness absence, and turnover intentions
This research examines two crucial outcomes of organizational change: employee health, operationalized as sickness absence, and turnover intentions. These align with the core principles of COR theory, which posits that individuals strive to build, protect, and foster their personal resources (Hobfoll et al., 2018). Stress arises when resources are lost, threatened, or difficult to acquire. Disruptive organizational changes, such as M&As, represent a stressful context where actual or perceived resource loss is prevalent (Oancea and Kamau, 2020). Such stress has been consistently linked to adverse outcomes, including diminished psychological and physical health (de Jong et al., 2016) and psychological withdrawal behaviors like increased turnover intentions (Fugate et al., 2011).
Drawing on COR theory, we propose that servant leadership acts as a resource-enhancing factor, counteracting the negative effects of resource loss during organizational change. Through support, empathy, and resource investment, servant leaders help employees replenish lost resources and buffer against stress. Empirical research consistently supports the positive impact of servant leadership on employee well-being (Roberts, 2020a) and positive work attitudes such as job satisfaction and organizational commitment (Eva et al., 2019). Emerging evidence further suggests that these benefits can extend to organizational change contexts (Sousa and Van Dierendonck, 2014).
Building on this, we hypothesize that servant leadership mitigates sickness absence and turnover intentions among employees during M&As. We posit that this effect operates indirectly through reduced emotional exhaustion, a core indicator of resource depletion. Servant leadership provides employees with access to resources that replenish energy and reduce emotional exhaustion (Lin et al., 2025; Rivkin et al., 2014), a key driver of both sickness absence and turnover intentions (Schaufeli et al., 2009). Thus, we specified the following hypotheses:
Servant leadership is negatively related to sickness absence through the mediating effect of reduced emotional exhaustion.
Servant leadership is negatively related to turnover intentions through the mediating effect of reduced emotional exhaustion.
The moderating role of perceived job insecurity
M&As pose tangible threats such as loss of status, benefits, opportunities, and even employment. Research consistently highlights that M&As heighten job insecurity, commonly conceptualized as the perception of a potential threat to job continuity (Lee et al., 2018; Probst et al., 2018). More broadly, job insecurity reflects an employee's subjective appraisal of job stability and the threat of losing their job or job-related resources (e.g. working conditions, career opportunities). Notably, it has been found that these perceptions are particularly high in the anticipation stage, when a merger is announced yet post-M&A integration is not ready (Seo and Hill, 2005). During this period, employees often struggle to assess and evaluate the impact of the change on their jobs and the availability of resources to manage it. Importantly, research suggests that in such volatile contexts, leadership and other organizational support systems often have limited direct influence over employees' perceptions of job insecurity. Instead, they interact with these perceptions in shaping employees' responses to change (Lee et al., 2018).
On this basis, we argue that perceived job insecurity exerts an interaction effect with servant leadership in predicting reduced emotional exhaustion. Under elevated job insecurity, employees are more vulnerable to resource depletion, making servant leadership's replenishing effect more pronounced. Conversely, when job insecurity is low, emotional strain is reduced, and additional leadership support becomes less critical. These predictions align with COR theory, which posits that in contexts of resource loss, individuals are especially sensitive to resource gains that help prevent further depletion and restore stability (Hobfoll et al., 2018). In such situations, servant leadership can serve as a valuable external resource that buffers employees against emotional exhaustion. Leaders may play a unique role compared to other organizational resources, as they are perceived as directly shaping employees' work experiences. Iindirect support for this argument comes from substitutes-for-leadership perspectives (Jermier and Kerr, 1997), which suggest that when employees feel secure about their job future, other resources, such as routines, team cohesion, or formal policies, can reduce the need for leadership intervention. However, under heightened job insecurity, these substitute resources are less salient, making leadership a more important factor in mitigating emotional exhaustion. Thus, we propose:
Perceived job insecurity moderates the negative effect of servant leadership on emotional exhaustion, such that higher job insecurity strengthens the effect.
Implicit to this prediction is the assumption that perceived job insecurity will conditionally influence the indirect effects of servant leadership on the focal outcomes (i.e. turnover intentions and sickness absence), demonstrating a moderated mediation pattern. Specifically, the indirect effect of servant leadership strengthens when both servant leadership and job insecurity are high, as heightened insecurity increases employees’ reliance on supportive leadership, making its resource-restoring effects more pronounced. Thus, we proposed:
The mediating effect of reduced emotional exhaustion in the negative relationship between servant leadership and sickness absence is moderated by perceived job insecurity. The higher the perceived job insecurity, the stronger the mediation effect.
The mediating effect of reduced emotional exhaustion in the negative relationship between servant leadership and turnover intentions is moderated by perceived job insecurity. The higher the perceived job insecurity, the stronger the mediation effect.
While we emphasize reduced emotional exhaustion as the primary pathway through which servant leadership mitigates sickness absence and turnover intentions, prior research suggests that additional mechanisms, beyond those captured in our model, are likely involved. For instance, servant leadership fulfills employees' basic psychological needs for autonomy, competence, and relatedness and fosters empowerment (Eva et al., 2019), which may further reduce turnover intentions and support health during change. The strength of these mechanisms, however, is likely contingent on the resource-depleting impact of job insecurity. Statistically, this implies that job insecurity also moderates the direct effects of servant leadership on sickness absence and turnover intentions. Given the complexity of additional mechanisms, we pose research questions (vs. hypotheses) to explore these relationships:
Does perceived job insecurity moderate the direct effect of servant leadership on reduced sickness absence such that higher levels of job insecurity strengthen the effect?
Does perceived job insecurity moderate the direct effect of servant leadership on reduced turnover intentions such that higher levels of job insecurity strengthen the effect?
Methods
Sample and procedure
Data were collected in the German subsidiary of a large multinational engineering and industrial consulting company (company A), in which top management had announced the successful acquisition of a competitor company (company B). The integration of the two companies was planned for the following year, leaving both companies operating separately in the interim.
The sample for our research exclusively comprises employees from company A (the acquiring company). It is important to highlight that the stakes for employees within company A were notably high as the acquisition was part of a comprehensive restructuring plan, reflecting company A's intention to overhaul significant aspects of its local operations and services. Shortly after the announcement, CEO and board member resignations occurred in both companies, and future integration and re-structuring strategies remained unclear, creating a turbulent “post-acquisition, pre-integration” context ideal for testing our theoretical model.
Two personalized surveys were sent to all potential participants at two points in time, separated by approximately 8 weeks. At Time 1, employees were asked to rate their immediate leader's leadership behavior and to provide personal information. At Time 2, respondents indicated their perceived job insecurity and individual work attitudes, i.e. emotional exhaustion and turnover intentions. Sickness absence data was obtained from the payroll manager at the end of the year.
Overall 586 out of 1,140 employees of the company responded at Time 1 and 491 out of 1,140 responded at Time 2, reaching a response rate of 47.24% on average. We excluded employees who completed only one survey, reported a supervisor change between Time 1 and Time 2, or had less than one month of tenure with their leader (1 case), as a short timeframe compromises rating reliability. The final sample comprised 215 employees, nested in 52 leaders (range: 1–16 members, average team size = 4.22, SD = 3.33). The average tenure with the leader was 1.76 years (SD = 2.02). Regarding demographics, 64% of participants were under 35, 30% were between 35 and 50, and 5% were between 50 and 60. The sample was predominantly male (80%), and 92% held a university degree.
Measures
Servant leadership. Servant leadership was captured with a 14-item measure developed by Ehrhart (2004). A sample item was: “My supervisor makes the personal development of his/her employees a priority”. Recent research by Van Dierendonck et al. (2023) validated this global servant leadership measure, demonstrating that it effectively captures the core elements of servant leadership. All items were measured on a 5-point Likert scale ranging from 1 (never) to 5 (frequently/almost always).
Perceived job insecurity. This variable was assessed with the four-item scale from Mauno and Kinnunen (2002). A sample item was: “How certain are you about what your future career picture looks like in your organization?” All items were scored on a 7-point scale (1 = not certain at all, 7 = very certain). Item scores were reversed so that high scores on the composite scale reflect a high level of perceived insecurity.
Emotional exhaustion. To measure emotional exhaustion we used the three-item measure provided by De Cuyper et al. (2012). A sample item was “I feel emotionally drained from my work.” Ratings were made on a 7-point Likert scale ranging from 1 (totally disagree) to 7 (totally agree).
Turnover intentions. Turnover intentions were assessed with two items adapted from the Michigan Organizational Assessment Questionnaire (Cammann et al., 1979) on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). A sample item was “I often think of leaving the organization.”
Sickness absence. We compiled a comprehensive record of annual sickness absence for each employee. Sickness absence duration was calculated in calendar days from the start of registered sick leave until the return to work. Documentation for each sick leave included the employee’s name, number of absence days, and the manager and business unit of affiliation at the time of absence. All employees, regardless of their employee status, needed a medical certificate for sick leaves of over 3 days. Absences due to caring for a sick child or maternity leave were not included in these files. Absences varied from 0 to 41 days. To comply with data protection and confidentiality regulations, we used categorical data for our analyses based on the reporting standards of German health insurance providers (Schenkel et al., 2024). Absences were grouped into the following categories: 0 days, 1–3 days, 4–7 days, 8–14 days, 15–21 days, 22–28 days, and 29–42 days.
Control variables. To test the unique contribution of servant leadership, we controlled for charismatic leadership as a competing correlated leadership style. We chose charismatic leadership because it has been portrayed as particularly suited for turbulent, unpredictable environments, including M&As (Waldman and Javidan, 2009). While charismatic and servant leadership share the focus on relationship building, charismatic leadership focuses more on aligning followers with the organization’s goals through articulating a clear vision and highlighting a shared purpose. Following Banks et al. (2017), we used the inspirational motivation and idealized influence behavior sub-scales from the Multifactor Leadership Questionnaire (Bass and Avolio, 1995) to measure charismatic leadership (e.g. “My leader talks optimistically about the future”). Responses were given on a 5-point scale (1 = never to 5 = frequently/almost always).
Results
Measurement model
To examine the integrity of the measurement model, we compared several theoretically viable models through confirmatory factor analysis (CFA). The hypothesized five-factor model (i.e. servant leadership, charismatic leadership, perceived job insecurity, turnover intentions, emotional exhaustion) demonstrated adequate fit (χ2 = 638.908, df = 395, p < 0.001, χ2/df = 1.617, CFI = 0.932, RMSEA = 0.054) and outperformed alternative models. Detailed CFA results are available in the supplementary material (Supplement A).
Descriptive statistics
Table 1 presents the descriptive statistics and correlations among the study variables. The mean score for perceived job insecurity (M = 3.977, SD = 1.423) was moderate but consistent with findings from other studies conducted during organizational change (Woehler et al., 2021). Notably, it exceeds the levels reported in studies of the general population in Germany (Roll et al., 2015).
Means, standard deviations, and correlations
| Variable | M | SD | 1 | 2 | 3 | 4 | 5 | ||
|---|---|---|---|---|---|---|---|---|---|
| 1 | Servant Leadership | 3.506 | 0.741 | (0.93) | |||||
| 2 | Perceived Job Insecurity | 3.977 | 1.423 | −0.447*** | (0.86) | ||||
| 3 | Charismatic Leadership | 3.726 | 0.718 | 0.725*** | −0.269*** | (0.86) | |||
| 4 | Emotional Exhaustion | 3.173 | 1.321 | −0.267*** | 0.366*** | −0.212** | (0.83) | ||
| 5 | Turnover Intention | 3.541 | 1.667 | −0.399*** | 0.503*** | −0.296*** | 0.430** | (0.85) | |
| 6 | Sickness Absence | 1.265 | 1.260 | −0.035 | 0.120† | −0.008 | 0.173* | 0.126† | (−) |
Note(s): †p < 0.10, **p < 0.01, ***p < 0.001. Cronbach’s alphas are reported on the diagonal. For turnover intentions, a two-item measure, the Spearman-Brown reliability estimate is provided
Hypothesis testing
We conceptualize servant leadership as an individual perception interacting with job insecurity as a personal experience, framing our hypotheses at the individual level. However, the nested structure of our data (followers within leaders) prompts concerns regarding the independence of observations and correct statistical inferences. Therefore, we examined whether multilevel modeling was needed. To this end, we computed the ICC(1) (i.e. the amount of variance that can be explained by group membership) and ICC(2) (i.e. the reliability of the means). For sickness absence, the ICC(1) was −0.05 (p = 0.84) and the ICC(2) was −0.28. The ICC(1) for emotional exhaustion was 0.05 (p = 0.14) and the ICC(2) was 0.20. The ICC(1) for turnover intentions was −0.02 (p = 0.66) and the ICC(2) was −0.11. The non-significant p-value suggests that the observed variance between groups is not significantly greater than what would be expected by chance. Negative ICC(1) values are mathematically possible but typically considered artifacts, often resulting from small sample sizes or high within-group variability. They indicate that within-group variance exceeds between-group variance, suggesting that the grouping structure is weak or meaningless. Similarly, negative ICC(2) values indicate that individual differences overshadow any group-level consistency, resulting in low group-level reliability. Given these patterns, we tested our predictions at the individual level.
We conducted path modeling using the PROCESS macro in SPSS (Hayes, 2018) to test the predicted effects. In a first step, two separate mediation analyses (PROCESS model 4) examined the indirect effects of servant leadership on sickness absence (Hypothesis 1) and turnover intentions (Hypothesis 2) via emotional exhaustion, controlling for charismatic leadership. Results showed that servant leadership was negatively related to emotional exhaustion (b = −0.426, p = 0.013), which, in turn, positively predicted both sickness absence (b = 0.180, p = 0.012) and turnover intentions (b = 0.440, p < 0.001). While the direct effect of servant leadership on sickness absence was not significant (b = −0.035, p = 0.847), the indirect effect via emotional exhaustion was significant (ab = −0.077, 95% CI [-0.186, −0.081]). For turnover intentions, servant leadership showed a significant direct effect (b = −0.686, p = 0.001) as well as a significant indirect effect through emotional exhaustion (ab = −0.188, 95% CI [-0.360, −0.038]). Charismatic leadership, included as a covariate, had no significant effect on emotional exhaustion (b = −0.071, p = 0.687), sickness absence (b = 0.081, p = 0.660), or turnover intentions (b = 0.000, p = 0.998), relative to servant leadership. These findings support Hypotheses 1 and 2, demonstrating the mediating role of reduced emotional exhaustion in the relationship between servant leadership and both sickness absence and turnover intentions.
Next, we tested the moderating role of job insecurity and included the interaction between servant leadership and perceived job insecurity in the regression models, using PROCESS model 8 (Table 2) [1]. As charismatic leadership exerted no significant effects in the mediation models, it was not included as a covariate in order to facilitate interpretability. All predictor variables were mean centered before starting the analysis.
Regression coefficients and confidence intervals
| Variable | B | SE | p | 95% CI | |
|---|---|---|---|---|---|
| LL | UL | ||||
| Outcome: EE | |||||
| Intercept | 3.244 | 0.094 | <0.001 | 3.066 | 3.421 |
| SL | −0.261 | 0.126 | 0.039 | −0.510 | −0.012 |
| JINS | 0.288 | 0.065 | <0.001 | 0.160 | 0.417 |
| SL × JINS | 0.150 | 0.075 | 0.048 | 0.001 | 0.298 |
| R2 | 0.163 | ||||
| F | 13.686 | <0.001 | |||
| Δ R2 | 0.016 | ||||
| Outcome: TOI | |||||
| Intercept | 2.648 | 0.265 | <0.001 | 2.124 | 3.172 |
| SL | −0.454 | 0.140 | 0.001 | −0.731 | −0.176 |
| EE | 0.309 | 0.075 | <0.001 | 0.159 | 0.458 |
| JINS | 0.392 | 0.075 | <0.001 | 0.243 | 0.540 |
| SL × JINS | 0.184 | 0.084 | 0.029 | 0.018 | 0.349 |
| R2 | 0.364 | ||||
| F | 30.084 | <0.001 | |||
| Δ R2 | 0.015 | ||||
| Outcome: SICK | |||||
| Intercept | 0.742 | 0.264 | 0.005 | 0.221 | 1.263 |
| SL | 0.093 | 0.139 | 0.504 | −0.182 | 0.369 |
| EE | 0.165 | 0.075 | 0.029 | 0.016 | 0.314 |
| JINS | 0.075 | 0.074 | 0.318 | −0.072 | 0.222 |
| SL × JINS | −0.066 | 0.083 | 0.429 | −0.230 | 0.098 |
| R2 | 0.038 | ||||
| F | 2.072 | 0.085 | |||
| Δ R2 | 0.003 | ||||
Note(s): N = 215. All coefficients are unstandardized. EE = Emotional Exhaustion, SL = Servant Leadership, CH = Charismatic Leadership, JINS = Job Insecurity, CI = Confidence Interval, LL = Lower Limit, UL = Upper Limit. Δ R2 refers to the inclusion of the interaction term. Bootstrap sample size = 5,000
With regard to emotional exhaustion, the results of the conditional analysis showed that it was predicted by both servant leadership (b = −0.261, p = 0.039) and job insecurity (b = 0.288, p < 0.001). The interaction between job insecurity and servant leadership was statistically significant (b = 0.150, p = 0.048). To facilitate interpretation, we tested simple slopes for the association between servant leadership and emotional exhaustion for low (one standard deviation below the mean), moderate (the mean), and high (one standard deviation above the mean) levels of job insecurity. Results showed that the effect of servant leadership was strongest at low levels of job insecurity (b = −0.483, p = 0.007), decreased at moderate levels (b = −0.295, p = 0.023), and became insignificant at high levels (b = 0.004, p = 0.980). This suggests that servant leadership becomes less effective for reducing emotional exhaustion, as job insecurity increases (Figure 2). Thus, Hypothesis 3 was not supported. A similar effect was found for the direct effect between servant leadership and turnover intentions (Q2). We found that turnover intentions were predicted by servant leadership (b = −0.454, p = 0.001), job insecurity (b = 0.392, p < 0.001), and emotional exhaustion (b = 0.309, p < 0.001). The interaction between servant leadership and job insecurity in predicting turnover intentions was significant (b = 0.184, p = 0.029). Simple slopes analysis revealed that the slopes of servant leadership predicting turnover intentions were significant at low (b = −0.726, p < 0.001) and medium (b = 0.495, p = 0.001), but not high levels of job insecurity (b = −0.127, p = 0.505). Again, this indicates that the negative relationship between servant leadership and turnover intentions becomes weaker as the value of job insecurity increases (Figure 2). Finally, while sickness absence was predicted by emotional exhaustion (b = 0.165, p = 0.029), neither servant leadership nor job insecurity showed direct effects, and the interaction between servant leadership and job insecurity was also not significant [2] (Q1).
The figure consists of two side-by-side line graphs. The left graph has the horizontal axis labeled “S L” with two labeled points from left to right, “Low S L” and “High S L.” The vertical axis represents “Emotional Exhaustion,” ranging from 1 to 7 in increments of 1 unit. Two lines are plotted, a solid line and a dashed line. The solid line labeled “Low J I N S” starts from approximately (Low S L, 3.236) and ends at (High S L, 2.506) and has a negative slope. The dashed black line labeled “High J I N S” starts from approximately (Low S L, 3.723) and ends at (High S L, 3.657) and has a slight negative slope. A legend on the right side shows that the solid line indicates “Low J I N S” and the dashed line indicates “High J I N S.” The right graph has the horizontal axis labeled “S L” with two labeled points from left to right, “Low S L” and “High S L.” The vertical axis represents “Turnover Intentions,” ranging from 1 to 7 in increments of 1 unit. Two lines are plotted, a solid line and a dashed line. The solid black line labeled “Low J I N S” starts from approximately (Low S L, 2.59) and ends at (High S L, 1.56) and has a negative slope. The dashed line labeled “High J I N S” starts from approximately (Low S L, 3.373) and ends at (High S L, 3.082) and has a slight negative slope. A legend on the right side shows the solid line indicates “Low J I N S” and the dashed line indicates “High J I N S.” Note: All numerical data values are approximated.The interaction effect of servant leadership and perceived job insecurity on emotional exhaustion and turnover intentions. Note. SL = Servant Leadership, JINS = Job Insecurity. Source: Authors’ own creation
The figure consists of two side-by-side line graphs. The left graph has the horizontal axis labeled “S L” with two labeled points from left to right, “Low S L” and “High S L.” The vertical axis represents “Emotional Exhaustion,” ranging from 1 to 7 in increments of 1 unit. Two lines are plotted, a solid line and a dashed line. The solid line labeled “Low J I N S” starts from approximately (Low S L, 3.236) and ends at (High S L, 2.506) and has a negative slope. The dashed black line labeled “High J I N S” starts from approximately (Low S L, 3.723) and ends at (High S L, 3.657) and has a slight negative slope. A legend on the right side shows that the solid line indicates “Low J I N S” and the dashed line indicates “High J I N S.” The right graph has the horizontal axis labeled “S L” with two labeled points from left to right, “Low S L” and “High S L.” The vertical axis represents “Turnover Intentions,” ranging from 1 to 7 in increments of 1 unit. Two lines are plotted, a solid line and a dashed line. The solid black line labeled “Low J I N S” starts from approximately (Low S L, 2.59) and ends at (High S L, 1.56) and has a negative slope. The dashed line labeled “High J I N S” starts from approximately (Low S L, 3.373) and ends at (High S L, 3.082) and has a slight negative slope. A legend on the right side shows the solid line indicates “Low J I N S” and the dashed line indicates “High J I N S.” Note: All numerical data values are approximated.The interaction effect of servant leadership and perceived job insecurity on emotional exhaustion and turnover intentions. Note. SL = Servant Leadership, JINS = Job Insecurity. Source: Authors’ own creation
The index of moderated mediation for the effect of servant leadership on sickness absence via emotional exhaustion was 0.025 and not significant, 95% CI [−0.005, 0.069]. Likewise, for the indirect effect on turnover intentions, the index of moderated mediation was 0.046 and not significant (95% CI [−0.007, 0.110]). These patterns indicate that the moderation effect of job insecurity on emotional exhaustion was not strong enough to impact the entire indirect effects significantly. Thus, Hypotheses 4a and 4b were not supported.
Additional results
We conducted additional analyses to further examine two key aspects of our findings. First, the high correlation between servant leadership and charismatic leadership (r = 0.725) raised concerns about multicollinearity and the stability of regression estimates. When analyzed alone, charismatic leadership significantly predicted emotional exhaustion (b = −0.390, p = 0.002) and turnover intentions (b = −0.496, p < 0.001). However, adding servant leadership rendered charismatic leadership non-significant in both cases, regardless of entry order, indicating no added explanatory value. Relative weight analyses (Nimon and Oswald, 2013) showed servant leadership as the dominant predictor of both outcomes, but the difference in relative weights was statistically significant only for turnover intentions. Thus, results for emotional exhaustion should be interpreted with caution. Full results are reported in the supplementary materials (Supplement C). However, combined with the finding that job insecurity moderated only the effects of servant leadership, these analyses underscore its unique role in shaping employee outcomes in our data.
Second, given the buffering effect of job insecurity on the relationship between servant leadership and key outcomes (emotional exhaustion and turnover intentions) at higher levels of insecurity, we tested whether this moderation followed a nonlinear pattern. Specifically, we examined whether the benefits of servant leadership might be strongest when insecurity is low, weaken at moderate levels, and re-emerge at high levels. Results did not support such quadratic effects. Full details are provided in the supplementary materials (Supplement D).
Discussion
This study explored how servant leadership and job insecurity interact in an M&A context to predict employee well-being and retention. While perceived job insecurity did not moderate the negative indirect effect of servant leadership on sickness absence and turnover intentions via emotional exhaustion, it affected the direct effects on emotional exhaustion and turnover intentions. Contrary to our predictions, the obtained moderating effects suggest that servant leadership becomes less effective when employees experience job insecurity.
Theoretical implications
A first insight from our research is that servant leadership is associated with lower levels of emotional exhaustion and reduced turnover intentions during change, outperforming charismatic leadership. In doing so, our study contributes to the ongoing conversation on the differential mechanisms and incremental value of servant leadership, particularly in comparison to transformational leadership, with charismatic leadership representing an important facet of this discussion (Hoch et al., 2018).
However, our findings also temper the prevailing optimism surrounding servant leadership (Baldomir and Hood, 2016; Roberts, 2020b). While consistent with prior research in showing that servant leadership is negatively related to emotional exhaustion and turnover intentions, we find that this effect diminishes, and can even disappear, when employees perceive high job insecurity. This does not suggest that servant leadership is detrimental during change, but rather that its effectiveness depends on specific recipient experiences, particularly perceptions of job insecurity. Thus, our results add a caveat to the few previous studies that have linked servant leadership to positive outcomes during mergers, such as increased engagement or enhanced organizational goal clarity (Heine et al., 2022; Sousa and Van Dierendonck, 2014). These positive effects might weaken when employees feel more insecure about their job situation.
The notion that servant leadership’s effectiveness varies across different situations, especially in the context of organizational change and varying levels of uncertainty, offers a critical opportunity to reassess core tenets of servant leadership theory. It particularly challenges the often unquestioned acclaim of servant leadership (Alvesson and Einola, 2019), encouraging a more context-sensitive examination of its efficacy and potential pitfalls. Specifically, from a COR theory perspective, it seems plausible that during the highly volatile anticipation phase of an impending merger, employees may benefit more from leadership qualities that extend beyond servant leadership's relational and supportive aspects. More research is needed to clarify which leadership resources are most effective for employees experiencing heightened job insecurity in this phase of change. It is conceivable that more directive forms of leadership may have a stabilizing effect in such scenarios (Rast et al., 2013). Specifically, directive leadership may help employees maintain a sense of role clarity and availability of role resources, which has been associated with reduced burnout (Vullinghs et al., 2020) and intentions to quit (Hassan, 2013). In contrast, servant leadership may be more effective in later stages of the merger, most notably in the integration period. Research by Schweizer and Patzelt (2012) suggests that the speed of change integration is positively related to change recipients' commitment, while this effect is amplified by perceptions of supportive and relational leadership. Servant leadership, with its inherent emphasis on interpersonal acceptance and developing others, may therefore be particularly suitable for facilitating human integration in the post-merger phase.
These discussions align with the notion that the effectiveness of servant leadership represents a continuum. Sun (2013) suggests that servant leaders are most effective when they are able to “compartmentalize” their leadership identity to suit different situational demands. In the context of organizational change, and in particular in stages where uncertainty is high, they may need to focus more on their identity as a transaction manager, emphasizing task management, organizing, and efficient planning. Conversely, a uniform application of servant leadership across contexts, especially when dealing with difficult situations or dilemmas, may be perceived by employees as a reluctance to make tough decisions or avoid responsibility, ultimately reducing its effectiveness.
Practical implications
Our findings underscore the practical value of servant leadership during organizational change. It effectively reduces emotional exhaustion and turnover intentions, benefits that extend beyond those of charismatic leadership. Organizations undergoing major transitions can therefore benefit from leadership development that fosters servant leadership. However, as heightened job insecurity appears to weaken these benefits, two key implications follow: First, organizations should systematically monitor perceived job insecurity during change. Second, they should invest in strategies to reduce it, such as transparent communication, employee involvement in decision-making, and fair treatment. A more implicit implication of our findings is that employees who continue to experience high job insecurity may require alternative forms of leadership support. This highlights the value of assessing employees' leadership expectations and support needs during change, enabling organizations to tailor leadership development more effectively to diverse employee experiences.
Limitations and future research
Our study has several limitations and raises questions for future research. First, as our findings are based on a single organizational change context, their generalizability is limited. Rather than offering definitive conclusions, the results suggest a possible pattern regarding the reduced effectiveness of servant leadership during change. Future research should replicate this work across diverse change settings and examine additional boundary conditions, such as team dynamics and individual employee resources, to better understand how servant leadership interacts with job insecurity.
Second, the psychological mechanisms underlying the observed moderation effect of job insecurity remain unclear and warrant future investigation. One possibility is that heightened job insecurity erodes trust in servant leaders, as employees may doubt their leader's power or competence, even if they recognize the leader’s supportive intentions. This could reduce receptiveness, limiting the leader's capacity to buffer emotional exhaustion and turnover intentions. Employees experiencing higher job insecurity may also become more resistant to change and its intended outcomes. As a result, they may perceive a servant leader's supportive behavior as cynical or even manipulative, interpreting it as a strategic attempt to elicit performance or loyalty rather than genuine concern (Meuser and Cao, 2022).
Third, measurement limitations should be considered. We used global measures for servant leadership and job insecurity, but multidimensional approaches may provide deeper insights. For instance, Van Dierendonck et al. (2017) distinguish between “leader” (e.g. empowerment, accountability) and “servant” (e.g. prioritizing employees’ needs, humility) behaviors, which may function differently during change. Similarly, job insecurity can be examined beyond general uncertainty to include affective concerns (e.g. job loss anxiety) or specific threats (e.g. transfer, dismissal, wage reductions). Such studies should also assess how insecurity is directly attributed to the M&A process. Finally, chronic illnesses (e.g. diabetes, allergies) may have influenced the sickness absence metric, warranting control in future studies.
Fourth, the cross-sectional nature of our data limits causal inferences and overlooks the dynamic nature of leader behaviors and job insecurity during change. Longitudinal studies could track employees over time, capturing shifts across key change stages. Experience sampling could offer real-time insights into short-term fluctuations in job insecurity and leadership perceptions, especially during critical change moments such as initial announcements or early implementation. These approaches could help explore when and how job insecurity emerges, and whether servant leadership mitigates it, challenging the assumption that insecurity is an inevitable consequence of change.
Finally, as we speculated that directive leadership may become more important as perceived job insecurity increases, future research should test this possibility. This opens the door to exploring the behavioral flexibility of servant leaders. Qualitative interviews with change recipients could offer deeper insight into how they interpret leadership efforts and under what conditions they respond differently to varying leader behaviors.
Conclusion
We have attempted to address a recurring question in the servant leadership literature, namely, whether and to what extent servant leadership is effective under significant organizational change. Our findings suggest that servant leadership is not always beneficial, as heightened job insecurity can offset its positive effects. If further supported by additional empirical research in diverse change settings, our findings can significantly extend scholars' and practitioners' understanding of the boundary conditions that regulate the effectiveness of servant leadership.
Notes
We also tested these models using a latent variable approach with the modsem package in R, replicating the reported findings. Full details are provided in the supplementary materials (Supplement B).
Including charismatic leadership as a covariate in these analyses did not alter the interaction effects of servant leadership, which remained virtually unchanged. When charismatic leadership was tested as the focal predictor moderated by job insecurity, the interaction terms were not significant, neither when examined alone nor when included alongside servant leadership.
The supplementary material for this article can be found online.
