Abstract
Motivating public service employees to greater effort is a key issue for managers and scholars. Transformational leadership concerns behaviors to develop, share, and sustain a vision for the organization and has been suggested as an important lever in this respect. However, we know little about the processes by which transformational leadership may stimulate work motivation. Integrating transformational leadership, public service motivation (PSM), and self-determination theory, this article sheds light on the psychological mechanisms underlying the motivational effects of transformational leadership. According to structural equation modeling, the relationships between transformational leadership and two types of autonomous work motivation—intrinsic motivation and PSM—are mediated by the satisfaction of the basic psychological needs for autonomy, competence, and relatedness. Our findings support the claim that the motivational effects of transformational leadership are mediated by need satisfaction, but also that satisfaction of individual needs is not equally important for intrinsic motivation and PSM, respectively.
Keywords
Introduction
Fostering work motivation is important, because work motivation has been related to job satisfaction (Cantarelli, Belardinelli, & Bellé, 2016) and performance (Andersen, Heinesen, & Pedersen, 2014; Bellé, 2013). In times of scarce resources and continuous demands for higher performance, abilities to increase work motivation may prove to be an important way for public organizations to reach organizational goals while keeping employees satisfied with their jobs. In the public administration literature, focus has centered on the concept of public service motivation (PSM). 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), PSM can be seen as a specific type of prosocial motivation (Perry, Wise, & Recascino, 2010). As such, PSM is often contrasted with other types of motivation based on the inherent enjoyment of an activity (intrinsic motivation) or the expected rewards/avoidance of sanctions (external regulation). Importantly, PSM not only energizes employees to make efforts toward interesting and enjoyable tasks with prosocial consequences, but it also helps explain why employees put extra efforts into tasks that are neither seen as inherently enjoyable nor incentivized by pecuniary rewards (Houston, 2011), making it important for public leaders to know how stimulate this type of motivation.
Transformational leadership entails behaviors that aim to direct and inspire efforts toward fulfilling organizational goals by articulating a vision that raises employees’ awareness of the importance of organizational values, mission, and outcomes. Aiming to make employees transcend their own self-interest for the sake of the organization (Jensen et al., 2016; Podsakoff, MacKenzie, & Bommer, 1996; Wright, Moynihan, & Pandey, 2012), this leadership strategy is increasingly recognized within public administration literature as an important antecedent of PSM (e.g., Wright et al., 2012). While recent empirical studies demonstrate positive correlations between transformational leadership and PSM (Krogsgaard, Thomsen, & Andersen, 2014; Park & Rainey, 2008; Vandenabeele, 2014; Wright et al., 2012), none of these studies integrate other types of motivation or explore the mechanisms underlying this relationship. A core expectation in transformational leadership theory is that transformational leaders “activate the higher-order needs of their employees” (Wright et al., 2012, p. 207), but this expectation has only been sparsely tested (see Hetland, Hetland, Andreassen, Pallesen, & Notelaers, 2011; Kovjanic, Schuh, & Jonas, 2013; Kovjanic, Schuh, Jonas, Quaquebeke, & Dick, 2012), and it has not been examined with motivation as the final outcome.
In this article, we build on insights from self-determination theory (SDT), as it introduces the importance of needs (for autonomy, competence, and relatedness) satisfaction and offers a framework for distinguishing different types of motivation (Deci & Ryan, 2000; Gagné & Deci, 2005). Introducing the satisfaction of basic needs as a mediating mechanism between transformational leadership and different types of autonomous work motivation is important for several reasons: First, understanding the psychological mechanisms through which organizational leadership can enhance or thwart employee work motivation is pivotal for managers. Ideally, managers can use such information in decisions on ways to motivate employees to achieve organizational goals. Second, motivation is a multifaceted concept (Wise, 2004), and multiple types of work motivation may be beneficial for reaching different desirable outcomes (Jensen & Andersen, 2015). Consequently, scholars provide a fuller picture of the motivational effects by including different types of work motivation simultaneously in empirical studies. Examining two distinct types of autonomous work motivation—intrinsic motivation and PSM—this article responds to such calls by offering a fuller picture on how to motivate employees through transformational leadership. Third and finally, integrating theories of self-determination, transformational leadership, and PSM has the potential to shed light on ways that leadership can shape employee perceptions of external steering mechanisms such as monetary incentive schemes or command systems to avoid “crowding out” of intrinsic motivation and PSM (e.g., Frey, 1997; Jacobsen, Hvitved, & Andersen, 2014).
The article tests the relationships between transformational leadership, satisfaction of basic needs, and the two types of motivation on a dataset consisting of survey responses from 1,481 teachers in Denmark. Using structural equation modeling (SEM) technique, the article examines the direct and indirect pathways (through the satisfaction of basic needs) between an aggregated measure of transformational leadership (at the organizational level) and individual employees’ intrinsic motivation and PSM, respectively.
Before we argue how basic need satisfaction mediates the link between transformational leadership and motivation, we first conceptualize transformational leadership and offer a distinction between intrinsic motivation and PSM from a SDT perspective. We then discuss our research design, data, and methods before presenting our results. The article concludes with a discussion of the main findings, their implications for research and practice, and the limitations of our study.
Conceptualizing Transformational Leadership
According to traditional transformational leadership literature, transformational leaders direct and inspire employee effort by articulating a vision that raises employees’ awareness and understanding of the importance of organizational values, mission, and outcomes (Wright et al., 2012). Consequently, central to the theory on transformational leadership is a strong emphasis on the role of a collective vision; that is, an idealized set of goals that the organization aspires to achieve one day (Carton, Murphy, & Clark, 2014). However, the visionary component has not only been core to the conceptualization of transformational leadership in contemporary management literature (e.g., Jung & Avolio, 2000) and public administration research (e.g., Paarlberg & Lavigna, 2010); it has also been highlighted as the key feature that makes transformational leadership particularly relevant in public and nonprofit organizations (given that such organizations rely on strong service- and community-oriented missions; see Wright & Pandey, 2010). In fact, transformational leadership has been linked to PSM, in that “the more engaging, attractive and worthwhile the mission is to people, the more the agency will be able to attract support from those people . . . and motivate them to perform well in the agency” (Rainey & Steinbauer, 1999, p. 16). Given the strong emphasis among management scholars (e.g., Jung & Avolio, 2000) and public administration researchers (e.g., Paarlberg & Lavigna, 2010; Wright et al., 2012) on visions as a key feature of transformational leadership, this article follows Jensen et al. (2016) by focusing on transformational leadership as a set of behaviors that involves developing, sharing, and sustaining an organizational vision.
Inspirational goals have long been considered an important driver of employee action and performance (see, for example, Latham & Yukl, 1975), and transformational leaders therefore strive to develop a clear vision for their organization. Next, transformational leaders seek to communicate the vision to employees. This includes establishing a clear connection between the vision and everyday work tasks to make individual employees understand how they contribute to designated outcomes of their organization. Finally, transformational leaders make an effort to sustain employees’ attention to the vision in both the short and long run by encouraging employees to work toward the vision (Wright et al., 2012). These three behaviors are intertwined, and we thus see them as reflections of the same (latent) endeavor to transform employee motivation. As noted above, transformational leaders engage in such behaviors with the objective to make employees transcend their own self-interest in favor of organizational goals, and we therefore follow Jacobsen and Andersen (2015) in defining transformational leadership as “behaviors seeking to develop, share, and sustain a vision intended to encourage employees to transcend their own self-interest and achieve organization goals” (p. 832).
Next, the article draws on SDT to distinguish between different types of work motivation, and to argue how satisfaction of the needs for autonomy, competence, and relatedness mediates the link between transformational leadership and two types of motivation—intrinsic motivation and PSM, respectively.
Intrinsic Motivation and PSM From a Self-Determination Perspective
According to SDT, people not only differ in terms of how much motivation they have for performing an activity but also in terms of the type of motivation behind the action (Ryan & Deci, 2000a). A basic distinction within the theory is between autonomous motivation and controlled motivation, determined by whether the motivation involves “a sense of volition and having the experience of choice” or “a sense of pressure and having to engage in actions” (Gagné & Deci, 2005, p. 334). The types of motivation can be seen as part of a motivational continuum, on which intrinsic motivation is prototypically autonomous (Gagné & Deci, 2005). Referring to whether an activity is seen as inherently interesting or enjoyable (Ryan & Deci, 2000b), intrinsic motivation is contrasted to other types of extrinsic motivation, which require some sort of instrumentality between the activity and a separable outcome (Gagné & Deci, 2005). When the activity is performed because of an external (or “as if” external) consequence (such as implicit approval, tangible rewards, or enhanced self-esteem), the motivation is said to be extrinsically controlling. However, when a behavioral regulation, and the value associated with it, has been internalized to the extent that an external consequence is no longer needed, the motivation is said to be extrinsically autonomous (Gagné & Deci, 2005). According to SDT, the latter case occurs when people identify with the value of an (unpleasant) activity for their own self-selected goals, or even see the behaviors as an integral part of who they are (the goals, values, and regulations of a behavior are coherent with other aspects of the person’s life; Gagné & Deci, 2005). Teachers could, for example, come to value the grammatical development of their students to the extent that they not only understand why they should perform their share of unpleasant tasks in this regard (such as correcting a lot of very similar grammatical tests) but also appreciate the importance of doing so.
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), PSM has both autonomous and extrinsic characteristics. When public service motivated activities are based on the expected external outcome of the activity (to do good), they may involve efforts directed at tasks that are not seen as inherently interesting. However, the expected outcome is not directed at (and is therefore not controlling for) the person performing the activity. Instead, such tasks are likely performed because “the activities are necessary for the ultimate delivery of service or because not performing the tasks would be unthinkable for people like themselves” (Houston, 2011, p. 763). PSM is thus an autonomous type of motivation that differs from intrinsic motivation by not being (primarily) characterized by the inherent interest in a given activity but by the activity being instrumentally important for other people and society. This is illustrated in Figure 1.

Motivation continuum.
According to SDT, satisfaction of an individual’s basic psychological needs for autonomy, competence, and relatedness is crucial for the internalization and maintenance of intrinsic motivation (Deci & Ryan, 2000; Gagné & Deci, 2005). Need satisfaction has also been considered a core feature of transformational leadership theory (Podsakoff et al., 1996). As noted by Kovjanic and colleagues (2012), it is by “appreciating and addressing needs that transformational leaders develop the potential of their followers and foster their commitment to and effort for the collective” (p. 1032). This argument is echoed by Wright and Pandey (2010), who state that
to direct and inspire individual effort, these [transformational] leaders transform their followers by raising their awareness of the importance of organizational outcomes, thereby activating their higher ordered needs and inducing them to transcend their own self-interest for the sake of the organization. (p. 76, emphasis added)
We therefore argue that SDT and its notion of needs satisfaction are necessary components for understanding how transformational leadership fosters intrinsic motivation and PSM, respectively. This argument is elaborated in the next section.
Transformational Leadership and Autonomous Motivation: The Mediating Role of Basic Need Satisfaction
The need for autonomy concerns the experience of choice and feeling like the initiator of one’s own actions (Baard, Deci, & Ryan, 2004). Transformational leadership, it is argued, strives to clarify the purpose of employees’ work tasks, in turn making employees experience their work as meaningful (Kovjanic et al., 2012). Establishing the broader organizational goals and objectives in the context of a vision, employees of transformational leaders do not require continuous guidance as to whether actions are supportive of the organization’s purpose. Moreover, transformational leadership distinguishes itself from leadership behaviors focusing on close monitoring of employee actions (e.g., to reward or punish employees) by the absence of “micromanagement” used to ensure that employees do not pursue different agendas. The few existing studies corroborate this perspective, demonstrating positive correlations between transformational leadership and satisfaction of the need for autonomy (Hetland et al., 2011; Kovjanic et al., 2013; Kovjanic et al., 2012). Transformational leaders thus instigate a sense of satisfaction of the need for autonomy by (a) clearly establishing the broader purpose of the organization, allowing employees to initiate everyday tasks that contribute to designated outcomes; and (b) avoiding micromanagement such as close monitoring of employee actions or setting very specific goals to direct the everyday tasks and effort of individual employees. We therefore expect transformational leadership to enhance employees’ satisfaction of the need for autonomy.
Second, the need for competence concerns succeeding at challenging tasks, being able to attain desired outcomes, and the feeling of generally being effective (Baard et al., 2004; Leary & Tangney, 2003). Articulating and communicating desirable future end states of the organization (vision) and establishing how the work of employees contributes to achieving designated outcomes, transformational leaders set desirable goals and express confidence that these goals can be achieved by the employees through their work. This is central to the feeling of competence among the employees (Latham & Yukl, 1975; Locke & Latham, 2002), something that is closely linked to feeling self-efficacious in the workplace.
Third, the need for relatedness refers to the desire to feel connected to others; that is, to “establish a sense of mutual respect and reliance with others” (Baard et al., 2004, p. 2046). Encouraging employees to “pull” in the same direction to achieve designated outcomes linked to the organization’s purpose, transformational leaders distinguish the organization—and its members—from other groups (Burns, 1978). In turn, this is expected to evoke an in-group feeling and a sense of being connected to others in the work. For this reason, transformational leadership is expected to enhance employees’ satisfaction of the need for relatedness. In sum, the article thus argues that transformational leaders prompt a sense of satisfaction of the basic psychological needs for autonomy, competence, and relatedness among employees:
According to SDT literature, the satisfaction of basic psychological needs is important to both intrinsic motivation and the internalization of extrinsic motivation (Gagné & Deci, 2005). Although intrinsic motivation is based on inherent interests, individuals cannot pursue all interests or natural inclinations. Rather, intrinsic motivation for performing a given activity will be nourished or thwarted depending on the employee’s experience of feeling autonomous, competent, and related to others while performing the activity (Gagné & Deci, 2005). As PSM can also be characterized as autonomous motivation and involves a sense of volition, it seems likely that individuals’ PSM also depends on need satisfaction. PSM is based on a consideration for others and a fundamental willingness to set one’s own needs aside for others and society (self-sacrifice; Perry, 1996; Perry & Wise, 1990). Employees’ PSM can therefore be thwarted if they feel commanded to perform an activity, as this reduces the individual’s perception of his or her own sacrifice in connection with the activity (Jacobsen et al., 2014). Furthermore, feeling incapable of or inefficacious in reaching goals that are deemed important may give rise to frustration and ultimately cause people to “give up.” Satisfaction of the need for competence is therefore expected to be related to PSM. The argument is supported by research on goal setting and goal commitment, which shows that individuals are likely to become more motivated when goals are difficult, but attainable, and when individuals feel capable of or efficacious in reaching the goals (Wright, 2001). Finally, with regard to the need to feel connected to others, studies on the importance of work climate show that individuals’ perceptions of their work environment (including work group characteristics such as cooperation, pride, and warmth) are positively related to work attitudes, motivation, and performance (Parker et al., 2003). In line with this research, the article expects that employees who experience a satisfaction of the need for relatedness with people at the workplace are willing to invest greater energy in work aimed at increasing the welfare of others.
In the field of public administration, one previous study deals with the interplay between transformational leadership, PSM, and the satisfaction of basic needs (Vandenabeele, 2014). Looking at 3,506 Belgian civil servants, Vandenabeele shows that the relationship between transformational leadership and PSM depends on satisfaction of the basic needs for autonomy and competence. This illustrates the importance of understanding the links between transformational leadership, basic need satisfaction, and PSM, but it offers little insight into the question of whether transformational leaders can influence need satisfaction to stimulate employee work motivation. The former link—between transformational leadership and basic need satisfaction—is discussed in SDT (Gagné & Deci, 2005), and has been empirically validated (e.g., Hetland et al., 2011; Kovjanic et al., 2013; Kovjanic et al., 2012). The study of Kovjanic et al. (2013) even demonstrated this using an experimental survey design. However, no study has, to our knowledge, examined whether transformational leadership stimulates PSM and intrinsic motivation through satisfaction of the basic psychological needs for autonomy, competence, and relatedness. To help close this gap, the article tests the following hypotheses:
Research Design and Data
The article draws on a cross-sectional research design with data on 1,481 teachers distributed to 129 private and public schools in Denmark. Danish schools constitute an appropriate research setting in our case for at least three reasons: First, schools resemble “classic” organizations in the sense that conceptions of what is conceived as desirable to pursue in this setting are fairly homogeneous. Public service organizations are multifaceted and serve a number of (sometimes conflicting) goals (Meier & Krause, 2005). Schools are no exception, but it still seems reasonable to expect that leaders (principals) and employees (teachers) share a core focus on educating and academically preparing students. Second, school principals in Denmark enjoy considerable managerial autonomy. For example, school principals in public schools are provided discretion to allocate the time teachers spend on various job tasks (e.g., preparations and actual teaching in the classrooms), and we expect that principals will use their managerial autonomy to motivate teachers in different ways. In a recent study of Danish secondary schools, Jacobsen and Andersen (2015) reported the use of transformational leadership behaviors to varying degrees. This indicates that transformational leadership is likely to represent relevant leadership behaviors in the educational sectors. In fact, transformational leadership may be particularly relevant in public service organizations, because such organizations have strong service- and community-oriented visions (Wright et al., 2012). Schools are prime examples of such organizations due to their educational purpose (Andersen et al., 2014), and this suggests that (Danish) schools constitute an empirical research setting that provides favorable conditions for testing the proposed hypotheses. Third, and related, teaching essentially concerns doing good for the students and society, and existing studies have found public service to be a relevant motivator among teachers (Andersen et al., 2014). Jacobsen et al. (2014) complement these findings, showing that teachers also express great interest and excitement in their job in its own right. We therefore expect excitement in one’s job tasks to represent a relevant motive for teachers to engage in service delivery. For these reasons, (Danish) schools offer a very well-suited research setting for testing the relationships between transformational leadership, basic need satisfaction, PSM, and intrinsic motivation.
Data were collected as a nationwide questionnaire survey distributed to all school principals in Denmark (N = 1,059). As most items originate from English, the items were translated to Danish and translated back to English by two language editors independently. Items were then pretested by a small group of professionals in the health care sector in Denmark. The questionnaire was distributed as an Internet-based survey in May–April 2014, and 482 school principals completed the questionnaire. Two hundred nineteen principals provided contact information on their 8,174 employees. A survey to the teachers—the basis for this study—was distributed in August–September 2014, and yielded 2,730 complete answers or a response rate of 33.4%. One school was excluded because less than five teachers completed the questionnaire. Retaining only respondents with complete and valid answers for the items on our focal constructs and control variables (teacher gender, age, job tenure), 1,486 teachers distributed to 129 schools were retained for analysis of the structural model. A total of 87% of the teachers were employed in public schools, 71% were female, and respondents were on average 46 years old with an average tenure in their current job position of approximately 6 years.
It is important to note that self-selection potentially limits the generalizability of our results. Self-selection associated with the sampling procedure renders it likely that the investigated group of teachers differs in systematic ways from the population of Danish teachers. Although a leader-based e-mail was distributed to all Danish principals, this was part of an invitation to participate in a free leadership training course, and only leaders who volunteered to participate were asked to provide contact information on the employees. This does not invalidate our coefficient estimates, but it does question whether we can generalize the specific results beyond the current sample. We discuss this issue and other limitations of our study in the “Discussion” section.
Measurement and Validation
To the extent possible, the article employs measures that have already been validated in previous studies. Appendix Table A1 presents all questionnaire items and central descriptive statistics. Responses to all questionnaire items follow a 5-point Likert-type scale ranging from 1, strongly disagree, to 5, strongly agree. First, we present the measure of each theoretical construct and its psychometric properties based on confirmatory factor analysis. Second, we examine the validity of the overall measurement model. In this second step, all latent constructs are included simultaneously (see Gould-Williams, Mostafa, & Bottomley, 2015, for a similar approach).
Transformational leadership
We use a set of modified items, partly inspired by existing studies from the management literature (MacKenzie, Podsakoff, & Rich, 2001; Podsakoff et al., 1996) and public management research (Moynihan, Pandey, & Wright, 2012), for two reasons: First, our measurement instrument should tap into the full set of transformational leadership behaviors (developing, sharing, and sustaining a vision for the organization) as conceptualized in the theoretical section, and the original items in the Multifactor Leadership Questionnaire (MLQ) do not encompass this. Second, the MLQ has been criticized for confounding operational terms with their proposed effects on employees and for poorly discriminating between different dimensions (Van Knippenberg & Sitkin, 2013). This is a serious critique, because we risk attributing transformational leadership with positive effects (e.g., raising employee motivation) per definition. For this reason, we use a four-item measure developed and validated by Jensen et al. (2016; see Table A1 for item wording).
The four-item model taps into the full set of transformational leadership behaviors as perceived by the employees and, according to confirmatory factor analysis, replicates our data well: χ2(2) = 3.10, p > 0.05 (comparative fit index [CFI] = 1.00, root mean square error of approximation [RMSEA] = 0.02, and standardized root mean square residual [SRMR] = 0.01; Williams, Vandenberg, & Edwards, 2009). 1 Standardized factor loadings (λ) are all significant and high: mean λ = 0.82, λlow = 0.75. Reliability scores display high internal consistency with Cronbach’s alpha of 0.89.
Basic need satisfaction
To measure basic need satisfaction, we draw on the “Basic Need Satisfaction at Work Scale” used by Deci et al. (2001) and a recent modification of these items (Vandenabeele, 2014). Based on the results of initial confirmatory factor analysis, six items are included to capture the satisfaction of the basic needs for autonomy, competence, and relatedness, respectively. The three subscales fit the data well, as indicated by the chi-square test of exact fit and absolute fit indices: χ2(6) = 8.32, p > 0.05 (CFI = 1.00, RMSEA = 0.01, SRMR = 0.01). The measurement model displays convergent validity with high standardized factor loadings: mean λautonomy = 0.88, mean λcompetence = 0.74, and mean λrelatedness = 0.78, and internal consistency with Cronbach’s alpha ranging from 0.70 to 0.87.
Intrinsic motivation
Intrinsic motivation is measured using three items based on Jacobsen et al.’s (2014) work on intrinsic motivation among teachers in Danish schools. Items tap into the enjoyment and excitement of performing one’s job. Cronbach’s alpha for the three-item scale is 0.84.
PSM
PSM is measured using items that have proven valid and reliable in existing studies (e.g., Andersen et al., 2014; Wright, Christensen, & Pandey, 2013). PSM can be seen as a second-order construct comprising four first-order factors/dimensions: “compassion,” “commitment to the public interest,” “self-sacrifice,” and “attraction to policymaking” (Kim & Vandenabeele, 2010; Perry, 1996). This operationalization is consistent with the common conceptualization of PSM as resting on affective, normative, and instrumental motives for public service delivery (Perry & Wise, 1990). The four-factor model displays acceptable fit to our data: χ2(38) = 132.37, p < .001 (CFI = 0.95, RMSEA = 0.03, SRMR = 0.06). All standardized factor loadings are significant and high: λmean = 0.71, λlow = 0.55. Cronbach’s alpha for the four PSM dimensions ranges between 0.69 2 and 0.80. As we do not have different expectations for the different dimensions of PSM, we focus on PSM as a single second-order construct. Hence, a composite PSM score is estimated for each respondent from the scores on the four first-order dimensions.
Full measurement model
To examine the validity of the overall measurement model, we included all latent constructs simultaneously and correlated these with each other to assess discriminant properties of each construct. The overall measurement model displays convergent validity with high standardized factor loadings: λmean = 0.79, λlow = 0.57, and acceptable fit to our data: χ2(216) = 573.96, p < 0.001 (CFI = 0.92, RMSEA = 0.03, SRMR = 0.08). To demonstrate discriminant validity, the average variance extracted (average variance shared by indicators of the same latent construct) should be greater than the variance shared with other latent constructs (Fornell & Larcker, 1981). Comparing the average variance extracted with the squared construct intercorrelations in Table 1, we note that average variances extracted are indeed greater for all latent constructs, suggesting that discriminant validity is present.
Intercorrelations and Estimates for Discriminant Validity and Reliability.
Note. Subdiagonal entries are correlations between latent constructs. Entries above the diagonal are the squared correlation estimates (shared variance). The first entry on the diagonal is Cronbach’s alpha for composite reliability. The second entry on the diagonal is the average variance extracted for each latent construct.
p < .05. **p < .01. ***p < .001.
Estimation Procedures and Robustness Checks
To estimate the complex mediation model, we use SEM. SEM has the distinct advantage of directly modeling measurement errors, which allows for stronger predictive power (Acock, 2013). We model all latent constructs simultaneously and specify correlations between related factors and paths between independent and dependent variables.
A challenge concerns the fact that all variables are measured in a single survey. Items on, for example, transformational leadership and PSM have positive connotations, and individual teachers may therefore answer in ways that conform to social norms rather than reflect their true attitudes (“social desirability bias”). Social desirability bias is one manifestation of a more general concern, namely bias due to the common source. Common source bias concerns a scenario where a nonrandom measurement error is shared among variables as a function of the same method—here a single questionnaire (Richardson, Simmering, & Sturman, 2009). To mitigate this issue, we aggregate teachers’ perceptions of school principals’ transformational leadership behaviors by school. Although common source bias may still arise because teachers share contextual characteristics related to the specific school and school principal (Favero & Bullock, 2015), using the school mean for the independent variable reduces individual-level common source bias and, thus, the risk of generating false positives.
Finally, we check the robustness of our results with a series of multilevel random effects models. These models offer the advantage of modeling the variation nested in different hierarchical levels in our data; that is, individual-level teacher variation within schools and organization-level variation between schools. The findings are referenced alongside the results in the next section, and may be found in full in Tables A2 and A3 in the supplemental appendix.
Findings
First, we present the results for the direct path from transformational leadership to intrinsic motivation and PSM (denoted “c’” in subsequent tables). Second, we proceed to each of the mediation terms; that is, the paths from transformational leadership to need satisfaction (H1, denoted “a”) and from need satisfaction to the two types of motivation (denoted “b”). Finally, we test for joint significance of the indirect paths to offer evidence of mediation (H2a and H2b). The structural model was tested using teachers’ gender, age, tenure in job position, and sector (indicator for private school) as control variables, and the structural model provides a good fit to our data: χ2(145) = 457.68, p < 0.001 (CFI = 0.98, RMSEA = 0.04, SRMR = 0.03). Parameter estimates for the SEM model are depicted in Figure 2.

Conceptual map.
Table 2 presents the standardized path coefficients from the structural equation model with jackknifed standard errors to account for clustering by schools (Cameron & Miller, 2015). Turning to the individual direct paths from transformational leadership to intrinsic motivation and PSM (denoted “c’” in Table 2), we note that our data offer partial support for direct paths from transformational leadership to the two types of autonomous work motivation. The coefficient for mean teacher perception of school principals’ transformational leadership behaviors is positive and statistically significant with respect to intrinsic motivation (β = 0.097, p < 0.05) but not with respect to PSM. These results indicate that teachers express greater enjoyment and excitement in their job when school principals rely on transformational leadership to a greater extent, but that there is no direct association between transformational leadership and teachers’ motivation to do good for others and society through their job.
Transformational Leadership, Basic Need Satisfaction, Intrinsic Motivation/Public Service Motivation.
Note. Standardized path coefficients (SE). Jackknife standard errors based on 129 clusters (schools) in data. For abbreviations, see Table 1. N = 1,481.
p < .05. ***p < .001.
To assess the hypotheses, we tested whether transformational leadership is directly related to satisfaction of the basic needs for autonomy, competence, and relatedness, and indirectly related to intrinsic motivation and PSM through these mechanisms. Here, we first observe positive and statistically significant direct paths from transformational leadership to the satisfaction of each of the three basic psychological needs (denoted “a” in Table 2): autonomy (β = 0.327, p < 0.001), competence (β = 0.077, p < 0.05), and relatedness (β = 0.084, p < 0.05). The results thus support H1: that transformational leadership may offer one way for leaders to satisfy employees’ basic psychological needs in the workplace. Prima facie evidence of mediation exists when “a” and “b” are statistically significant (Gould-Williams et al., 2015), and this is further supported in part when we look at the direct path from each of the basic needs to intrinsic motivation and PSM (denoted “b” in Table 2), respectively. Here, we find statistically significant direct paths from the satisfaction of the need for autonomy to intrinsic motivation (β = 0.164, p < 0.001), from the satisfaction of the need for competence to intrinsic motivation (β = 0.959, p < 0.001) and PSM (β = 0.145, p < 0.001), and from the satisfaction of the need for relatedness to PSM (β = 0.149, p < 0.001). Taken together, these results provide partial support for our argument that the relationship between transformational leadership and autonomous work motivation is mediated by the satisfaction of the basic needs for autonomy, competence, and relatedness. 3 However, the results also suggest that satisfying the needs for autonomy, competence, and relatedness is not necessarily equally important for intrinsic motivation and PSM, respectively. We return to this observation in the “Discussion” section below.
To evaluate mediation, Table 3 reports the coefficient estimates associated with the indirect paths (“a” × “b”) and their corresponding Sobel test score for statistical significance. Consistent with the evidence from Table 2, we find four statistically significant indirect paths. Specifically, and in regard to H2a, we observe that the relationship between transformational leadership and intrinsic motivation is mediated by the satisfaction of the needs for autonomy and competence but not by the satisfaction of the need for relatedness. About 35.8% and 43.3% of the total variance in intrinsic motivation explained by transformational leadership were accounted for by the satisfaction of the needs for autonomy and competence, suggesting that other important mediators need to be examined in future work. With regard to PSM and H2b, the relationship between transformational leadership and this type of autonomous work motivation is mediated by the satisfaction of the needs for competence and relatedness. The proportion of mediation 4 in these cases is greater than 0.5, indicating that the satisfaction of the needs for competence and relatedness is likely to be very important for understanding the link between transformational leadership and PSM.
Mediation Test: Standardized Path Coefficients for Indirect Effects of Transformational Leadership on Intrinsic Motivation and Public Service Motivation Through Basic Need Satisfaction (SE).
Note. Ratio of direct-to-total effects = 1 − ratio of indirect-to-total effects. For abbreviations, see Table 1.
p < .05. **p < .01.
Discussion and Conclusion
This article set out to examine whether transformational leadership relates to autonomous work motivation—intrinsic motivation and PSM—of public service providers by satisfying basic psychological needs for autonomy, competence, and relatedness in the workplace. Our empirical analysis reveals several interesting results: First, it supports H1 in finding a positive correlation between transformational leadership and the satisfaction of the needs for autonomy, competence, and relatedness. Second, the results offer partial support for H2a and H2b, namely, that the satisfaction of the basic needs mediates the relationship between transformational leadership and the two types of autonomous motivation. For intrinsic motivation, the mediation effect is statistically significant for the satisfaction of the needs for autonomy and competence but not for the satisfaction of the need for relatedness. These results corroborate the notion within SDT that satisfaction of the needs for autonomy (i.e., feeling like the initiator of one’s own actions) and competence is pivotal for the inherent interest in an activity, whereas satisfaction of the need for relatedness is less important (Deci & Ryan, 2000; Ryan & Deci, 2000b). Satisfying the need for relatedness—along with the need for competence—is important in regard to PSM, though. The finding that satisfaction of the need for autonomy does not seem to mediate the relationship between transformational leadership and PSM squares with previous research, which has not been able to show a positive association between the concepts (in fact, Vandenabeele, 2013 finds a negative association in his study of Belgian civil servants). Taken together, our findings thus suggest that transformational leaders may give weight to promoting the satisfaction of employees’ needs to feel autonomous, efficacious, or connected to other people in the workplace, depending on the desire to enhance employee intrinsic motivation or PSM. Importantly, however, in our study transformational leadership seems to increase the satisfaction of each of the three basic psychological needs, and it therefore offers a useful tool for managers to increase both intrinsic motivation and PSM.
Finally, the direct relationships between transformational leadership and the two types of work motivation should be considered. Our findings show a positive correlation between transformational leadership and intrinsic motivation, but the result for PSM is not as straightforward. We do not find a positive direct path from transformational leadership to PSM, as has been suggested by other studies (e.g., Wright et al., 2012). This may be somewhat surprising, given the strong arguments in public administration literature for the motivational effects of transformational leadership. According to Rainey and Steinbauer (1999), agencies will be able to motivate people to perform better when the mission is perceived as worthwhile, and this may be particularly relevant in public organizations characterized by strong service- and community-oriented visions (Wright et al., 2012) and a workforce motivated by public service (Perry & Wise, 1990). At least three potential explanations may account for this discrepancy.
First, endogeneity constitutes a concern. For example, leaders may choose to exercise transformational leadership behaviors in response to a workforce with low levels of PSM (for the very same reasons outlined above). Simultaneity could thus cancel out a positive direct association between transformational leadership and PSM. Moreover, potential selection biases may obscure this result. For example, highly transformational leaders may be allocated to turn around poor-performing schools (which are likely to have employees with lower levels of motivation as well), potentially canceling out a positive direct relationship between transformational leadership and PSM.
Second, studies on the relationship between transformational leadership and PSM have generally not taken contextual factors such as potential mediators or moderators into account. This is important, among other things, because Krogsgaard et al. (2014) recently showed that a relationship between transformational leadership and PSM only exists when managers and public service employees do not disagree on organizational values.
Third and finally, the discrepancy between results could potentially be attributed to the measurement of transformational leadership. Previous studies (e.g., Wright et al., 2012) measure transformational leadership at the individual level, whereas this study treats transformational leadership as an organization-level construct. The former approach may be reasonable, because individuals are likely to perceive the leadership behaviors of the same leader differently (Jacobsen & Andersen, 2015). Indeed, research shows that variation in employee perceptions of transformational leadership between comparable organizations is much smaller compared with variation within organizations (Jensen & Jacobsen, 2016), suggesting that much variation is purged from our data by aggregating individuals’ perceptions of transformational leadership. Certainly, this makes it more difficult to establish empirical relationships with PSM, which is an inherently individual-level phenomenon. For PSM, this argument would imply that the direct motivational effect of transformational leadership strongly depends on employees’ individual perceptions of their contribution to a society-oriented vision. 5 However, the use of individual perceptions of transformational leadership and PSM measured with survey data at single point in time may confound results because of common source bias (Meier & O’Toole, 2012). In this article, we prioritize this concern by aggregating teachers’ perceptions of school principals’ transformational leadership behaviors by school and present the findings for transformational leadership as an organization-level variable. Although common source bias may still arise from teachers sharing contextual characteristics related to the specific school and school principal (Favero & Bullock, 2015), aggregating on the independent variable reduces individual-level common source bias and, thus, the risk of generating false positives.
Limitations
The study has some limitations to which the reader should be attentive: First, transformational leadership, basic need satisfaction, and the two types of motivation are all measured at one point in time, and this makes it difficult to clearly establish the temporal sequence of the variables empirically. The article thus relies on its theoretical account to offer arguments for the specified model. The argument that transformational leaders increase the satisfaction of the basic psychological needs of their followers is also corroborated by existing experimental studies (see Kovjanic et al., 2013), but the link between the satisfaction of basic needs and motivation is more uncertain. Although SDT explicitly argues for the causal sequence proposed in this article by stating that “the satisfaction of basic psychological needs provides the nutriments for intrinsic motivation and internalization” (Gagné & Deci, 2005, p. 336), it is possible, for example, that highly motivated employees make a greater effort in their jobs, receive positive feedback, and, as a consequence, to a greater extent experience their need to feel competent as fulfilled.
Second, the findings on the relationship between need satisfaction and the two types of motivation could be inflated because of common source bias. While the aggregation of transformational leadership decreases such bias in correlations between this concept and the endogenous variables (i.e., the satisfaction of the needs and the two types of motivation, respectively), both need satisfaction and motivation are psychological factors embedded within individuals and, therefore, inherently difficult to measure more objectively or through external data sources. Future studies could attempt to accommodate the problems of endogeneity, such as simultaneity or common source bias, by using experimental designs. Such studies could attempt to manipulate the perceived need satisfaction by comparing levels of motivation between treated and nontreated groups. Still, it is important to note that the aim of this article is to strengthen our knowledge of the mechanisms between transformational leadership and the two types of motivation, and the results generally seem to support our theoretical expectations.
Third, the sample only consists of teachers in Denmark whose leaders (school principals) volunteered to participate in a free leadership training program. This could be problematic in terms of generalizability to the population of teachers whose leaders did not sign up for the program; to employees with different types of jobs; and to other cultures. Regarding generalizability to the population of teachers, a selection bias may, on one hand, exist in the sense that leaders who volunteer underperform compared with nonparticipants and look for ways to strengthen their leadership skills. On the other hand, it may be the case that leaders who already exert higher levels of leadership or have more resources (e.g., time) select into the program. Different types of self-selection potentially have different implications for employee motivation (i.e., lower or higher depending on existing leadership behaviors or organizational resources), and if both types of self-selection take place, their effects may cancel each other out. In the latter case, it will be less problematic for the generalizability of our findings to the population of Danish schools, but we cannot be sure that such selection biases exist and exactly how they look in the present case. Regarding generalizability to nonteachers, the impact of transformational leadership on motivation may differ depending on the kind of work undertaken by different employees. Existing studies already suggest that the correlation between transformational leadership and PSM is stronger when employees are not already aware of the prosocial impact of their job—and that such perceptions are lowest in jobs with low levels of citizen contact (Bro, Andersen, & Bøllingtoft, 2016). However, the fact that teaching is a relatively “high citizen contact job” only makes it more likely that our findings hold in sectors where this is not the case. This is the opposite of what could be the case when comparing different national cultures. Differentiating, for example, between collectivistic cultures (understood as the extent to which individual goals are more aligned with those of the collective) and individualistic cultures (understood as the extent to which members of a group are individualistic in their goals and objectives in life; Hofstede, 1980, 1991), an argument could be made that the contribution of leadership to, for example, the satisfaction of the need for relatedness is weaker in collectivistic cultures, simply because this need is already fulfilled by the work context (Kovjanic et al., 2012). As we have little knowledge of the impact of potential sector or culture differences on the motivational effects of transformational leadership, we encourage future studies to take up such issues.
A final potential limitation on the generalizability of our findings concerns particular contextual factors. The article finds a positive correlation between transformational leadership and satisfaction of the need for autonomy, but this finding may be ascribed to a systematic tendency for employees in our sample to agree with the vision espoused by their leader. Employees may experience a diminished feeling of autonomy if they disagree with the vision, or even if they agree with the vision, but feel that a previous self-determined will to work toward such ends is now imposed from above (i.e., through the vision). In fact, one previous study of the relationship between transformational leadership and PSM (Krogsgaard et al., 2014) suggests that the relationship depends on the level of value conflict between the leader and the employees. Specifically, the authors find a positive relationship between transformational leadership and PSM when there is little to no conflict on organizational values, but no relationship when leaders and employees hold very different conceptions of what is desirable for the organization. Hence, yet another interesting and important area of future research points to the contextual factors, such as value or goal conflict, that affect the relationships between transformational leadership, basic need satisfaction, and employee motivation.
Implications
As underlined by recent public administration studies, PSM is an important type of motivation to stimulate when aiming to increase performance in the public sector. The article not only corroborates existing studies on transformational leadership as an important antecedent in this respect (Krogsgaard et al., 2014; Park & Rainey, 2008; Vandenabeele, 2014; Wright et al., 2012), but it also sheds light on the psychological mechanisms underlying this relationship and compares different types of autonomous motivation. Differentiating between intrinsic motivation and PSM—and understanding the mediating mechanisms that link transformational leadership and autonomous motivation—is important, because not all tasks are necessarily easy to connect to a social purpose or contribution, just as not all tasks are inherently interesting. If managers aim to stimulate motivation based on inherent joy and excitement, it thus seems especially relevant to focus on satisfying the needs for autonomy and competence; that is, for example, to offer employees a say in the performance of an activity and avoiding micromanagement as well as expressing confidence in the employees. While satisfying the need for competence is also relevant in regard to PSM, the really interesting implication of the findings is that intrinsic motivation does not necessarily require the transformational leader to establish in-group feelings among employees, just as PSM does not necessarily rely on a feeling of self-determination in work activities. These results not only point to the theoretical value of integrating theories of self-determination, transformational leadership, and PSM but also to the practical significance of distinguishing between different task characteristics. When tasks are easy to connect to external prosocial consequences, managers should thus work to increase feelings of competence and relatedness among employees, while tasks that do not exhibit such characteristics require relatively more attention to instigate a feeling of autonomy among employees.
Finally, the study draws lines to research on how managers can ensure that employees perceive other managerial tools, such as pecuniary rewards or command systems, as supportive of their work. Motivation crowding theory (Frey, 1997) argues that perceptions of monetary incentives (associated with transactional leadership) depend on the extent to which these tools satisfy employees’ basic psychological needs for autonomy, competence, and relatedness. In light of our findings, transformational leaders may use visions to support perceptions of other managerial tools, such as pecuniary incentive systems, because visions illustrate how more specific goals contribute to the organization’s purpose and broader social contributions. This is in line with other scholars suggesting augmented effects of transformational and transactional leadership (e.g., Hater & Bass, 1988), though very little research has investigated whether the combination of transformational and transactional leadership indeed stimulates even higher levels of PSM and intrinsic motivation. Scholars are therefore strongly encouraged to pursue such questions in future endeavors.
Footnotes
Appendix
Descriptive Statistics and Standardized Factor Loadings.
| Constructs and items | M | SD | Minimum | Maximum | λ |
|---|---|---|---|---|---|
| Transformational leadership | |||||
| Concretizes a clear vision for the school’s future | 3.25 | 0.494 | 1 | 5 | .820 |
| Seeks to make teachers accept common goals for the organization | 3.68 | 0.376 | 2.50 | 5 | .760 |
| Strives to get teachers to work together in the direction of the vision | 3.44 | 0.475 | 2 | 5 | .864 |
| Strives to clarify for teachers how they can contribute to achieve the school’s goals | 3.33 | 0.451 | 1.75 | 5 | .849 |
| Basic need satisfaction | |||||
| Autonomy | |||||
| I feel I can make a great deal of input in deciding how my job gets done | 3.90 | 1.01 | 1 | 5 | .911 |
| I have good opportunities to decide for myself how to go about my work | 3.83 | 1.03 | 1 | 5 | .831 |
| Competence | |||||
| I feel very competent when I am at work | 4.22 | 0.703 | 1 | 5 | .681 |
| Most days I feel a sense of accomplishment from working | 4.22 | 0.691 | 1 | 5 | .831 |
| Relatedness | |||||
| I really like the people I work with | 4.39 | 0.641 | 1 | 5 | .811 |
| I feel connected to the people I work with | 4.13 | 0.761 | 1 | 5 | .760 |
| Public service motivation | |||||
| Compassion | |||||
| It is difficult for me to contain my feelings when I see people in distress | 4.27 | 0.765 | 1 | 5 | .615 |
| For me, considering the welfare of others is very important | 4.26 | 0.693 | 1 | 5 | .756 |
| I am often reminded by daily events of how dependent we are on one another | 4.06 | 0.784 | 1 | 5 | .552 |
| Commitment to the public interest | |||||
| It is important for me to contribute to the common good | 4.23 | 0.641 | 1 | 5 | .786 |
| I consider public service my civic duty | 3.91 | 0.741 | 1 | 5 | .678 |
| Meaningful public service is very important to me | 4.24 | 0.674 | 1 | 5 | .619 |
| Self-sacrifice | |||||
| I believe in putting duty before self | 3.06 | 0.917 | 1 | 5 | .672 |
| I am willing to risk personal loss to help society | 3.23 | 0.924 | 1 | 5 | .837 |
| I am prepared to make sacrifices for the good of society | 3.09 | 0.936 | 1 | 5 | .771 |
| Attraction to policy making | |||||
| I generally associate politics with something positive | 2.96 | 1.02 | 1 | 5 | .675 |
| I do not care much for politicians (reversed) | 3.32 | 1.08 | 1 | 5 | .847 |
| Intrinsic motivation | |||||
| I very much enjoy my daily work | 3.97 | 0.885 | 1 | 5 | .864 |
| My work is very exciting | 4.15 | 0.773 | 1 | 5 | .836 |
| I like performing most of my work processes | 4.21 | 0.698 | 1 | 5 | .781 |
Note. λ = standardized factor loading from confirmatory factor analysis.
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: This study was supported by the Danish Council for Independent Research Grant 1327-00015B.
Notes
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
