The Curvilinear Effects of Leader Public Service Motivation on Person–Supervisor Fit and Subordinate Emotional Exhaustion: Evidence From Field and Experimental Studies

Abstract

Past research has shed important light on the dark side of individuals’ public service motivation (PSM) in relation to their own well-being. This study turns attention to the role of leader PSM and asks whether it could have a curvilinear relationship with subordinate emotional exhaustion. Drawing from the person–environment (P-E) fit perspective, this study proposes that this curvilinear relationship is mediated by perceptions of person–supervisor (PS) fit and moderated by subordinate PSM. The results from the field and two experimental vignette studies in Thailand provide support for the proposed hypotheses. In particular, higher levels of emotional exhaustion and lower levels of PS fit were observed at the low and high levels of leader PSM, whereas the moderate level of leader PSM was associated with lower emotional exhaustion and higher PS fit. The results from the experimental studies further indicate that individuals with high PSM, in comparison with those with low PSM, perceived higher PS fit with leaders who have moderate to high levels of PSM, in turn, experiencing less emotional exhaustion. These findings highlight the potential dark side of leader PSM, which lends further credence to the too-much-of-a-good-thing effect. Nevertheless, these effects also depend on employees’ PSM levels.

Keywords

leader PSM emotional exhaustion person–supervisor (PS) fit value congruence curvilinear too-much-of-a-good-thing (TMGT) effect experimental vignette the dark side of PSM

Introduction

Past research in public administration (PA) has shed important light on the dark side of public service motivation (PSM)—the desire to do good for others and society (J. L. Perry & Hondeghem, 2008)—indicating that it could have negative consequences for individuals’ well-being (Schott & Ritz, 2017). In particular, this body of work suggests that high-PSM individuals may have too high expectations of their own performance contributions in the face of public-sector constraints (e.g., Giauque et al., 2012; Potipiroon & Ford, 2017; van Loon et al., 2015) and that they may sacrifice too much for society (e.g., Gross et al., 2019; Jensen, Andersen, & Holten, 2019) while reaping too little (Conway et al., 2014) such that they can experience lower psychological and physical well-being.

This research turns attention to the role of leader PSM and asks whether it could have a negative effect on subordinate well-being, rather on their own. Leader PSM, much like employee PSM, can be construed as the extent to which leaders (are perceived to) embrace the beliefs and values regarding the importance of making a difference for society at large, even if it requires personal sacrifice (cf. J. L. Perry & Wise, 1990). While a growing body of work has shed light on the bright side of leader PSM, indicating that it could lead to several positive outcomes both at the organizational (e.g., Coursey et al., 2012) and individual (e.g., Wright et al., 2016) levels, it is important to acknowledge that leaders who are overly dedicated to public service could be perceived less favorably by employees. It is proposed that highly public service–motivated leaders may place excessive emphasis on serving the public rather than on attending to followers’ needs and values such that followers’ well-being could be jeopardized. To date, while we are quite acquainted with stories of hard-driving leaders in the private sector such as Steve Jobs who are known for wearing out employees (Kottler, 2018), much less is known about whether highly public service–motivated leaders can undermine public employees’ well-being.

This study aims to fill this research void by examining whether leader PSM can result in employee emotional exhaustion. In particular, it is proposed that there could be a curvilinear (U-shaped) relationship between leader PSM and subordinate emotional exhaustion. From the too-much-of-a-good-thing (TMGT) effect perspective (Grant & Schwartz, 2011; Pierce & Aguinis, 2013), an extreme level of something positive may have unintended negative consequences when reaching a certain “inflection” point (i.e., threshold) such that the benefits that it once had no longer exist beyond the inflection point. Although leader PSM is an inherently positive construct, this study argues that leaders who work tirelessly for the betterment of the public could be perceived by employees as a source of job stress that undermines their well-being. As far-fetched as this argument may appear, it is important to acknowledge that this type of phenomenon is not uncommon. For example, research has shown that leaders who are too transformational (Molines et al., 2022), too ethical (Stouten et al., 2013), or too smart (Antonakis et al., 2017) tend to be perceived less favorably by followers. These findings suggest that an extreme level of leaders’ PSM may have a perverse negative effect on employees’ well-being as well.

To address this critical issue, this study draws from the person–environment (P-E) fit perspective (Kristof-Brown et al., 2005) to propose that perceptions of person–supervisor (PS) fit (i.e., leader-follower value congruence), or the lack thereof, may provide a possible psychological explanation for the curvilinear relationship between leader PSM and subordinate emotional exhaustion. While there is a general consensus in the extant literature that followers perceive higher leader–follower value congruence when leaders embody altruistic, other-focused values (e.g., Bao & Li, 2019; Hoffman et al., 2011; Jung & Avolio, 2000; Tang et al., 2015), previous research has overlooked the possibility of such value-based leadership may have a nonlinear effect on PS fit when it embodies an extreme level of transcendent values. Furthermore, it is important to acknowledge that not all individuals will perceive the characteristics of their leaders in the same way (Guay et al., 2019). This study seeks to examine the possibility that employee PSM may interact with leader PSM to predict emotional exhaustion via PS fit. This argument is consistent with the P-E fit perspective (Edwards et al., 2006) and also PSM research (e.g., Bottomley et al., 2016; Kroll & Vogel, 2014), which suggests that followers’ PSM can determine how leaders are perceived.

This study contributes to the PSM and public leadership literature by investigating the potential dark side of leader PSM in relation to employee well-being. To date, the negative consequences of public leadership remain largely underexplored (Jensen, Andersen, & Jacobsen, 2019; Vogel & Werkmeister, 2021). While PA scholars have begun to consider the perverse effects of positive leadership styles on employees’ well-being (e.g, Molines et al., 2022), it remains an empirical yet practical question as to whether leader PSM may exert a similar undesirable effect, and, if so, when and how.

Theory and Hypotheses

Leader PSM

PSM can be defined as “an individual’s orientation to delivering service to people with the purpose of doing good for others and society” (Hondeghem & Perry, 2009, p. 6). In line with this, Rainey and Steinbauer (1999) view PSM as a “general altruistic motivation to serve the interests of a community of people, a state, a nation or humanity” (p. 23). Similarly, PSM has been regarded a set of beliefs and values that extend beyond personal and organizational interests toward the benefit of a broader public (Vandenabeele, 2007, p. 547). Thus, individuals with high PSM are generally characterized by the belief that (a) their personal interests are less important than those of the public, (b) meaningful public service is highly important, and (c) standing up for the rights of others is necessary even if it takes personal sacrifice (J. L. Perry et al., 2010; J. L. Perry & Hondeghem, 2008; J. L. Perry & Wise, 1990).

The central focus of this research is on the role of PSM among public-sector leaders. While PA scholars have long been interested in the association between leadership and employees’ PSM (e.g., Hameduddin & Engbers, 2022; Paarlberg & Lavigna, 2010; Vandenabeele, 2014), there is a dearth of research on leader PSM and its potential influence on important outcomes in the public sector. Arguably, the importance of leader PSM may manifest at multiple levels of analysis including the organizational and interpersonal levels. Note however that personal values such as PSM involve deeply held beliefs that exist in a person, and, as such, their manifestation is not necessarily an intentional or a conscious process but may occur inadvertently (Berson et al., 2008).

At the organizational level, PSM can determine how leaders make strategic decisions that impact their organizations. From the upper echelons perspective (Hambrick & Mason, 1984), attributes of organizational leaders (e.g., traits, beliefs, and values) can determine the strategic choice that affects organizational outcomes. Indeed, PSM has also been shown to influence leaders’ views on public policies (Coursey et al., 2012) and decision-making (Stazyk & Davis, 2015). To illustrate, a high-PSM leader may favor investing in public projects that benefit the greater good over those that serve the interest of a select few. Such a decision could have a significant impact not only on the public sector’s performance but also on the public’s welfare.

At the same time, PSM can also influence how leaders manage and lead followers. Decades of research in applied psychology indicates that leaders’ personality can determine their leadership styles (Bono & Judge, 2004) and their perceived effectiveness (Judge et al., 2002, 2009). From the social learning perspective (Bandura, 1977), leaders with high PSM serve as role models in terms of their public service dedication, in turn inspiring followers to display similar positive behaviors (Paarlberg & Lavigna, 2010; Wright et al., 2016). Indeed, previous research indicates that the virtuous influence of public leadership represents the informal institutional socialization process, which serves as a bridge that facilitates the transmission and maintenance of important institutional values (J. L. Perry & Vandenabeele, 2008; Vandenabeele, 2007). Furthermore, public-service–motivated leaders are known to transform the focus of their followers from self-interests toward a broader collective goal and to inspire them to perform above and beyond the call of duties (Paarlberg & Lavigna, 2010; Shamir et al., 1993).

Existing, albeit limited, empirical evidence provides support for the above arguments. For example, in a study of civil servants in Belgium, Vandenabeele (2014) reported that leaders’ PSM (i.e., the promotion of public service values) can significantly alter employees’ PSM by fulfilling their basic psychological needs. In another study, Wright et al. (2016) found that high-PSM leaders are more likely to be perceived by followers as ethical leaders, in turn triggering followers’ PSM and their willingness to report unethical behavior. More recently, in the management literature, Tekleab et al. (2021) also showed that leaders with high prosocial motivation influence their followers to engage in CSR practices, which leads to higher CSR performance.

Given the above empirical evidence, it is not a surprise that leader PSM is generally viewed as a positive force that almost certainly results in positive outcomes. As J. L. Perry and Hondeghem (2008, p. 8) noted, “if public servants are general altruists, then we will be inclined to rely on them to do good at all times.” However, from the subordinates’ point of view, high-PSM leaders may be perceived as a source of job stress that undermines their well-being. This study brings attention to emotional exhaustion, which is a core dimension of job burnout (Maslach & Jackson, 1981) and an important indicator of well-being (Maslach et al., 2001). Emotional exhaustion describes the extent to which individuals experience constant depletion of mental, physical, and emotional energy and resources that is generally a result of prolonged and extensive exposure to job-related stress (Maslach & Jackson, 1981). As discussed subsequently, this study draws upon the P-E fit theory (Kristof-Brown et al., 2005) as a theoretical lens through which to understand how excessive levels of leader PSM can have an undesirable effect on subordinate emotional exhaustion via perceptions of (mis)fit with the leader.

The P-E Fit Perspective

According to P-E fit theory (Kristof-Brown et al., 2005), the match or similarity between individuals’ characteristics and the work environment (e.g., jobs, groups, and organizations) can have a significant impact on their attitudes and behaviors. P-E fit can be attained in two important ways: (a) supplementary fit and (b) complementary fit (Kristof-Brown et al., 2005). Supplementary fit occurs when the work environment and individuals possess similar values or goals, whereas complementary fit pertains to a situation in which individuals’ unmet needs are satisfied by the work environment and vice versa (Kristof-Brown et al., 2005). Both types of fit have been shown to predict several employee outcomes including the intent to stay, job satisfaction, and organizational identification (Cable & Edwards, 2004) as well as citizenship behavior (Guan et al., 2011).

However, when individuals perceive a lack of correspondence with the work environment, it can lead to various psychological and physical effects including emotional exhaustion (Siegall & McDonald, 2004) and depression (Lamiani et al., 2018). According to the P-E perspective on stress (Edwards & Cooper, 1990), stress can be characterized as the incompatibility between the person’s characteristics (e.g., values, goals, and abilities) and the work environment (e.g., demands and supplies). This argument is consistent with other theoretical perspectives on stress, which indicate that stress can occur as a result of the perceived demands from the work environment that exceed a person’s capabilities or resources (McGrath, 1976) or when the work environment fails to supply what the person needs or values (French et al., 1982). These P-E fit perspectives on stress serve as a theoretical lens through which to understand the effects of leader PSM.

The Curvilinear Effects of Leader PSM on PS Fit and Subordinate Emotional Exhaustion

This study calls upon the above PE fit perspective to explain how leader PSM can exert a curvilinear effect on subordinate emotional exhaustion. As discussed, this study is interested in probing the TMGT effects of leader PSM. That is, it is expected that the beneficial effects of leader PSM on follower outcomes may begin to diminish once reaching a certain critical point. The paragraphs below discuss how and why leader PSM can exert a nonlinear effect on perceptions of fit with the leader and subordinate emotional exhaustion.

While there are different types of P-E fits including person–organization (PO) fit, person–job (PJ) fit, or person–group (PG) fit (Kristof-Brown et al., 2005), this research focuses on person–supervisor (PS) fit, which is most relevant when considering leader-follower relationships. PS fit can be achieved both supplementarily or complementarily (Kristof-Brown et al., 2005). Indeed, PS fit has been conceptualized in terms of goal congruence (Witt, 1998) and personality similarity (Guay et al., 2019; Schaubroeck & Lam, 2002) as well as needs–supplies fit (Kroll & Vogel, 2014) and needs–values fit (Marstand et al., 2017). Given the central focus on leaders’ emphasis on public service values, this research focuses on the supplementary fit, which is defined as perceived value congruence between the leader and the subordinate in a dyadic relationship (Kristof-Brown et al., 2005). PS fit has been shown to be influenced by several leadership styles including transformational leadership (Hoffman et al., 2011; Jung & Avolio, 2000), ethical leadership (Tang et al., 2015), moral leadership (Bao & Li, 2019), and servant leadership (Safavi & Bouzari, 2020). Unfortunately, previous research has overlooked the possibility that high levels of such value-based leadership may relate curvilinearly to PS fit and work outcomes.

This study proposes that excessive levels of leaders’ PSM may be perceived by followers as a job demand that leads to a decline in PS fit and their well-being. Unlike other leadership styles such as transformational leadership, ethical leadership, or servant leadership, leaders with high PSM are, by definition, those who put a primary emphasis on serving the public rather than their followers (Coursey et al., 2012; Pedersen, 2014; Ritz, 2015). In particular, leaders high in PSM encourage followers to rise above their personal self-interest to achieve important public goals and missions (Vandenabeele, 2014), which may include equal treatment and equity, responsiveness and representation, and the protection of individual rights (Stazyk & Davis, 2015). In so doing, they emphasize to employees the importance of showing compassion to fellow citizens and the need to forgo self-interests (Vandenabeele, 2014). Furthermore, leaders who are high in PSM are known to put in more work effort and work longer hours than typical leaders (Pedersen, 2014; Ritz, 2015), and, as such, they may demand the same level of work contributions from followers.

Accordingly, this study argues that leader PSM could have an undesirable effect on employees’ well-being to the extent that the leader’s societal focus is perceived as incompatible with employees’ values and goals. In particular, excessive levels of leader PSM could lower PS fit as the leader’s dedication to public service may signal to employees that their values and work styles are different. As noted by J. L. Perry and Wise (1990, p. 371), “individuals motivated by public service may carry their commitment beyond reasonable boundaries.” In fact, perceived pressure to fulfill the leader’s expectations can lead to perceptions of role overload and strain because employees may fear falling short of the leader’s expectations (Syrek & Antoni, 2014). While this type of job demand—also termed a “challenge” job demand—can encourage mastery, learning, and personal growth, it can also contribute to emotional exhaustion (Crawford et al., 2010).

Empirical findings from the field provide support for this central argument. For example, leaders’ achievement striving has been shown to lead to poorer work–life balance among team members (Robertson et al., 2014). Also, excessive levels of leaders’ conscientiousness (i.e., the tendency to exhibit self-discipline and perfectionism) have been shown to result in a nonlinear increase in subordinates’ emotional exhaustion (S. J. Perry et al., 2011). A recent study in PA by Molines et al. (2022) also reported that high levels of transformational leadership lead to a nonlinear increase in the police’s emotional exhaustion via a decline in the quality of leader–member exchange (LMX). Similarly, Nielsen and Daniels (2016) showed that under transformational leaders, employees feel compelled to show up at work even when ill. In another study, Diebig et al. (2016) showed that leaders’ visionary behaviors significantly increase subordinate’ cortisol, an objective indicator of stress. These findings suggest that highly committed leaders—a reasonable assumption of high-PSM leaders—may anchor their performance expectations at a comparatively higher level than others such that the desire to meet or surpass their aspirations could result in their expectations being unduly imposed on their own subordinates, hence, putting the latter’s well-being at risk.

However, under leaders with low PSM, employees may also experience low PS fit and high emotional exhaustion, albeit for different reasons. In particular, low-PSM leaders are known to be more concerned about personal agenda than the pursuit of important public service goals (Pedersen, 2014; J. L. Perry & Wise, 1990), and it is expected that they may be evaluated less favorably by employees. Indeed, it has been shown that individuals whose values are not aligned with public service dedication tend to receive lower performance ratings (Alonso & Lewis, 2001; Naff & Crum, 1999). It is thus plausible that employees will perceive a lower fit with low-PSM leaders. Furthermore, low-PSM leaders may fail to set a clear example of how to perform important public service. This lack of role modeling from the leader could result in poorer PS fit (Safavi & Bouzari, 2020), in turn, leading to a decline in their well-being (Enwereuzor et al., 2020). In fact, low-PSM leaders may even be perceived as a “hindrance” job demand to the extent that they prevent employees from accomplishing valued tasks (Crawford et al., 2010).

For these reasons, it is expected that those who work under leaders with high and low levels of PSM could experience high levels of emotional exhaustion and lower PS fit. In contrast, leaders who show moderate levels of PSM are expected to be perceived most favorably. Following this line of reasoning, this study proposes that leader PSM will have a direct curvilinear relationship with both subordinate emotional exhaustion and PS fit. At the same time, PS fit is expected to mediate the curvilinear relationship between leader PSM and emotional exhaustion. This argument is based on the insights from previous research, which indicates that PS fit is a more proximal predictor of work outcomes than leadership styles (e.g., Guay et al., 2019; Jung & Avolio, 2000; Safavi & Bouzari, 2020; Tang et al., 2015). This leads to the following hypotheses.

Hypothesis 1: Leader PSM has a direct curvilinear (U-shaped) relationship with subordinate emotional exhaustion such that low and high levels of leader PSM are associated with higher subordinate emotional exhaustion than the moderate level of leader PSM.

Hypothesis 2: Leader PSM has a direct curvilinear (inverted U-shaped) relationship with PS fit, such that low and high levels of leader PSM are associated with lower PS fit than the moderate level of leader PSM.

Hypothesis 3: PS fit mediates the curvilinear relationship between leader PSM and subordinate emotional exhaustion.

The Moderating Role of Employee PSM

While the idea of perceived fit in the preceding section (PS fit) focuses on a direct assessment of compatibility between the leader and the subordinate, which is based on the focal person’s comparison of the leader to certain psychological or social standards (Edwards et al., 2006), this section aims to shed light on the possibility that employee PSM may interact with leader PSM. Such an interactive effect is also known as “objective fit,” which has been shown to predict PS fit (Guay et al., 2019). The central argument for the moderating role of employee PSM is that there could be subtle differences among subordinates with varying levels of PSM in terms of how they evaluate their leaders. Indeed, subordinates are in a unique position to judge the leader’s effectiveness and their PSM is especially relevant when considering leader PSM (see Kristof-Brown et al., 2005 for a discussion on commensurate measurement).

From the similarity-attraction perspective (Byrne, 1971), which is a central paradigm for explaining the dynamics of leader-follower value congruence (Guay et al., 2019; Tang et al., 2015), employees’ preferences for and reactions to certain leadership styles are largely determined by the extent to which they share similar attributes and values with the leader. Although excessive levels of leader PSM are expected to be perceived less favorably by employees, this study proposes that high-PSM leaders will gain greater acceptance from high-PSM employees in comparison to those with low PSM. In particular, individuals with high PSM define their public service roles more broadly than just complying with a bare minimum (Potipiroon & Faerman, 2016), and, as such, they should view the characteristics of high-PSM leaders as congruent with their personal values and also as the opportunity to act on their prosocial values (Kroll & Vogel, 2014).

Accordingly, it is expected that high-PSM employees will identify less with low-PSM leaders. On one hand, tasks allocated by low-PSM leaders could be mundane and far removed from the true source of high-PSM employees’ motivation. On the contrary, the contributions of high-PSM employees may not be recognized or valued by low-PSM leaders (cf. Christensen et al., 2013). As a result, high-PSM employees may feel hurt by the leader’s failure to honor their work contributions (cf. Conway et al., 2014). On the contrary, individuals with low PSM have a weak motivation to serve the public and, as a result, they may perceive lower PS fit and experience higher emotional exhaustion under high-PSM leaders. Indeed, low-PSM individuals have been shown to be more prone to experience role overload and stress under high work-demand situations (Shim et al., 2017). This leads to the following hypothesis.

Hypothesis 4: The effect of leader PSM on PS fit is moderated by subordinate PSM such that those with high PSM will perceive higher PS fit with high-PSM leaders.

Combining the ideas discussed previously, this study also proposes that the indirect curvilinear effect of leader PSM on emotional exhaustion via the mediating role of PS fit will be moderated by employees’ PSM. This argument is based on Edwards et al.’s (2006) work and also emerging empirical evidence (Guay et al., 2019), which indicate that objective fit (i.e., the interactive effect of leaders’ and employees’ characteristics) serves as a more distal predictor of work outcomes than perceived fit (i.e., PS fit). This leads to the final hypothesis.

Hypothesis 5: The indirect curvilinear effect of leader PSM on subordinate emotional exhaustion via the mediating role of PS fit is moderated by subordinate PSM.

Study 1: Field Research Methodology

Sample and Data Collection Procedures

In Study 1, field data were collected from employees working in Subdistrict Administrative Organizations (SAOs) in Thailand, which are the primary local administrative government unit with a nation-wide presence. SAO employees work closely under the supervision of Chief SAO Officers (CSO) who report directly to mayors and city councils. Survey packets containing one CSO questionnaire and ten employee questionnaires were mailed to 50 SAOs located in three LMX provinces in the central region of the country. Each survey packet contained a postage-paid return envelope and detailed instructions on how to administer the questionnaires. The CSOs were asked to assess their own PSM, whereas employee respondents were asked to assess their emotional exhaustion and provide information on their gender, age, education, and dyad tenure. Leaders’ ratings of their own PSM have been recommended by previous research (e.g., Ritz, 2015), which alleviates concerns about common method bias (CMB; Podsakoff et al., 2003). After a period of 2 months, a total of 214 surveys from 41 SAOs were returned to the author, with a response rate of 46.8% (employees) and 82% (CSOs). In particular, 71% of the employees were female. About 82% of them were college-level graduates. Their average age was 35.74 years (SD = 7.83), and the average dyad tenure with the CSO was 4.86 years (SD = 3.85). As for the CSOs, 48.8% of them were male and their average age is 44 years old (SD = 10.03).

Measurement

Unless stated otherwise, the scales used for this research were based on a 5-point format where 1 = strongly disagree and 5 = strongly agree. Leader PSM (α = .70) was assessed by having the CSOs assess their own PSM. This measurement was based on the five-item global measure adapted from J. L. Perry’s (1996) 24-item scale, which has been used in several PSM studies (e.g., Belle, 2013; Wright et al., 2013). Sample items include “Meaningful public service is very important to me” and “Making a difference in society means more to me than personal achievements.” PS fit (α = .83) was measured using the three-item subjective fit scale by Cable and DeRue (2002), in which the referent was changed from “organization” to “supervisor.” Sample items include “My supervisor’s values provide a good fit with the things that I value in life” and “The things that I value in life are very similar to the things that my supervisor values.” This was based on a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree). Emotional exhaustion (α = .88) was measured with five items derived from the Maslach Burnout Inventory (Maslach & Jackson, 1981). Sample items include “I feel emotionally drained from my work” and “I feel burned out from my work.” A confirmatory factor analysis (CFA) showed that a three-factor model had an excellent fit to the data (χ² = 91.75, df = 62, p < .001; comparative fit index [CFI] = .97; Tucker–Lewis index = .96; root mean square error of approximation [RMSEA] = .05). All the analyses that follow also controlled for employee age, gender (male = 0, female = 1), education (0 = below bachelors’ degree, 1 = a bachelor’s degree, 2 =a master’s degree, and 3 = a doctoral degree) and dyad tenure. Bivariate correlations of the study variables, their means, standard deviations (SDs), and Cronbach’s alpha values (α) are reported in Table 1.

Table 1.

Bivariate Correlations, Means, Standard Deviation, and Reliabilities (Study 1).

Variables	M	SD	1	2	3	4	5	6	7
1. Emotional exhaustion	2.61	0.83	(.89)
2. PS fit	5.46	1.13	−.40**	(.83)
3. Leader PSM	4.11	0.41	−.04	.05	(.70)
4. Leader PSM squared	17.01	3.33	.16*	−.21**	−.33**
5. Gender	0.78	0.41	−.02	−.08	.09	−.01	—
6. Age	35.75	7.83	−.20**	.12	−.14*	.00	−.16*	—
7. Education	1.09	0.67	.03	.02	.12	.13*	.05	.04	—
8. Dyad tenure	58.32	46.22	−.14*	.07	−.05	−.02	.00	.39**	−.00

Note. N = 214 in 41 SAOs. Tenure in months. Alpha values are displayed on the diagonal. PS = person–supervisor; PSM = public service motivation; SAO = Subdistrict Administrative Organizations.

p < .05. **p < .01.

Analytic Procedures

As employees within the same SAOs were likely to share similar levels of PS fit and emotional exhaustion, the nonindependence assumption of Ordinary Least Square could be violated. Thus, the degrees of nonindependence in these variables were calculated using a one-way analysis of variance (ANOVA) and the intraclass correlation coefficient (ICC). The result showed that both PS fit and emotional exhaustion had significant between-organization variance (F = 2.15, p < .001, ICC = 18.00; F = 1.71, p < .01, ICC = 11.40, respectively). This suggests that it is more appropriate to use multilevel modeling to conduct the analyses. Hierarchical Linear Modeling (Raudenbush & Bryk, 2002) in STATA Version 13.0 (StataCorp, 2012) was used for all the analyses. In particular, leader PSM was modeled as an organization-level predictor that has a cross-level influence on employees’ PS-fit and emotional exhaustion.

For the main regression analyses, a random-intercept and fixed-slope model was estimated in which leader PSM was entered as a Level 2 predictor of the random intercepts along with all the Level 1 control variables. The purpose of this procedure was to test a direct cross-level effect of leader PSM on employee emotional exhaustion and PS fit. Then, to test the curvilinear effects of leader PSM, its quadratic term (leader PSM squared) was entered in the regressions. Finally, to test the indirect curvilinear effect of leader PSM on employee emotional exhaustion via PS fit, all the terms (i.e., leader PSM, leader PSM squared, and PS fit) were entered in the regression. Note that the Level 2 terms were all grand-mean centered (Hofmann & Gavin, 1998) to reduce the nonessential multicollinearity between the linear and the quadratic terms of leaders’ PSM (Aiken & West, 1991; Dawson, 2013). To test the indirect curvilinear effect, Hayes and Preacher’s (2010) MEDCURVE Macro was performed in Mplus Version 8.2 (Muthén & Muthén, 1998–2018).

Hypothesis Testing

As shown in Table 2, the results revealed that the direct effects of leader PSM on emotional exhaustion and PS fit were nonsignificant (Model 1: b = −.12, p > .10; Model 3: b = .17, p > .10). However, the quadratic term of leader PSM was significant in the predicted direction on both emotional exhaustion (Model 2: b = .51, p < .05) and PS fit (Model 4: b = −.95, p < .01). These results suggest that leader PSM has a curvilinear relationship with both emotional exhaustion (U-shaped) and PS fit (inverted U-shaped). The results are portrayed in Figures 1 and 2.

Table 2.

Hierarchical Linear Modeling (HLM) (Study 1).

	Model 1	Model 2	Model 3	Model 4	Model 5
Independent variables	Emotional exhaustion	Emotional exhaustion	PS fit	PS fit	Emotional exhaustion
Intercept	3.45***	3.33***	5.03***	5.03***	4.88***
Level 1 (Controls)
Age	−.02 (.00)**	−.02 (.00)**	.01 (.01)	.01 (.01)	−.01 (.00)
Gender	−.14 (.13)	−.14 (.13)	−.17 (.18)	−.17 (.18)	−.19 (.12)
Education	.05 (.08)	.06 (.08)	−.01 (.11)	−.01 (.11)	.04 (.07)
Dyad tenure	−.00 (.00)	−.00 (.00)	.00 (.00)	.00 (.00)	−.00 (.00)
Level 1 (Mediator)
PS fit	—	—	—	—	−.31 (.04)***
Level 2 (Cross-level)
Leader PSM	−.12 (.16)	−.01 (.16)	.17 (.24)	−.03 (.24)	−.02 (.15)
Leader PSM squared	—	.51 (.24)*	—	−.95 (.36)**	.20 (.23)
R² Level 1	.054	.079	.026	.072	.238
R² Level 2	.094	.175	.046	.175	.273

Note. N = 214 in 41 organizations. Values in parentheses are standard errors. R² was calculated with a formula presented by Snijders and Bosker (1999), using the command “mltrsq” in STATA. Reported coefficients are unstandardized. PS = person–supervisor; PSM = public service motivation.

p < .05. **p < .01. ***p < .001

Figure 1.

The Curvilinear (U-Shaped) Relationship Between Leader PSM and Subordinate Emotional Exhaustion (Study 1).

Figure 2.

The Curvilinear (Inverted U-Shaped) Relationship Between Leader PSM and PS Fit (Study 1).

In terms of the indirect curvilinear effect of leader PSM, the results in Model 5 showed a significant effect for PS fit on emotional exhaustion (b = −.31, p < .001), whereas the effect of leader PSM and its quadratic term were non-significant (b = −.02, p > .80; b =.20, p > .30, respectively). This provides support to Hypothesis 3. Furthermore, the MEDCURVE results showed that PS fit mediated the effects of low-PSM leaders (θ = −.22, p < .01, 95% confidence interval [CI]: [−.50, −.03]) and high-PSM leaders (θ = .19, p < .05, 95% CI [.00, .45]), whereas the effect of moderate-PSM leaders was not significantly mediated (θ = −.01, p > .10, 95% CI [−.15, .11]).

Studies 2 and 3: Experimental Vignette Research Method

Results from Study 1 provide preliminary evidence for the indirect curvilinear effect of leader PSM on subordinate emotional exhaustion. Studies 2 and 3 aim to replicate and extend this important finding by examining the moderating role of subordinate PSM (Hypotheses 4 and 5). Furthermore, Studies 2 and 3 aim to address two important limitations of Study 1. First, while the strength of Study 1 lies in the use of multiple sources of field data, its internal validity (i.e., a causal inference about the proposed relationships) may be lacking. Second, leaders’ self-ratings of their own PSM in Study 1 may introduce social desirability bias (i.e., their PSM levels may be overreported; Podsakoff et al., 2003). To address these inherent limitations, Studies 2 and 3 were conducted using the experimental vignette methodology (EVM; Aguinis & Bradley, 2014). In the context of this research, the EVM involves presenting hypothetical scenarios pertaining to different levels of leader PSM. Specifically, a between-subjects design was used in which three different vignettes (i.e., low-, moderate-, and high-PSM leaders) were presented to each group of participants.

Samples and Experimental Procedures

In Study 2, a total of 120 undergraduate PA students from a large public university in Thailand were invited to participate in the study. Undergraduate student samples have been successfully used in previous experimental research in PA (e.g., Bouwman & Grimmelikhuijsen, 2016; Christensen & Wright, 2018). Four participants were removed from the analyses because their responses involved a number of missing values. Of the participants (N = 116), 81.9% were female and were on average 20 years old (SD = 1.10). Study 3 was conducted to enhance the external validity of Study 2 by recruiting a sample of working adults who are familiar with the public sector. Specifically, 48 working professionals enrolled in the Master of Public Administration (MPA) Program from the same university were invited to take part in the study. Of these participants, 75% were female and were on average 36 years old (SD = 6.15). All of these participants had worked in the public sector, with an average organizational tenure of 7.50 years (SD = 4.30).

Two weeks before the experiments were conducted, all the participants were asked to rate their own PSM. During the experiments, the author randomly assigned the participants to one of three conditions of leader PSM (i.e., low [coded as 0], moderate [coded as 1], and high [coded as 2]). The participants were asked to immerse themselves in a situation where they were working for a district chief named “Mr.Wanchai.” Although the participants were likely aware of the significance of district chiefs’ roles in the local context, they were provided with baseline information to rule out the possibility that any effects are due to differences in the lack of knowledge about the nature of the district chief’s work. This procedure was recommended by Aguinis and Bradley (2014). The baseline information was provided as follows:

Imagine that you are working for a district chief officer named Mr.Wanchai. As you probably know, district chiefs serve as the important point of contact between the central government and the people. Appointed by the Ministry of Interior, district chiefs are in charge of enforcing a wide variety of laws while also implementing major government programs in their jurisdiction that serve two specific goals: (1) to alleviate the suffering and (2) to promote the well-being of all Thai people. As the chief magistrate of the district, their responsibilities include but are not limited to poverty reduction, security enforcement . . .[information omitted to save space]. Their work is mostly on the ground helping people, rather than in the office.

After reading the baseline information, the participants in each of the three conditions were asked to read one of the three vignettes and assess Mr. Wanchai’s PSM. The manipulation of leader PSM (i.e., the design of the vignettes) was carefully crafted using the general characterizations of public service–motivated individuals (e.g., Broekema et al., 2019; Coursey et al., 2012; J. L. Perry & Wise, 1990; Ritz et al., 2016; Schott & Ritz, 2017; Vandenabeele, 2007). The characterization of the leader with high PSM served as an anchor upon which the characterizations of leaders with low to moderate levels of PSM were further contextualized. This approach is consistent with previous experimental vignette research in the leadership literature (e.g., Stouten et al., 2013). The details of the three vignettes were provided as follows.

High-PSM leader condition

In the high-PSM leader condition, the district chief is described as a person who is highly dedicated to public service, even at the expense of his self-interest (in terms of the time and money) (Brewer, 2003; Houston, 2006):

“Mr.Wanchai is known for his dedication to helping the people in his district. In fact, no other district chiefs can compare to him in terms of his public service commitment. He spends most of his time on the ground helping the needy, working day and night to achieve his public service goals. To Mr. Wanchai, public service must take precedence over anything else including his personal matters. In fact, he goes to great lengths to help the needy, even if it costs him his own money.”

Moderate-PSM leader condition

In the moderate-PSM leader condition, the district chief was described as a person who is still dedicated to public service but not to the point of jeopardizing his self-interest and personal goals:

“Mr.Wanchai is known for his work in helping the people in his district. His public service is recognized by the local people. He spends a lot of time on the ground helping the needy while also making sure that he makes time for his personal matters. To Mr.Wanchai, it is important to sacrifice his personal time to public service to the extent that circumstances call for it.”

Low-PSM leader condition

In the low-PSM leader condition, the district chief was described as a person who cares more about pursuing his personal achievements than public service dedication (J. L. Perry, 1996):

“Mr.Wanchai is known for being uninterested in helping people in his district. He spent most of his time in the office, meeting with important people so that his own career could be advanced. Oftentimes, he abandons his work while also refusing to meet with the locals who seek help from him.”

Measurement

All the measurement items were assessed using a 5-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). Both the leader’s PSM and the participants’ PSM were assessed using the same five-item PSM scale as in Study 1. Note that the items for the leader’s PSM were reworded to reflect the participants’ perceptions of Mr. Wanchai’s PSM. Sample items include “Meaningful public service appears to be very important to Mr. Wanchai” and “Making a difference in society seems to mean more to Mr. Wanchai than his personal achievements.” Employees’ assessment of leaders’ (prosocial) motivation has been employed in previous research (e.g., Tekleab et al., 2021). The participants were then asked to assess their PS fit with Mr. Wanchai. This measure was based on the same three-item scale as in Study 1 but the referent was changed to “Mr. Wanchai.” Finally, the participants were asked to assess the level of emotional exhaustion that they expect to experience from working for Mr. Wanchai. This was based on the same five-item measure in Study 1. This approach—although not a perfect way to measure emotional exhaustion—has been used successfully in several previous vignette experimental studies in applied psychology (e.g., Cameron et al., 2016; Clinton & Pollini, 2021; García-Romero & Martinez-Iñigo, 2021; Martínez-Íñigo et al., 2015). In terms of the measurement quality, the CFA results from Study 2 indicated that the proposed four-factor model had a good fit to the data (χ² = 219.00, df = 129, p < .001; CFI = .96; TLI = .95; RMSEA = .08). As for Study 3, the results indicated that the four-factor model also had an acceptable fit to the data (χ² = 201.13, df = 129, p < .001; CFI = .91; TLI = .90; RMSEA = .09). Bivariate correlations among the variables, their means, SDs and Cronbach’s αs are presented in Table 3 (Study 2) and Table 4 (Study 3).

Table 3.

Bivariate Correlations, Means, Standard Deviation, and Reliabilities (Study 2).

Variables	M	SD	1	2	3	4
1. Emotional exhaustion (Time 2)	2.57	1.12	(.94)
2. PS fit (Time 2)	3.08	1.14	−.69**	(.95)
3. Leader PSM (Time 2)	3.69	1.40	−.56**	.73**	(.93)
4. Employee PSM (Time 1)	3.58	0.67	−.09	.22*	.14	(.83)

Note. N = 116. Alpha values are displayed on the diagonal. Time 1 = 2 weeks prior to the experiment; PS = person–supervisor; PSM = public service motivation.

p < .05. **p < .01.

Table 4.

Bivariate Correlations, Means, Standard Deviation, and Reliabilities (Study 3).

Variables	M	SD	1	2	3	4
1. Emotional exhaustion (Time 2)	3.12	1.02	(.92)
2. PS fit (Time 2)	2.87	1.05	−.75**	(.92)
3. Leader PSM (Time 2)	3.74	1.34	−.44**	.56**	(.94)
4. Employee PSM (Time 1)	3.35	0.66	−.31*	.33*	.25	(.75)

Note. N = 48. Alpha values are displayed on the diagonal. Time 1 = 2 weeks prior to the experiment. PS = person–supervisor; PSM = public service motivation.

p < .05. **p < .01.

Random Assignment and Manipulation

The results from Study 2 indicated that the participants across the three conditions did not differ in their grade point average, F(2, 113) =.409, p > .60, or PSM levels, F(2, 113) = .053, p > .90. There was also a good mix of men and women in each of the three conditions, χ²(2) = 3.737, p >.10. This suggests that the random assignment of the participants was successful. In terms of the manipulation of leader PSM, the results showed that the participants’ ratings of leader PSM differed significantly across the three conditions, F (2, 113) = 253.65, p < .001. Post hoc tests also showed that the low-PSM leader condition (M = 1.57; SD = .90) was perceived to be lower on PSM than the moderate-PSM leader condition (M =4.29; SD =.52, p < .001) and the high-PSM leader condition (M = 4.65; SD = .32, p < .001). The moderate-PSM leader condition was also found to be significantly lower on PSM than the high-PSM leader condition, p < .01. These results indicate that the manipulation was successful.

As for Study 3, the results also indicated that there was a good mix of men and women in each of the three conditions, χ²(2) = 5.21, p >.26. The participants also did not differ in their age, F(2, 45) =.980, p > .39, organizational tenure, F(2, 45) =.519, p > .60, or PSM levels, F(2, 45) = 1.902, p > .15. This suggests that the random assignment of the participants was successful. As for the manipulation of leader PSM, the participants’ ratings of leader PSM differed significantly across the three conditions, F(2, 45) = 151.51, p < .001. Post hoc tests also showed that the low-PSM leader condition (M = 1.61; SD = .70) was perceived to be lower on PSM than the moderate-PSM leader condition (M =4.22; SD =.54, p < .001) and the high-PSM leader condition (M = 4.65; SD = .29, p < .001). The moderate-PSM leader condition was also found to be significantly lower on PSM than the high-PSM leader condition, p < .05. These results indicate that the manipulation was successful.

Analytic Procedures

All the analyses were conducted using the General Linear Model (GLM) for univariate analyses (see Stouten et al., 2013 for a similar procedure). The indirect curvilinear effect was tested using Hayes and Preacher’s (2010) MEDCURVE Macro in Mplus Version 8.2 (Muthén & Muthén, 1998/2018). To test the curvilinear moderated mediation effects (i.e., the instantaneous indirect effects of leader PSM on employee emotional exhaustion via PS fit at low and high levels of employee PSM), this study followed the combined statistical procedures of Hayes and Preacher’s (2010) MEDCURVE and Edwards and Lambert’s (2007) bootstrapping-based approach with 10,000 samples (see also Jiang et al., 2022 for a detailed procedure).

Study 2: Hypothesis Testing

To test Hypothesis 1 (i.e., the curvilinear [U-shaped] relationship between leader PSM and employee emotional exhaustion), a between-subjects ANOVA was performed. As shown in Table 5, the results revealed a significant effect of leader PSM, F(2, 113) = 64.14, p < .001, η² = 0.53. Post hoc analysis also showed that the low-PSM leader condition was more strongly associated with employee emotional exhaustion (M = 3.80; SD = .89) than was the moderate-PSM leader condition (M = 1.80, SD = .62, p < .001) as well as high-PSM leaders (M = 2.70, SD = .88, p < .001). The moderate- and high-PSM leader conditions also differed significantly, p < .001. The polynomial contrast test showed that the effect of leader PSM on emotional exhaustion supports a curvilinear (U-shaped) pattern (95% CI [.95, 1.41]). This offers support for Hypothesis 1.

Table 5.

Means and ANOVA Results (Study 2).

DV	Low-PSM leader		Moderate-PSM leader		High-PSM leader		F-test	p-value	η²
DV	M	SD	M	SD	M	SD	F-test	p-value	η²
Exhaustion	3.80	.89	1.80	.62	2.70	.88	64.14	<.001	.53
PS fit	1.64	.63	3.80	.72	3.25	.83	83.69	<.001	.59

Note. ANOVA = analysis of variance; PS = person–supervisor; DV = dependent variables.

Then, to test Hypothesis 2 (i.e., the curvilinear [inverted U-shaped] relationship between leader PSM and PS fit), a between-subjects ANOVA was also conducted. As shown in Table 5, the results showed that leader PSM had a positive effect on PS fit, F(2, 113) = 83.69, p < .001, η² = 0.59. Post hoc tests also showed that the moderate-PSM leader condition resulted in higher PS fit (M = 3.80, SD = .72) than the high-PSM leader condition (M = 3.25, SD = .83, p < .01) as well as the low-PSM leader condition (M = 1.64, SD = .63, p < .001). The high- and low-PSM leader conditions also differed significantly, p < .001. The polynomial contrast test also showed that the effect of leader PSM on PS fit supports a curvilinear (inverted U-shaped) pattern (95% CI [−1.33, −.88]). This supports Hypothesis 2.

To test Hypothesis 3 (i.e., the mediating effect of PS fit), an analysis of covariance (ANCOVA) was conducted. PS fit was treated as a covariate in the ANCOVA. As can be seen in Table 6, the results showed a significant direct effect for PS fit, F(1, 112) = 13.702, p < .001, η² = 0.11, whereas the effect of leader PSM remained significant, F(2, 113) = 13.628, p < .001, η² = 0.19. To test the significance of the indirect curvilinear effect of leader PSM, the MEDCURVE procedure was performed. The results showed that PS fit mediated the effects of the low-PSM leader condition (θ = −1.87, p < .001, 95% CI [−2.33, −1.40]), the moderate-PSM leader condition (θ = −.50, p < .001, 95% CI [−.66, −.33]), and the high-PSM leader condition (θ = .87, p < .001, 95% CI [.50, 1.23]). This provides support to Hypothesis 3.

Table 6.

Main Results (Study 2).

IV	ANCOVADV = emotional exhaustion			3 × 2 ANOVADV = PS fit
IV	F-test	p value	η²	F-test	p value	η²
PS fit	13.70	<.001	.11	—	—	—
Leader PSM^a	13.62	<.001	.19	92.34	<.001	.62
Employee PSM^b	—	—	—	3.41	<.10	.03
Leader PSM × employee PSM	—	—	—	4.72	<.05	.08

Note. ANCOVA = analysis of covariance; ANOVA = analysis of variance; PS = person–supervisor; PSM = public service motivation; IV = independent variables.

Leader PSM was coded as 0, 1, 2. ^bEmployee PSM was coded as 0 and 1, based on the median split.

Next, to test Hypothesis 4 (i.e., the moderating effect of employee PSM on PS fit), a 3×2 between-subjects ANOVA was conducted. Employee PSM was categorized as low (coded as 0) and high (coded as 1), based on the median split (median = 3.40). As can be seen in Table 6, the results showed that the main effect of leader PSM was significant, F(2, 110) = 92.34 p < .001, η² = 0.62, whereas the main effect of employee PSM was marginally significant, F(2, 110) = 3.41 p < .05, η² = 0.03. The results further showed that the interaction between leader PSM and employee PSM was significant as predicted, F(2, 110) = 4.72, p < .05, η² = 0.08. Tests of simple effects further indicated that the effect of leader PSM on PS fit was significant for both high-PSM employees, F(2, 55) = 51.10, p < .001, η² = 0.65, and low-PSM employees, F(2, 55) = 43.86, p < .001, η² = 0.61. Nevertheless, as can be seen in Table 7 and Figure 3, the observed relationships between leader PSM and PS fit were different between the two groups. At the high level of leader PSM, employees with high PSM perceived higher PS fit (M =3.58, SD = .19) than those with low PSM (M =2.89, SD = .14), F(2, 110) = 8.03, p < .01, η² = 0.07). Similarly, at the moderate level of leader PSM, employees with high PSM perceived higher PS fit (M =4.01, SD = .15) than those with low PSM (M =3.60, SD = .11), F(2, 110) = 4.48, p < .05, η² = 0.04. However, at the low level of leader PSM, although there was a difference in PS fit between employees with high PSM (M =1.46, SD = .20) and those with low PSM (M =1.82, SD = .15), it was not statistically significant, F(2, 110) = 1.92, p > .10, η² = 0.02. This provides partial support to Hypothesis 4. Finally, in terms of the curvilinear moderated mediation effects, the results showed that the indirect effects of leader PSM on employees’ emotional exhaustion via PS fit were significantly different between low- and high-PSM employees at the low level of leader PSM (Δ = −7.36, p < .001, 95% CI = [−9.63, −5.10]), the moderate level of leader PSM (Δ = −8.98, p < .001, 95% CI = [−11.77, −6.19]), and the high level of leader PSM (Δ = −10.60, p < .001, 95% CI = [−13.91, −7.28]). This provides support to Hypothesis 5.

Table 7.

Comparisons of Marginal Means of PS Fit (3×2 ANOVA) (Study 2).

Employee PSM	Low-PSM leader condition		Moderate-PSM leader condition		High-PSM leader condition
Employee PSM	M	SD	M	SD	M	SD
Low-PSM employees	1.82	.15	3.60	.11	2.89	.14
High-PSM employees	1.46	.20	4.01	.15	3.58	19
F-test	1.92, p > .10		4.48, p < .05		8.03, p < .01
η²	.02		.04		.07

Note. ANOVA = analysis of variance; PSM = public service motivation.

Figure 3.

The Interaction Plot (3×2 ANOVA) (Study 2).

Study 3: Hypothesis Testing

To test Hypothesis 1 (i.e., the curvilinear [U-shaped] relationship between leader PSM and emotional exhaustion), a between-subjects ANOVA was performed. As shown in Table 8, the results revealed a significant effect of leader PSM, F(2, 45) = 13.23, p < .001, η² = 0.37. Post hoc analysis also showed that the low-PSM leader condition was more strongly associated with employee emotional exhaustion (M = 4.05; SD = .72) than was the moderate-PSM leader condition (M = 2.43, SD = .87, p < .001) as well as high-PSM leaders (M = 3.14, SD = .86, p < .05). The moderate and high-PSM leader conditions also differed significantly, p < .05. The polynomial contrast test showed that the effect of leader PSM on employee emotional exhaustion supports a curvilinear (U-shaped) pattern (95% CI [.53, 1.36]). This offers a strong support for Hypothesis 1.

Table 8.

Means and ANOVA Results (Study 3).

DV	Low-PSM leader		Moderate-PSM leader		High-PSM leader		F-test	p-value	η²
DV	M	SD	M	SD	M	SD	F-test	p-value	η²
Exhaustion	4.05	.72	2.43	.87	3.14	.86	13.23	<.001	.37
PS fit	1.58	.45	3.68	.61	2.96	.85	32.91	<.001	.59

Note. ANOVA = analysis of variance; PS = person–supervisor.

Then, to test Hypothesis 2 (i.e., the curvilinear [inverted U-shaped] relationship between leader PSM and PS fit), a between-subjects ANOVA was also performed. As shown in Table 8, the results showed that leader PSM had a positive effect on PS fit, F(2, 45) = 32.91, p < .001, η² = 0.59. Post hoc tests also showed that the moderate-PSM leader condition resulted in higher PS fit (M = 3.68; SD = .61) than the high-PSM leader condition (M = 2.96, SD = .85, p < .01) as well as the low-PSM leader condition (M = 1.58, SD = .45, p < .001). The high- and low-PSM leader conditions also differed significantly, p < .001. The polynomial contrast test also showed that the effect of leader PSM on PS fit supports a curvilinear (inverted U-shaped) pattern (95% CI [−1.49, −0.81]). This supports Hypothesis 2.

To test Hypothesis 3 (i.e., the mediating effect of PS fit), an ANCOVA was performed. PS fit was treated as a covariate in the ANCOVA. As can be seen from Table 9, the results showed a significant effect for PS fit, F(1, 44) = 23.17, p < .001, η² = 0.34, whereas the effect of leader PSM was no longer significant, F(2, 45) = .34, p > .70, η² = 0.01, suggesting that the influence of leader PSM was fully mediated. To test the indirect curvilinear effect of leader PSM, the same MEDCURVE procedure was performed. The results showed that PS fit mediated the effects of the low-PSM leader condition (θ = −1.79, p < .001, 95% CI [−2.62, −1.06]), the moderate-PSM leader condition (θ = −.19, p < .05, 95% CI [−.45, −.02]), and the high-PSM leader condition (θ = 1.40, p < .001, 95% CI [.75, 2.36]). This provides full support for Hypothesis 3.

Table 9.

Main Results (Study 3).

IV	ANCOVADV = emotional exhaustion			3 × 2 ANOVADV = PS fit
IV	F-test	p value	η²	F-test	p value	η²
PS fit	23.17	<.001	.34	—	—	—
Leader PSM^a	.34	<.001	.01	24.53	<.001	.54
Employee PSM^b	—	—	—	.23	>.10	.01
Leader PSM × Employee PSM	—	—	—	4.81	<.05	.19

Note. ANCOVA = analysis of covariance; ANOVA = analysis of variance; PS = person–supervisor; PSM = public service motivation.

Leader PSM was coded as 0, 1, 2. ^bEmployee PSM was coded as 0 and 1, based on the median split.

Next, to test Hypothesis 4 (i.e., the moderating effect of employee PSM on PS fit), a 3×2 between-subjects ANOVA was conducted. Employee PSM was categorized as low (coded as 0) and high (coded as 1), based on the median split (median = 3.30). As can be seen from Table 9, the results further showed that the interaction between leader PSM and employee PSM was significant as predicted, F(2, 42) = 4.81, p < .05, η² = 0.19. Tests of simple effects indicated that the effect of leader PSM on PS fit was significant for both high-PSM employees, F(2, 25) = 27.05, p < .001, η² = 0.68, and low-PSM employees, F(2, 17) = 3.710, p < .05, η² = 0.30. Nevertheless, as can be seen in Figure 4 and Table 10, the observed relationships between leader PSM and PS fit were different between the two groups. At the low level of leader PSM, employees with high PSM perceived lower PS fit (M =1.33, SD = .32) than those with low PSM (M =2.12, SD = .23), F(2, 42) = 4.06, p < .05, η² = 0.09). At the moderate level of leader PSM, employees with high PSM perceived higher PS fit (M =3.88, SD = .18) than those with low PSM (M =3.20, SD = .28), F(2, 42) = 4.07, p < .05, η² = 0.09. However, at the low level of leader PSM, although there was a difference in PS fit between employees with high PSM (M =3.16, SD = .19) and those with low PSM (M =2.61, SD = .24), it was only marginally significant, F(2, 42) = 3.22, p < .10, η² = 0.07. This provides support to Hypothesis 4. Finally, in terms of the curvilinear moderated mediation effects, the results showed that the indirect effects of leader PSM on employees’ emotional exhaustion via PS fit were significantly different between low- and high-PSM employees at the low level of leader PSM (Δ = −4.76, p < .05, 95% CI = [−8.43, −1.10]), the moderate level of leader PSM (Δ = −5.78, p < .05, 95% CI = [−10.43, −1.13]), and the high level of leader PSM (Δ = −6.80, p < .05, 95% CI = [−12.44, −1.15]). This provides support to Hypothesis 5.

Figure 4.

The Interaction Plot (3×2 ANOVA) (Study 3).

Table 10.

Comparisons of Marginal Means of PS Fit (3×2 ANOVA) (Study 3).

Employee PSM	Low-PSM leader condition		Moderate-PSM leader condition		High-PSM leader condition
Employee PSM	M	SD	M	SD	M	SD
Low-PSM employees	2.12	.23	3.20	.28	2.61	.24
High-PSM employees	1.33	.32	3.88	.18	3.16	.19
F-test	4.06, p < .05		4.07, p < .05		3.22, p < .10
η²	.09		.09		.07

Note. PSM = public service motivation.

Discussion

This study contributes to the PSM and public leadership literature by investigating a curvilinear relationship between leader PSM and subordinate emotional exhaustion via the mediating role of PS fit. Results from both field and experimental studies provide convergent support for the proposed hypotheses. In particular, it was found that employees generally prefer leaders with a moderate level of PSM over those with high or low levels of PSM. The experimental results further showed that these effects were contingent upon employees’ PSM. Theoretical and practical implications as well as avenues for future research are discussed below.

Theoretical Implications

First, while past research suggests that individuals’ PSM may have unintended negative consequences for their own well-being (Schott & Ritz, 2017), this current research extends this body of knowledge by turning attention to the role of leader PSM and its potentially negative influence on employee well-being. This research found that the relationship between leader PSM and employee emotional exhaustion is not a simple linear one but follows a non-linear pattern: the positive influence of leader PSM begins to diminish at the higher levels. Thus, while PSM is generally recognized as a “lever” that produces primarily, if not exclusively beneficial effects for society at large (e.g., Coursey et al., 2012; J. L. Perry & Hondeghem, 2008), it is important to acknowledge that leader PSM may also carry paradoxical utility for subordinates. The current findings add to the TMGT literature (Grant & Schwartz, 2011; Pierce & Aguinis, 2013) by demonstrating that there could be two sides to being a public service–motivated leader. While the TMGT effects of leadership are believed to be ubiquitous across different disciplines (Pierce & Aguinis, 2013), very few studies in PA have shed light on this phenomenon in the public-sector sphere. Furthermore, although recent PA research has begun to shed light on the dark side of positive leadership styles (e.g., Jensen, Andersen, & Jacobsen, 2019; Molines et al., 2022), this study is among the first to show that leaders with high levels of PSM may actually undermine subordinates’ well-being. The current findings are consistent with the study by Noblet and Rodwell (2009), which showed that job demands in the public sector can have a curvilinear effect on employees’ well-being.

Second, this research adds to the PSM literature by shedding light on the interactive effects of leaders’ and employees’ PSM on PS fit. Although the results indicate that the moderate level of leader PSM is generally preferred by employees, there are significant differences between high-PSM and low-PSM individuals with respect to how they view their leaders’ PSM. In particular, those with low levels of PSM appeared to perceive lower PS fit with leaders who exhibit moderate to high levels of PSM. On the contrary, for employees with high PSM, the inflection point of the curvilinear relationship was somewhat delayed in comparison to those with low PSM. That is, high-PSM employees, in comparison to those with low PSM, generally perceive higher PS fit with moderate- and high-PSM leaders and, in turn, experience lower emotional exhaustion. These findings point to the importance of supplementary fit (i.e., leaders and followers sharing similar characteristics; Byrne, 1971) while suggesting that PSM may serve as a valuable resource that allows a person to cope with the job demands that emanate from high-PSM leaders (Bakker, 2015). Furthermore, the current findings are consistent with the trait activation theory (Tett et al., 2021), which posits that trait-like variables (e.g., employee PSM) may exert a stronger influence on employees’ outcomes (e.g., PS fit) when accompanied by situational factors that allow them to thrive (e.g., leader PSM). Interestingly, it was also found that individuals with high PSM, in comparison to those with low PSM, were also found to have lower PS fit with low-PSM leaders. This finding points to the possibility that individuals with high PSM may anchor their expectations at a comparatively higher level and in turn experience lower fit with the leaders whose values are incompatible with theirs.

Third, the use of both objective fit (i.e., the interactive effect of leaders’ and employees’ PSM) and perceived fit (i.e., PS fit) in this current research also contributes to the P-E fit literature. In particular, the findings lend credence to the utility of conceptualizing the interaction between leaders’ and employees’ characteristics as a predictor of perceived fit, which permits a more nuanced understanding of the leader-follower value congruence phenomenon (Edwards et al., 2006; Guay et al., 2019). In addition, while P-E fit research on the interaction between leadership styles and employee PSM has gained increasing attention from PA researchers (e.g., Belle, 2013; Kroll & Vogel, 2014; Potipiroon & Ford, 2017), few studies have investigated the fit between the leader and subordinate using commensurate measurement (i.e., measurement items with identical content domains), which is a fundamental requirement of P-E fit (Kristof-Brown et al., 2005). The investigation of PS fit as a mediating mechanism also deserves particular attention from PA scholars. To date, PS fit still remains largely underexplored in comparison to other fit dimensions (Guay et al., 2019). While PSM researchers have used the P-E fit perspective to shed light on the possibility that PSM can enhance person–organization fit (e.g., Bright, 2007; Kim, 2012; Wright & Pandey, 2008) and person–job fit (Christensen & Wright, 2011) or both (Van Loon et al., 2017), no previous research has shed light on the possibility that PS fit can be influenced by employee PSM and, more importantly, by its interaction with leader PSM. This is a worthy line of inquiry because after all, employees are largely affected by the everyday encounters with their superiors who not only have significant control over the nature of employees’ jobs and resource allocation but also how employees view their organizations and, as such, their perceived fit with leaders can have a strong implication for their well-being.

Nevertheless, it is important to acknowledge that the curvilinear effect of leader PSM on subordinate emotional exhaustion may not be driven only by supplemental fit (i.e., PS fit) but also by complementary fit (i.e., needs-supplies and demand-ability fits). For example, under high-PSM leaders, employees may need to spend extra effort and energy on their tasks beyond their abilities and resources (Audenaert et al., 2018). At the same time, low-PSM leaders may fail to acquire sufficient resources for employees to accomplish their work goals. This merits future investigation.

Practical Implications

The above findings are clear indications that public leaders need to exercise moderation in their leadership. Leaders who demonstrate moderation are not only more likely to gain acceptance from followers but are also more likely to avoid overworking themselves to the detriment of subordinates’ well-being. It is therefore imperative that the emphasis on leadership moderation be included in public leadership training programs. It is important to note that this moderation does not mean mediocrity but rather it suggests that leaders must try to strike a balance between encouraging employees to go above and beyond the call of duty and keeping their stress at the productive level. After all, burned-out employees cannot effectively serve citizens (Eldor, 2018). Nevertheless, the fact that employees with high PSM still perceive relatively high PS fit with high-PSM leaders further highlights the importance of recruiting and selecting high-PSM individuals into the public sector. This is not to suggest that public managers should exploit the will and motivation of high-PSM employees by assigning more work responsibility to them; rather, it is necessary for them to provide employees with the needed emotional support especially when workload is perceived to exceed capacity. Moreover, it is still necessary to engage those with low PSM in a participative goal-setting process to ensure important goal alignment.

Study Limitations

Despite the study’s findings, several important limitations should be acknowledged. First, because the survey data in Study 1 were collected via registered mail, it is plausible that the selected respondents may have more favorable views of their leaders. Although the observed variation in the study variables suggests that this was unlikely to be the issue, future research should try to overcome this limitation by employing a data collection procedure that minimizes possible bias in survey responses. Second, the student samples used in the two experiments cannot provide conclusive empirical evidence regarding the impact of leader PSM. While the topic of this research (i.e., imagining working with a prospective leader, assessing PS fit and indicating the anticipated levels of emotional exhaustion) involves very little technicality and is relatively easy to understand for most laymen (Aguinis & Bradley, 2014), future research should use a more diverse sample situated in actual organizational settings. As noted by Hassan and Wright (2020, p. 165),

experimental vignette studies may have very low external validity because participants may respond very differently to vignettes than to actual work settings in which they are more likely to face consequences or other relevant factors may weaken or even reverse their effects.

Relatedly, it is important to acknowledge that the current experiment design only captures the expected level of emotional exhaustion. Although this approach has been used successfully in previous research in applied psychology (e.g., Cameron et al., 2016; Clinton & Pollini, 2021; Martínez-Íñigo et al., 2015), there is still a discrepancy between anticipating to feel something and actually experiencing it (Hassan & Wright, 2020). With that said, it is important to note that, in the PA literature, vignette experiments are widely used to predict attitudes and intentions (e.g., Prysmakova & Evans, 2022) and also the likelihood of engaging in behaviors (e.g., Christensen & Wright, 2018). Furthermore, several techniques were employed to enhance realism in the vignettes including the provision of baseline information to the participants. In particular, the fact that the participants’ PSM, which was measured 2 weeks prior to the experiments, was found to interact with leader PSM does indicate that the participants experienced a high level of immersion during the experiments (Aguinis & Bradley, 2014). Last but not least, the fact that both the field and the experimental data showed convergent support for the curvilinear effects provides some confidence about the credibility of the findings.

Looking forward, given that curvilinear relationships can be difficult to detect (e.g., field data may be restricted in range), future research may find it beneficial to use the “too much/too little” scale format to enhance the possibility of detecting similar effects that exist in the field (Vergauwe et al., 2017). Indeed, the field study results were observed to be much weaker in comparison to those of the experimental studies (e.g., the effects of low-PSM leaders were much more pronounced in the experiments). Furthermore, it is important to acknowledge that the vignettes of leader PSM still need further cross-cultural validation. This is because the meanings of PSM (Mikkelsen et al., 2021) and leadership (Dickson et al., 2012) can vary across cultural contexts. For example, in collectivistic cultures such as Thailand, leaders are expected to show concern for their subordinates and make decisions in the best interest of the group (Potipiroon & Chumphong, 2022), which may be different from what is expected in individualistic cultures. This clearly suggests that much more research in different cultural contexts is still needed to corroborate the current findings.

Conclusion

This research draws attention to the effect of leader PSM as an example of the TMGT phenomenon. The results showed that high levels of leader PSM can have unintended perverse effects on employees’ emotional exhaustion, which can be explained by perceptions of (mis)fit with the leader. These findings illustrate that leader PSM can be a double-edged sword, not only affecting those exhibiting it but also for those on the receiving end. Nevertheless, the silver lining is that these effects also depend on employee PSM. It is hoped that this study will inspire future research to continue investigating the significance of leader PSM and its impact on employee and organizational outcomes.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The previous version of this paper was presented at the 2022 Academy of Management Conference in Seattle, with financial support provided by the Faculty of Management Sciences, Prince of Songkla University, Hat Yai.

ORCID iD

Wisanupong Potipiroon

Data Availability

The data used in this study are available upon request from the corresponding author.

Author Biography

Wisanupong Potipiroon is an associate professor at the Faculty of Management Sciences, Prince of Songkla University in Thailand. He has served as Director of the PhD Program in Management since 2015. He conducts research on public-sector leadership, public service motivation, and organizational corruption. His work has appeared in several leading public management journals including Public Management Review, International Public Management Journal, Review of Public Personnel Administration, Public Personnel Management, and Public Performance and Management Review.

References

Aguinis

Bradley

K. J.

(2014). Best practice recommendations for designing and implementing experimental vignette methodology studies. Organizational Research Methods, 17(4), 351–371.

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. SAGE.

Alonso

Lewis

G. B.

(2001). Public service motivation and job performance: Evidence from the federal sector. American Review of Public Administration, 31(4), 363–380.

Antonakis

House

R. J.

Simonton

D. K.

(2017). Can super smart leaders suffer from too much of a good thing? The curvilinear effect of intelligence on perceived leadership behavior. Journal of Applied Psychology, 102(7), 1003–1021.

Audenaert

Carette

Shore

L. M.

Lange

Van Waeyenberg

Decramer

(2018). Leader-employee congruence of expected contributions in the employee-organization relationship. The Leadership Quarterly, 29(3), 414–422.

Bakker

A. B.

(2015). A job demands–resources approach to public service motivation. Public Administration Review, 75(5), 723–732.

Bandura

(1977). Social learning theory. Prentice-Hall.

Bao

(2019). From moral leadership to positive work behaviors: The mediating roles of value congruence and leader-member exchange. Frontiers of Business Research in China, 13(1), 1–18.

Belle

(2013). Experimental evidence on the relationship between public service motivation and job performance. Public Administration Review, 73(1), 143–153.

10.

Berson

Oreg

Dvir

(2008). CEO values, organizational culture and firm outcomes. Journal of Organizational Behavior, 29(5), 615–633.

11.

Bono

J. E.

Judge

T. A.

(2004). Personality and transformational and transactional leadership: A meta-analysis. Journal of Applied Psychology, 89(5), 901–910.

12.

Bottomley

Mostafa

A. M. S.

Gould-Williams

J. S.

León-Cázares

(2016). The impact of transformational leadership on organizational citizenship behaviours: The contingent role of public service motivation. British Journal of Management, 27(2), 390–405.

13.

Bouwman

Grimmelikhuijsen

(2016). Experimental public administration from 1992 to 2014: A systematic literature review and ways forward. International Journal of Public Sector Management, 29(2), 110–131.

14.

Brewer

G. A.

(2003). Building social capital: Civic attitudes and behavior of public servants. Journal of Public Administration Research and Theory, 13(1), 5–26.

15.

Bright

(2007). Does person-organization fit mediate the relationship between public service motivation and the job performance of public employees? Review of Public Personnel Administration, 27(4), 361–379.

16.

Broekema

Porth

Steen

Torenvlied

(2019). Public leaders’ organizational learning orientations in the wake of a crisis and the role of public service motivation. Safety Science, 113, 200–209.

17.

Byrne

D. E.

(1971). The attraction paradigm (Vol. 462). Academic Press.

18.

Cable

D. M.

DeRue

(2002). The convergent and discriminant validity of subjective fit perceptions. Journal of Applied Psychology, 87(5), 875–884.

19.

Cable

D. M.

Edwards

J. R.

(2004). Complementary and supplementary fit: A theoretical and empirical integration. Journal of Applied Psychology, 89(5), 822–834.

20.

Cameron

C. D.

Harris

L. T.

Payne

B. K.

(2016). The emotional cost of humanity: Anticipated exhaustion motivates dehumanization of stigmatized targets. Social Psychological and Personality Science, 7(2), 105–112.

21.

Christensen

R. K.

Whiting

S. W.

Rho

Stritch

J. M.

Park

(2013). Public service motivation, task, and non-task behavior: A performance appraisal experiment with Korean MPA and MBA students. International Public Management Journal, 16(1), 28–52.

22.

Christensen

R. K.

Wright

B. E.

(2011). The effects of public service motivation on job choice decisions: Disentangling the contributions of person-organization fit and person-job fit. Journal of Public Administration Research and Theory, 21(4), 723–743.

23.

Christensen

R. K.

Wright

B. E.

(2018). Public service motivation and ethical behavior: Evidence from three experiments. Journal of Behavioral Public Administration, 1(1), 1–8.

24.

Clinton

A. J.

Pollini

R. A.

(2021). Using positive empathy interventions to reduce stigma toward people who inject drugs. Frontiers in Psychology, 12, 2836.

25.

Conway

Kiefer

Hartley

Briner

R. B.

(2014). Doing more with less? Employee reactions to psychological contract breach via target similarity or spillover during public sector organizational change. British Journal of Management, 25(4), 737–754.

26.

Coursey

Pandey

S. K.

Yang

(2012). Public service motivation (PSM) and support for citizen participation: A test of Perry and Vandenabeele’s reformulation of PSM theory. Public Administration Review, 72(4), 572–582.

27.

Crawford

E. R.

LePine

J. A.

Rich

B. L.

(2010). Linking job demands and resources to employee engagement and burnout: A theoretical extension and meta-analytic test. Journal of Applied Psychology, 95(5), 834–848.

28.

Dawson

J. F.

(2013). Moderation in management research: What, why, when, and how. Journal of Business and Psychology, 29(1), 1–19.

29.

Dickson

M. W.

Castaño

Magomaeva

Den Hartog

D. N.

(2012). Conceptualizing leadership across cultures. Journal of World Business, 47(4), 483–492.

30.

Diebig

Bormann

K. C.

Rowold

(2016). A double-edged sword: Relationship between full-range leadership behaviors and followers’ hair cortisol level. The Leadership Quarterly, 27(4), 684–696.

31.

Edwards

J. R.

Cable

D. M.

Williamson

I. O.

Lambert

L. S.

Shipp

A. J.

(2006). The phenomenology of fit: Linking the person and environment to the subjective experience of person-environment fit. Journal of Applied Psychology, 91(4), 802–827.

32.

Edwards

J. R.

Cooper

C. L.

(1990). The person-environment fit approach to stress: Recurring problems and some suggested solutions. Journal of Organizational Behavior, 11, 293–307.

33.

Edwards

J. R.

Lambert

L. S.

(2007). Methods for integrating moderation and mediation: A general analytical framework using moderated path analysis. Psychological Methods, 12(1), 1–22.

34.

Eldor

(2018). Public service sector: The compassionate workplace—The effect of compassion and stress on employee engagement, burnout, and performance. Journal of Public Administration Research and Theory, 28(1), 86–103.

35.

Enwereuzor

I. K.

Ugwu

L. E.

Nnadozie

E. E.

(2020). Dynamics of person-supervisor fit in relationship quality and well-being of university academicians. Personnel Review, 50(5), 1356–1371.

36.

French

J. R.

Caplan

R. D.

Van Harrison

(1982). The mechanisms of job stress and strain. Wiley.

37.

García-Romero

Martinez-Iñigo

(2021). Validation of an attributional and distributive justice mediational model on the effects of surface acting on emotional exhaustion: An experimental study. International Journal of Environmental Research and Public Health, 18(14), 7505.

38.

Giauque

Ritz

Varone

Anderfuhren-Biget

(2012). Resigned but satisfied: The negative impact of public service motivation and red tape on work satisfaction. Public Administration, 90(1), 175–193.

39.

Grant

A. M.

Schwartz

(2011). Too much of a good thing: The challenge and opportunity of the inverted U. Perspectives on Psychological Science, 6(1), 61–76.

40.

Gross

H. P.

Thaler

Winter

(2019). Integrating public service motivation in the job-demands-resources model: An empirical analysis to explain employees’ performance, absenteeism, and presenteeism. International Public Management Journal, 22(1), 176–206.

41.

Guan

Deng

Risavy

S. D.

Bond

M. H.

(2011). Supplementary fit, complementary fit, and work-related outcomes: The role of self-construal. Applied Psychology, 60(2), 286–310.

42.

Guay

R. P.

Kim

Y.-J.

I.-S.

Vogel

R. M.

(2019). The interaction effects of leader and follower conscientiousness on person-supervisor fit perceptions and follower outcomes: A cross-level moderated indirect effects model. Human Performance, 32(3–4), 181–199.

43.

Hambrick

D. C.

Mason

P. A.

(1984). Upper echelons: The organization as a reflection of its top managers. Academy of Management Review, 9(2), 193–206.

44.

Hameduddin

Engbers

(2022). Leadership and public service motivation: A systematic synthesis. International Public Management Journal, 25(1), 86–119.

45.

Hassan

Wright

B. E.

(2020). The behavioral public administration movement: A critical reflection. Public Administration Review, 80(1), 163–167.

46.

Hayes

A. F.

Preacher

K. J.

(2010). Quantifying and testing indirect effects in simple mediation models when the constituent paths are nonlinear. Multivariate Behavioral Research, 45(4), 627–660.

47.

Hoffman

B. J.

Bynum

B. H.

Piccolo

R. F.

Sutton

A. W.

(2011). Person-organization value congruence: How transformational leaders influence work group effectiveness. Academy of Management Journal, 54(4), 779–796.

48.

Hofmann

D. A.

Gavin

M. B.

(1998). Centering decisions in hierarchical linear models: Implications for research in organizations. Journal of Management, 24(5), 623–641.

49.

Hondeghem

Perry

J. L.

(2009). EGPA symposium on public service motivation and performance: Introduction. International Review of Administrative Sciences, 75(1), 5–9.

50.

Houston

D. J.

(2006). “Walking the walk” of public service motivation: Public employees and charitable gifts of time, blood, and money. Journal of Public Administration Research and Theory, 16(1), 67–86.

51.

Jensen

U. T.

Andersen

L. B.

Holten

A.-L.

(2019). Explaining a dark side: Public service motivation, presenteeism, and absenteeism. Review of Public Personnel Administration, 39(4), 487–510.

52.

Jensen

U. T.

Andersen

L. B.

Jacobsen

C. B.

(2019). Only when we agree! How value congruence moderates the impact of goal-oriented leadership on public service motivation. Public Administration Review, 79(1), 12–24.

53.

Jiang

Dong

Liu

Guan

(2022). Leaders’ response to employee overqualification: An explanation of the curvilinear moderated relationship. Journal of Occupational and Organizational Psychology, 95(2), 459–494.

54.

Judge

T. A.

Bono

J. E.

Ilies

Gerhardt

M. W.

(2002). Personality and leadership: A qualitative and quantitative review. Journal of Applied Psychology, 87(4), 765–780.

55.

Judge

T. A.

Piccolo

R. F.

Kosalka

(2009). The bright and dark sides of leader traits: A review and theoretical extension of the leader trait paradigm. The Leadership Quarterly, 20(6), 855–875.

56.

Jung

D. I.

Avolio

B. J.

(2000). Opening the black box: An experimental investigation of the mediating effects of trust and value congruence on transformational and transactional leadership. Journal of Organizational Behavior, 21(8), 949–964.

57.

Kim

(2012). Does Person-organization fit matter in the public-sector? Testing the mediating effect of Person-organization fit in the relationship between public service motivation and work attitudes. Public Administration Review, 72(6), 830–840.

58.

Kottler

J. A.

(2018). What you don’t know about leadership, but probably should: Applications to daily life. Oxford University Press.

59.

Kristof-Brown

A. L.

Zimmerman

R. D.

Johnson

E. C.

(2005). Consequences of individuals’ fit at work: A meta-analysis of person-job, person-organization, person-group, and person-supervisor fit. Personnel Psychology, 58(2), 281–342.

60.

Kroll

Vogel

(2014). The PSM–leadership fit: A model of performance information use. Public Administration, 92(4), 974–991.

61.

Lamiani

Dordoni

Argentero

(2018). Value congruence and depressive symptoms among critical care clinicians: The mediating role of moral distress. Stress and Health, 34(1), 135–142.

62.

Marstand

A. F.

Martin

Epitropaki

(2017). Complementary person-supervisor fit: An investigation of supplies-values (SV) fit, leader-member exchange (LMX) and work outcomes. The Leadership Quarterly, 28(3), 418–437.

63.

Martínez-Íñigo

Mercado

Totterdell

(2015). Using interpersonal affect regulation in simulated healthcare consultations: An experimental investigation of self-control resource depletion. Frontiers in Psychology, 6, 1485.

64.

Maslach

Jackson

S. E.

(1981). The measurement of experienced burnout. Journal of Organizational Behavior, 2(2), 99–113.

65.

Maslach

Schaufeli

W. B.

Leiter

M. P.

(2001). Job burnout. Annual Review of Psychology, 52(1), 397–422.

66.

McGrath

(1976). Sress and behavior in organizations. In Dunnette

(Ed.), Handbook of industrial and oranizational psychology (pp. 1351–1395). Rand McNally.

67.

Mikkelsen

K. S.

Schuster

Meyer-Sahling

J.-H.

(2021). A cross-cultural basis for public service? Public service motivation measurement invariance in an original survey of 23,000 public servants in ten countries and four world regions. International Public Management Journal, 24(6), 739–761.

68.

Molines

El Akremi

Storme

Celik

(2022). Beyond the tipping point: The curvilinear relationships of transformational leadership, leader–member exchange, and emotional exhaustion in the French police. Public Management Review, 24(1), 80–105.

69.

Muthén

(2018). Mplus user’s guide. (Original work published 1998)

70.

Naff

K. C.

Crum

(1999). Working for America: Does public service motivation make a difference? Review of Public Personnel Administration, 19(4), 5–16.

71.

Nielsen

Daniels

(2016). The relationship between transformational leadership and follower sickness absence: The role of presenteeism. Work & Stress, 30(2), 193–208.

72.

Noblet

A. J.

Rodwell

J. J.

(2009). Integrating job stress and social exchange theories to predict employee strain in reformed public sector contexts. Journal of Public Administration Research and Theory, 19(3), 555–578.

73.

Paarlberg

L. E.

Lavigna

(2010). Transformational leadership and public service motivation: Driving individual and organizational performance. Public Administration Review, 70(5), 710–718.

74.

Pedersen

L. H.

(2014). Committed to the public interest? Motivation and behavioural outcomes among local councillors. Public Administration, 92(4), 886–901.

75.

Perry

J. L.

(1996). Measuring public service motivation: An assessment of construct reliability and validity. Journal of Public Administration Research & Theory, 6(1), 5–22.

76.

Perry

J. L.

Hondeghem

(2008). Motivation in public management: The call of public service. Oxford University Press.

77.

Perry

J. L.

Hondeghem

Wise

L. R.

(2010). Revisiting the motivational bases of public service: Twenty years of research and an agenda for the future. Public Administration Review, 70(5), 681–690.

78.

Perry

J. L.

Vandenabeele

(2008). Behavioral dynamics: Institutions, identities, and self-regulation. In Perry

J. L.

Hondeghem

(Eds.), Motivation in public management: The call of public service (pp. 56–79). Oxford University Press.

79.

Perry

J. L.

Wise

L. R.

(1990). The motivational bases of public service. Public Administration Review, 50(3), 367–373.

80.

Perry

S. J.

Rubino

Witt

(2011). General job performance of first-line supervisors: The role of conscientiousness in determining its effects on subordinate exhaustion. Stress and Health, 27(2), 83–93.

81.

Pierce

J. R.

Aguinis

(2013). The too-much-of-a-good-thing effect in management. Journal of Management, 39(2), 313–338.

82.

Podsakoff

P. M.

MacKenzie

S. B.

Jeong-Yeon

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903.

83.

Potipiroon

Chumphong

(2022). Authoritarian leadership and firm-level voluntary turnover among SMEs in Thailand: Does benevolent leadership matter? International Journal of Emerging Markets. Advance online publication. https://doi.org/10.1108/ijoem-07-2021-1144

84.

Potipiroon

Faerman

(2016). What difference do ethical leaders make? Exploring the mediating role of interpersonal justice and the moderating role of public service motivation. International Public Management Journal, 19(2), 171–207.

85.

Potipiroon

Ford

M. T.

(2017). Does public service motivation always lead to organizational commitment? Examining the moderating roles of intrinsic motivation and ethical leadership. Public Personnel Management, 46(3), 211–238.

86.

Prysmakova

Evans

M. D.

(2022). Whistleblowing motivation and gender: Vignette-based study in a local government. Review of Public Personnel Administration, 42(1), 165–190.

87.

Rainey

H. G.

Steinbauer

(1999). Galloping elephants: Developing elements of a theory of effective government organizations. Journal of Public Administration Research & Theory, 9(1), 1–32.

88.

Raudenbush

S. W.

Bryk

A. S.

(2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). SAGE.

89.

Ritz

(2015). Public service motivation and politics: Behavioural consequences among local councillors in Switzerland. Public Administration, 93(4), 1121–1137.

90.

Ritz

Brewer

G. A.

Neumann

(2016). Public service motivation: A systematic literature review and outlook. Public Administration Review, 76(3), 414–426.

91.

Robertson

Healey

M. P.

Hodgkinson

G. P.

Flint-Taylor

Jones

(2014). Leader personality and employees’ experience of workplace stressors. Journal of Organizational Effectiveness, 1(3), 281–295.

92.

Safavi

H. P.

Bouzari

(2020). How can leaders enhance employees’ psychological capital? Mediation effect of person-group and person-supervisor fit. Tourism Management Perspectives, 33, 100626.

93.

Schaubroeck

Lam

S. S.

(2002). How similarity to peers and supervisor influences organizational advancement in different cultures. Academy of Management Journal, 45(6), 1120–1136.

94.

Schott

Ritz

(2017). The dark sides of public service motivation: A multi-level theoretical framework. Perspectives on Public Management and Governance, 1(1), 29–42.

95.

Shamir

House

R. J.

Arthur

M. B.

(1993). The motivational effects of charismatic leadership: A self-concept based theory. Organization Science, 4(4), 577–594.

96.

Shim

D. C.

Park

H. H.

Eom

T. H.

(2017). Street-level bureaucrats’ turnover intention: Does public service motivation matter? International Review of Administrative Sciences, 83(3), 563–582.

97.

Siegall

McDonald

(2004). Person-organization value congruence, burnout and diversion of resources. Personnel Review, 33(3), 291–301.

98.

Snijders

Bosker

(1999). Multilevel modeling: An introduction to basic and advanced multilevel modeling. SAGE.

99.

StataCorp. (2012). Stata statistical software.

100.

Stazyk

E. C.

Davis

R. S.

(2015). Taking the ‘high road’: Does public service motivation alter ethical decision making processes? Public Administration, 93(3), 627–645.

101.

Stouten

Van Dijke

Mayer

D. M.

De Cremer

Euwema

M. C.

(2013). Can a leader be seen as too ethical? The curvilinear effects of ethical leadership. The Leadership Quarterly, 24(5), 680–695.

102.

Syrek

C. J.

Antoni

C. H.

(2014). Unfinished tasks foster rumination and impair sleeping—Particularly if leaders have high performance expectations. Journal of Occupational Health Psychology, 19(4), 490–499.

103.

Tang

Cai

Liu

Zhu

Yang

(2015). The importance of ethical leadership in employees’ value congruence and turnover. Cornell Hospitality Quarterly, 56(4), 397–410.

104.

Tekleab

A. G.

Reagan

P. M.

Levi

Lichtman

(2021). Translating corporate social responsibility into action: A social learning perspective. Journal of Business Ethics, 171(4), 741–756.

105.

Tett

R. P.

Toich

M. J.

Ozkum

S. B.

(2021). Trait activation theory: A review of the literature and applications to five lines of personality dynamics research. Annual Review of Organizational Psychology and Organizational Behavior, 8, 199–233.

106.

Vandenabeele

(2007). Toward a public administration theory of public service motivation: An institutional approach. Public Management Review, 9(4), 545–556.

107.

Vandenabeele

(2014). Explaining public service motivation: The role of leadership and basic needs satisfaction. Review of Public Personnel Administration, 34(2), 153–173.

108.

van Loon

N. M.

Vandenabeele

Leisink

. (2015). On the bright and dark side of public service motivation: The relationship between PSM and employee wellbeing. Public Money & Management, 35(5), 349–356.

109.

Van Loon

N. M.

Vandenabeele

Leisink

. (2017). Clarifying the relationship between public service motivation and in-role and extra-role behaviors: The relative contributions of person-job and person-organization fit. The American Review of Public Administration, 47(6), 699–713.

110.

Vergauwe

Wille

Hofmans

Kaiser

R. B.

Fruyt

F. D.

(2017). The too little/too much scale: A new rating format for detecting curvilinear effects. Organizational Research Methods, 20(3), 518–544.

111.

Vogel

Werkmeister

(2021). What is public about public leadership? Exploring implicit public leadership theories. Journal of Public Administration Research and Theory, 31(1), 166–183.

112.

Witt

(1998). Enhancing organizational goal congruence: A solution to organizational politics. Journal of Applied Psychology, 83(4), 666–674.

113.

Wright

B. E.

Christensen

R. K.

Pandey

S. K.

(2013). Measuring public service motivation: Exploring the equivalence of existing global measures. International Public Management Journal, 16(2), 197–223.

114.

Wright

B. E.

Hassan

Park

(2016). Does a public service ethic encourage ethical behavior? Public service motivation, ethical leadership and the report ethical problems. Public Administration, 94(3), 647–663.

115.

Wright

B. E.

Pandey

S. K.

(2008). Public service motivation and the assumption of person-organization fit: Testing the mediating effect of value congruence. Administration & Society, 40(5), 502–521.