Introduction

The competitiveness and survival of work organizations relies on the capacity to react adequately to challenges caused by globalization and technological developments which affect behaviors of clients, customers, and competing organizations (Herrmann & Felfe, 2013; Messmann et al., 2017; Saeed et al., 2019). One category of work organizations are schools. Schools need innovative solutions as part of their classroom learning environments as well within the school organization to deal with social and technological developments and to provide sustained high-quality education (Thurlings et al., 2015; Vincent-Lancrin et al., 2019). Regarding this context, the OECD (2019) put forward the critical question how education will reinvent itself to anticipate to the major trends that are forming the future of our societies and to train learners for their future rather than the past (Vincent-Lancrin et al., 2019). As a response to this question, teachers need to contribute to the development of innovative solutions, which is referred to as innovative work behavior (IWB). Teachers’ IWB may range from applying upcoming technologies in the classroom to innovating the whole curriculum. Applying new technologies, for example, can be a challenge as teachers often do not know exactly if, when and how they can use it (Johnson et al., 2016). IWB is associated with professional development (Coetzer et al., 2020), employability (Lecat et al., 2018), career success (Dan et al., 2018), and for teachers also with increased student results (Ross & Bruce, 2007). However, negative effects, such as conflicts among colleagues, may also occur, for example, when an employee proposes ideas for changing established routines and by doing so challenges the established framework of theories and practices shared by his or her colleagues (Janssen, 2003). In addition, risks concerning the process and outcome of the innovation process need to be managed. For example, a prototype of teaching materials with promising technologies needs to be introduced with care. Such challenges and risks highlight the importance of the social context for overcoming or dealing with such hurdles and, thus, call for attention to social factors such as leadership.

In accordance with the leadership literature, support from colleagues and one’s immediate supervisor is crucial for enabling teachers to overcome hurdles and risks during the innovation process (Di Vincenzo & Iacopino, 2022; Le Hoang et al., 2021; Thurlings et al., 2015). These complementary aspects of support are pronounced in the constructs distributed leadership (DL) and leader–member exchange (LMX), respectively. DL is defined as employees’ taking initiative in the course of actions, which is granted or supported by the other members of a team while aligning the team activities (Thorpe et al., 2011). For example, a teacher who is good at keeping an overview of topics takes the lead in working towards better integrating topics into the curriculum. Other teachers support this teacher in doing so, while everyone simultaneously ensures that activities and goals are aligned. These kinds of leadership practices help teachers to realize innovative solutions at work. The relationship between DL from this theoretical perspective and IWB has not been investigated before, neither within nor outside the education sector. Complementary to DL, LMX refers to a dyadic relationship of high quality among leaders and employees based on sympathy, loyalty, and professional respect (Dienesch & Liden, 1986; Graen, 2004; Graen & Uhl-Bien, 1995; Somech, 2010). In education one may think of a teacher and a team or school leader who have similar ideas about the quality of education and who work closely together while introducing a technology-enhanced project-based learning environment. This relationship between LMX and IWB has previously been studied in education (Vermeulen et al., 2022; Zeng & Xu, 2020), but not in combination with DL.

Because DL and LMX may play such a vital role in fostering teachers’ IWB, it may also be important to investigate how these effects unfold, for example by zooming in on the function of motivational factors in these relationships (Javed et al., 2019, 2020). In particular, we build on self-determination theory (Deci & Ryan, 2000) which posits that positive work-related behaviours and outcomes require the satisfaction of the basic psychological needs for autonomy, competence, and relatedness (Gagné & Deci, 2005). For example, experiencing these needs may provide employees, such as teachers, with the motivational energy to endure the challenges and uncertainties associated with being actively involved in an innovation process (Bowers & Khorakian, 2014; Janssen, 2003; Kanter, 1988). Thus, we posit in the current contribution that the mechanism of basic psychological needs satisfaction (BPNS) provides a motivational explanation regarding the link between DL and LMX on the one hand and teachers’ IWB on the other (Coun et al., 2019; Liu et al., 2021a; Martin et al., 2016). It is surprising that in this regard BPNS has not yet been studied as a mediator, given that all behavior, including IWB, ultimately is driven by motivational factors.

Moreover, as IWB was predominantly studied in profit companies (Choi & Choi, 2014; Messmann et al., 2017), we conducted our study in non-profit organizations, more specifically in secondary education schools. Furthermore, as many studies in this context are cross-sectional in nature, we used a time-lag design. To conclude, the aim of our research was to examine BPNS as a mediator in the relation between DL and LMX on the one hand and teachers’ IWB on the other. To that end, we formulated several research questions that aim at disentangling this complex relationship by first viewing the respective connections of DL and LMX with IWB (i.e., RQ1 and RQ2) before introducing BPNS as a mediator (i.e., RQ3 and RQ4):

  • RQ1. What is the relation between DL and teachers’ IWB?

  • RQ2. What is the relation between LMX and teachers’ IWB?

  • RQ3. Does BPNS mediate the relation between DL and teachers’ IWB?

  • RQ4. Does BPNS mediate the relation between LMX and teachers’ IWB?

Conceptualizing IWB

Innovations can be products (e.g., devices such as an educational game) or processes (e.g., services or procedures such as assessment) which are new, adequate, applicable, and beneficial for realizing processes or tasks in a social or work context (Messmann et al., 2017; West & Farr, 1989). Individual contributions to the advancement of innovative products or processes are named IWB. In accordance with the literature (De Jong & Den Hartog, 2005; Kanter, 1988; Scott & Bruce, 1994; Thurlings et al., 2015; West & Farr, 1989), teachers’ IWB refers to the intentional exploration, generation, promotion, and realization of innovative ideas that may be favorable for individual teachers, work groups, or the entire school organization. Regarding opportunity exploration, it is required that innovating teachers make other members of the school organization pay attention to opportunities for renewal or to needs for new solutions for educational problems. Subsequently, idea generation is focused on the creation of corresponding concepts for improvements of processes or products, either in classrooms or in the wider school context. Furthermore, idea promotion is necessary to generate support within the organization from colleagues or the school leader to enable the realization of innovative ideas. Finally, idea realization encompasses the development of ideas into an actual prototype, such as an educational game, that can be tested and sustained in practice.

However, as throughout all these endeavors teachers need to deal with uncertainty about the consequences of decisions, the accomplishment of goals, the chance of success, and possible opposing reactions of others (Bowers & Khorakian, 2014; Janssen, 2003; Kanter, 1988; Messmann et al., 2022), personal and job-related resources, like social and motivational factors, may be facilitators of teachers’ IWB.

DL and LMX as facilitators of IWB

Practical and socio-political support offered in the work context is crucial for the feasibility of developing an innovation. In this respect, the daily interaction with colleagues and one’s immediate supervisor may contribute to a constructive work climate in which employees, such as teachers, feel safe to come up with critical issues or new ideas and in which they experience that support is available when hurdles have to be overcome (Messmann & Mulder, 2017; Thurlings et al., 2015). Taking into account the complementary roles of colleagues and supervisors, the two leadership constructs DL and LMX both highlight this pivotal role that exchange, support, and safety play for enhancing innovation processes.

DL can be viewed from two different perspectives (Tian et al., 2016). The first perspective sees DL as a formal distribution of leadership responsibilities (e.g., Hulpia et al., 2012). The second perspective sees DL more informally, that is, it focuses on daily practices in teams of teachers rather than on formal leadership responsibilities. We adhere to the second perspective and consider DL as a temporal and situational phenomenon that is built on informal roles rather than on formal positions (García Torres, 2019; Kelley & Dikkers, 2016) that occurs in concurrence with formal school leadership. Specifically, DL entails that employees, such as teachers, who are in some way connected to a group of colleagues take initiative in their course of actions and exert influence on group processes based on their own expertise, which in turn is granted and thus accepted by the other group members. In addition, it is a requirement for taking and granting temporary leadership that group activities share ‘common aspects’ and in this way are aligned (Thorpe et al., 2011, p. 245).

DL makes it possible that within a particular social work context a diversity of qualities (i.e., expertise, talents) can be put to good use at the very moment they are needed. This in turn creates a climate of readiness and safety that may be encouraging for each individual teacher to strive for an innovative solution when problems or challenges need to be tackled. For example, when a group of teachers that is used to distributing leadership responsibilities is confronted with a classroom-based problem. They may tend to combine their innovative ideas and efforts in the development of a learning environment that fits the interests and needs of students and is responsive of societal demands for education (Thurlings et al., 2015). This is supported by quantitative evidence from teachers’ work context indicating that teachers and their supervisor perceived DL as vital for educational change (Amels et al., 2020). Empirically, the specific relation between DL and teachers’ IWB has been established, although in these studies DL is defined differently than in our study, namely, as just the involvement of stakeholders in school decision-making (Buyukgoze et al., 2022; Lin, 2022). In addition, studies on similar leadership constructs outside the educational context showed that IWB may be enhanced if employees experience inclusive leadership (Javed et al., 2018, 2019, 2020; Portugal & Yukl, 1994) and shared leadership (Liu et al., 2021a; Svensson & Kang, 2019). Thus, we formulated the hypothesis:

  • Hypothesis 1: There will be a positive relationship between DL and teachers’ IWB.

Despite focusing on formal leadership, LMX similarly highlights the importance of social connections and addresses the quality of dyadic relations between leaders and their followers (Graen, 2004; Graen & Uhl-Bien, 1995). Specifically, LMX refers to relationships of high quality between leaders and employees that go beyond the minimum amount of formal exchanges (e.g., defined in employment contracts) and that are instead based on social exchange or exchanges of psychological benefits founded on sympathy, loyalty, and professional respect (Dienesch & Liden, 1986). Such high-quality relationships may lead to more extra role behavior of employees (Vermeulen et al., 2022). For example, teachers may be inclined to go beyond their role description and explore innovative solutions in the course of their work. In part, such efforts may be due to the fact that a vital relationship with their principle enables teachers to deal with the uncertainty associated with IWB. That is, teachers do not need to worry that their relationship will be at stake if during an innovation process something goes wrong. Conversely, low-quality LMX can lead to low self-esteem (Zeng & Xu, 2020) which may hinder IWB. In two studies in the context of teachers’ work the connection between LMX and IWB has already been demonstrated empirically (Vermeulen et al., 2022; Zeng & Xu, 2020). Accordingly, we hypothesized:

  • Hypothesis 2: There will be a positive relationship between LMX and teachers’ IWB.

BPNS as a Mediator between DL, LMX, and IWB

The role of DL and LMX as complementary aspects of leadership for creating a facilitative social work environment for IWB may be further explained by considering DL and LMX as sources of satisfaction of basic psychological needs as outlined in self-determination theory (Deci & Ryan, 2000; Gagné & Deci, 2005; Loopers et al., 2023). Perceived autonomy refers to the feeling of having control over one’s actions and contemplating these actions as a manifestation of the self. Perceived competence concerns feelings of accomplishment and effectiveness concerning one’s actions. And perceived relatedness means the feelings of closeness and integration when interacting with other persons. In accordance with self-determination theory, these basic psychological needs must be satisfied as a basis for the internalization of influences from the social work environment, the development of self-determined motivation, and ultimately for effective functioning, growth, and well-being (Deci & Ryan, 2000; Gagné & Deci, 2005; Van Nuland et al., 2012).

Experiencing autonomy, competence and relatedness may play a crucial role in enabling employees, such as teachers, to take initiative and become proactive (Messmann et al., 2022). For example, when teachers introduce new ideas into their workplace, it may often be that this introduction is not fully planned, which requires numerous just-in-time decisions and the opportunity to experiment. Feeling autonomous may be beneficial for this process as it enables teachers to make decisions and plans by themselves and experiment with ideas, for example for new teaching methods, in a self-determined way. Furthermore, as the development of innovative solutions is a complex process, feeling competent is an important requirement for mastering and enduring the cognitive demands and requirements of this process. That is, feeling confident about their competences will make it easier for teachers to further advance their knowledge and vision about new ideas and be more assertive during difficult phases of the innovation processes. Finally, feeling related to one’s immediate social work environment and experiencing the support of colleagues and the socio-political backing of one’s immediate supervisor represent powerful levers, for example when challenges and opposing social reactions at one’s schools need to be overcome during the innovation process. Empirical studies in teachers’ work context supported these assumptions about the link between BPNS and IWB (Klaeijsen, 2015; Messmann et al., 2022).

With respect to sources of BPNS, both DL and LMX may be important. With regard to DL, temporarily taking on leadership responsibilities amongst a group of colleagues, which involves both taking and granting control over decisions, may bolster teachers’ feeling of autonomy (Liu et al., 2021a). Furthermore, the flexible, relaxed, and mutually supportive working environment that surrounds DL practices allows teachers to demonstrate their diverse capabilities as well as to give and receive support and impactful feedback (Liu et al., 2021a). This, in turn, may bolster teachers’ sense of accomplishment and their feeling to be acknowledged for their competent actions. Altogether this fulfills teachers’ need for competence. Finally, as DL practices imply that individuals are tightly related with their peers and have the opportunity to discuss problems and joy, they may feel more inclined to form and nurture close social bonds with their colleagues (Coun et al., 2019). Thus, DL is also likely to bolster feelings of relatedness.

With regard to the provision of empirical findings, two studies revealed a positive relation between DL and teachers’ perception of autonomy (Lin, 2022; Liu et al., 2021b). In addition, the relationship between DL and perceived competence of teachers received support by a qualitative study (Aas & Paulsen, 2019). Although there seems to be no study on the relationship between DL and teachers’ relatedness, significant, yet weak relationships between shared leadership and the perceptions of autonomy and relatedness were reported in one study outside the educational sector (Coun et al., 2019). Thus, we formulated our third hypothesis:

  • Hypothesis 3: BPNS will mediate the relationship between DL and teachers’ IWB.

In addition to the above considerations about DL practice in teacher teams or other groups of employees, formal leadership may still play a crucial role for teams or groups (Binkhorst et al., 2018) and their perceptions of autonomy, competence, and relatedness (Rigby & Ryan, 2018). As far as LMX being a source of BPNS is concerned, having high-quality relationships with their immediate supervisor provides teachers with the necessary job discretion for making choices and decisions at work, thus fueling their sense of autonomy (Martin et al., 2016). Likewise, the fact that a relationship of high quality with one’s immediate supervisor is likely to include active help with problems and feedback that facilitates effective performance, enables teachers to develop confidence in their capabilities and experience feelings of competence. Finally, as LMX involves having a strong interpersonal relationship with their immediate supervisor, this may give teachers the feeling of psychological encouragement and safety that is necessary to build a sense of relatedness (Martin et al., 2016). Concerning empirical evidence, no studies were found that have examined the relationship between LMX and BPNS in teachers’ work context. Nevertheless, supportive evidence for this connection was provided by studies among employment counsellors (Dose et al., 2019) and working professionals (Graves & Luciano, 2013). Thus, we hypothesized:

  • Hypothesis 4: BPNS will mediate the relationship between LMX and teachers’ IWB.

Method

Sample and data collection

In order to address our research questions, a quantitative study with secondary education teachers in the Netherlands was carried out. The data collection was executed by a certified company for online marketing research operating in the Netherlands. In order to acquire participants for the study, a unique hyperlink to an online questionnaire was sent to a large national teacher database. In accordance with the causal order of the presumed connections among variables in our study and to reduce possible issues with common method variance (Podsakoff et al., 2003), a time-lag design with two points of measurement was implemented with four and a half months in between the data collections (i.e., T1 in December 2017: N = 654; T2 in mid April 2018: N = 293; response rate of the repeated measurement: 44.8%). Specifically, the predictor variables DL and LMX were measured at T1 and, thus, temporally separated from IWB which was measured at T2. Furthermore, BPNS was measured at T1 and T2 in order to allow investigating its mediating role more adequately than with just one measurement at either T1 or T2. Four and a half months were considered an appropriate time lag as it requires time for leadership practices to influence psychological processes such as BPNS and, in turn, work behaviors such as IWB. For information on background variables, see Table 1. The core characteristics of the current sample are in line with characteristics of secondary school teachers in the Netherlands who have an average age of approximately 44 years, a percentage of 55% female teachers and an average working time equaling approximately 79% of a full-time employment (Ministerie van Onderwijs, 2021).

Table 1 Sample characteristics

Measures

IWB

For measuring IWB, seven items were derived from De Jong and Den Hartog (2005) that assess individual contributions to developing innovative workplace solutions (e.g., ‘I take many initiatives for change in my work’, α = .93).

DL

For the measurement of DL, a scale developed by Evers et al., (n. d.) was used. The scale consists of nine items that tap whether individual employees engage in temporal leadership-in-action (three items; e.g., ‘In my work group we exert our influence on basis of the qualities we possess’); whether individual employees are granted to enact temporal leadership-in-action (three items; e.g., ‘In my work group we support colleagues who use their expertise’); and whether individual employees perceive alignment concerning goals, tasks and expectations in their social work environment (three items; e.g., ‘In my work group everyone’s tasks are clear’). Exploratory factor analysis (principal axis factoring, promax rotation, Kaiser criterion extraction) showed that all items loaded on a single factor (α = .93). In conjunction with the measurement of DL it was furthermore controlled for group type. That is, teachers were requested to indicate the organizational structure of the work group they are interacting with most frequently at work (i.e., 1 = ‘work team’, 2 = ‘subject area’, 3 = ‘department’, 4 = ‘work group’) and to relate the measurement items to the social interactions in this group. In addition, participants were asked to provide information about the corresponding group size.

LMX

For assessing LMX, a seven-item scale from Janssen and Van Yperen (2004) and used in Van Dam et al. (2013) was adapted. It captures whether the interaction of individual employees with their immediate supervisor is characterized by quality features such as support, trust, and appreciation (e.g., ‘My supervisor would be personally inclined to help me solve problems in my work.’, α = .96).

BPNS

For measuring BPNS a Dutch instrument developed by Kreijns et al., 2019 and further validated by Messmann et al. (2022) was used. It comprises three subscales for measuring perceived autonomy (four items; e.g., ‘In my work I can realize all my ideas’, α = .92/92 at T1/T2), perceived competence (four items; e.g., ‘In my work I always manage to complete difficult tasks’, α = .71/.75 at T1/T2), and perceived relatedness (four items; e.g., ‘In my work my colleagues and I accept each other for who we are’, α = .83/.87 at T1/T2).

The measures used for IWB, distributed leadership, and BPNS employed a seven-point Likert-type scale which ranged from 1 = ‘does not apply at all’ to 7 = ‘fully applies’ while the measure used for LMX contained a five-point Likert-type scale which ranged from 1 = ‘fully disagree’ to 5 = ‘fully agree’. The original language of all scales was Dutch.

Analyses

First, a descriptive analysis was conducted by examining means and standard deviations. Then, a correlation analysis (Pearson’s r) was carried out for all study variables. In addition, to identify bivariate effects of background variables (i.e., gender, age, working time, education, leadership responsibility, innovation participation, group type, and group size) on endogenous variables (i.e., BPNS and IWB), group comparisons (i.e., t-tests and analyses of variance) and correlation analyses were employed. For education and group type, dummy variables were used to assess effects of the single categories.

For hypothesis testing, SEM in Mplus 8 (Muthén & Muthén, 1998-2018) was carried out. As a first step, a confirmatory factor analysis (CFA) was conducted in which the measurement items were used as indicators of the underlying latent variables. During CFA, three separate measurement models were specified to test our hypotheses that all contained IWB measured at T2 and DL and LMX measured at T1 but that differed with respect to BPNS. That is, the first model did not contain BPNS, the second model contained BPNS measured at T1, and the third model contained BPNS measured at T2. Item parceling was employed to make the measurement model more parsimonious (Little et al., 2002). Subsequently, the measurement models were advanced into structural models by specifying directed paths among the study variables in line with the hypothesized effects. Accordingly, three separate structural models were specified, that is, one model without BPNS, one model with BPNS measured at T1, and one model with BPNS measured at T2. Finally, background variables were included in all models to control for significant effects of background variables on IWB. For evaluating model fit, recommended fit indices and cut-off criteria (Schermelleh-Engel et al., 2003) including the χ2-test statistic (0 ≤ χ2 ≤ 2df), the χ2/df-ratio (0 ≤ χ2/df ≤ 2), the comparative fit index (0.97 ≤ CFI ≤ 1.00), the non-normed fit index (0.97 ≤ NNFI ≤ 1.00), the root mean square error of approximation (0 ≤ RMSEA ≤ .05), and the standardized root mean square residual (0 ≤ SRMR ≤ .05) were employed. In order to evaluate the significance of effects, we considered the sizes and p-values (i.e., p < .05, .01, or .001) of the effects.

Results

Descriptives and correlations

Descriptive statistics (Table 2) showed that teachers frequently carried out activities relating to the development of innovative processes and products in their work context (M = 5.18, SD = 1.09). Furthermore, teachers reported that DL was a vital characteristic of their work context (M = 5.08, SD = 1.19). Regarding group type as a reference point for DL, 61.1% mentioned the group of teachers within their subject domain; 30.4% mentioned a work team they were part of; 5.8% mentioned the colleagues they worked with in the same department; and 2.7% did not provide any additional information on group type. On average, these groups consisted of roughly 12 members (M = 12.01, SD = 8.84). Moreover, the teachers perceived that the exchange with their leader (LMX) was of high-quality (M = 3.62, SD = 1.04). And concerning BPNS, T1 and T2 values showed that the teachers felt competent (M = 5.12/5.17, SD = 0.91/0.89), socially related (M = 5.02/5.02, SD = 1.11/1.18), and, to a somewhat lesser degree, autonomous at work (M = 4.57/4.65, SD = 1.32/1.31).

Table 2 Descriptives and correlations

Regarding correlations among the study variables (Table 2), both DL and LMX were substantially related to IWB. Likewise, significant correlations between IWB and perceived autonomy, competence, and relatedness measured at T1 and T2 were found. Regarding the link between leadership and BPNS measured at T1 and T2, DL most strongly correlated with perceived autonomy, followed by perceived relatedness, and perceived competence. A similar but slightly weaker correlational pattern was found for the relationship between LMX and the dimensions of BPNS measured at T1 and T2. With respect to inter-correlations, a moderate correlation was found between DL and LMX. Furthermore, strong correlations among the corresponding dimensions of BPNS indicated that teachers’ perceptions of autonomy, competence, and relatedness were stable over time. Furthermore, concerning connections among dimensions of BPNS measured at T1 and T2, moderate correlations of perceived autonomy with perceived competence and with perceived relatedness and a weak correlation between perceived competence and perceived relatedness were determined.

With regard to effects of background variables, concerning gender female teachers were more active in the development of innovative solutions (t = 2.59, p < .05). Moreover, teachers who had recently participated in an innovation process were substantially more active in the development of innovative products or processes at work (t = 4.19, p < .001).

Main results of hypothesis testing

For each of the three separate structural equation models that were specified for testing the assumed direct and indirect effects, a good model fit (cf. Schermelleh-Engel et al., 2003) was found (see Table 3).

Table 3 Fit indices for the three specified structural models

Hypotheses 1, 2, and 3 received empirical support, but Hypothesis 4 did not, which now will be explained in detail. As far as the relative importance of the two investigated aspects of leadership was concerned, the test of the first structural model (Table 4) showed that DL (β = .21, p < .01) as well as LMX (β = .22, p < .001) significantly predicted teachers’ IWB. Thus, Hypothesis 1 and 2 received full empirical support.

Table 4 Standardized estimates for effects of leadership on IWB

Concerning BPNS as a mediator in the relationships between distributed leadership and LMX on the one hand and IWB on the other, the links between the two investigated aspects of leadership and BPNS and between BPNS and IWB were inspected first. With respect to the leadership–BPNS link, the two structural models depicted in Fig. 1 (containing BPNS at T1 and T2, respectively) showed that DL significantly predicted teachers’ perceptions of autonomy (β = .53/.36, p < .001), competence (β = .37/.25, p < .001) and relatedness (β = .56/.36, p < .001). Conversely, LMX was only predictive of perceived autonomy (β = .36/.28, p < .001) and relatedness (β = .15/.15, p < .01/.05) while its link to perceived competence was not significant (β = .03/.09, ns). With regard to the BPNS–innovation link, the two structural models depicted in Fig. 1 show that perceived competence had the strongest predictive effect on IWB (β = .26/.40, p < .001) followed by perceived autonomy which had a significant predictive effect at T2 (β = .24, p < .001) but merely showed a positive tendency at T1 (β = .18, p < .10). Likewise, perceived relatedness had a predictive effect at T2 (β = .23, p < .001) but no effect at T1 (β = .00, ns). The explained variances for IWB were R2 = .27, p < .001 for the model containing BPNS T1 and R2 = .44, p < .001 for the model containing BPNS T2.

Fig. 1
figure 1

Standardized estimates for effects on IWB. Note. p < .10, *p < .05, **p < .01, ***p < .001. N = 293. Effects of background variables and zero-order correlations between exogenous variables are omitted. Indirect effects of distributed leadership (DL) and leader–member exchange (LMX) on innovative work behaviour (IWB) mediated by perceived autonomy (PA, T1/T2), perceived competence (PC, T1/T2), and perceived relatedness (PR, T1/T2): DL → PA → IWB, β = .10/.09**; DL → PC → IWB, β = .09**/.10**; DL → PR → IWB, β = .00, ns/.08**; LMX → PA → IWB, β = .06/.07**; LMX → PC → IWB, β = .01, ns/.04, ns; and LMX → PR → IWB, β = .00, ns/.03.

After inspecting these direct links, the hypothesized role of BPNS as a mediator was examined. To begin with, indirect effects of distributed leadership on IWB were mediated by perceived competence measured at T1 and T2 (β = .09/.10, p < .01), perceived autonomy measured at T2 (β = .09, p < .01), and perceived relatedness measured at T2 (β = .08, p < .01). In addition, a positive tendency was determined for an indirect effect mediated by perceived autonomy measured at T1 (β = .10, p < .10) while no indirect effect mediated by perceived relatedness measured at T1 was found (β = .00, ns). In addition, the direct effect of DL on IWB was close to zero (β = .02/-.04, ns), thus indicating full mediation. Accordingly, the study’s results with few exceptions supported the assumptions of Hypothesis 3.

By contrast, the effects of LMX on IWB were hardly mediated by BPNS. That is, only one significant indirect effect that was mediated by perceived autonomy measured at T2 (β = .07, p < .01) and positive tendencies for indirect effects mediated by perceived autonomy measured at T1 (β = .06, p < .10) and perceived relatedness measured at T2 (β = .03, p < .10) were found. However, no indirect effects were mediated by perceived competence (β = .01/.04, ns) and by perceived relatedness measured at T1 (β = .00, ns). These inconclusive findings together with the result that the direct effect of LMX remained significant in the mediation model containing BPNS T1 (β = .15, p < .05) and was not close to zero in the model containing BPNS T2 (β = .09, ns) indicated that BPNS hardly acts as mediator in the relation between LMX and IWB. Hence, the findings of the present study did not support Hypothesis 4.

Lastly, regarding the impact of background variables the outcomes of all three structural models showed that the engagement in IWB was higher for female teachers (β = .12/.13/.14, p < .05/.05/.01) and for teachers who had recently participated in an innovation process (β = .24/.22/.23, p < .001).

Discussion

The primary goal of this study was to investigate whether BPNS mediates predictive effects of DL and LMX on IWB. This is a crucial topic as IWB is essential to respond to changes caused by globalization and fast digital developments and therefore in the end the survival of organizations. IWB has been linked with positive consequences such as employability (Lecat et al., 2018) and career success (Dan et al., 2018) but also with negative effects such as conflicts with co-workers (Janssen, 2003). As the development of an innovation is a demanding socio-political process, the social work environment is an important resource for facilitating employees’ IWB. The current study investigated this topic from the angle of motivational effects that result from leadership practices in schools, integrating both leadership practices distributed among teachers and formal leadership practices taking place in the dyadic relation among teachers and their direct supervisor. This specific topic was addressed in a quantitative study with teachers working in secondary education employing a time-lag design. Based on the results we may conclude that BPNS mediates the relationship between DL and IWB but not that between LMX and IWB. This main conclusion will now be explained in detail.

To begin with, the results of our study support the assumption that DL enhances teachers’ IWB because when a group of teachers engages in DL practices this will create an innovative climate in which a diversity of qualities is readily available which in turn will encourage each group member to make an individual contribution to developing an innovative solution when needed (Amels et al., 2020; Thurlings et al., 2015). This result is consistent with studies defining DL as participation of stakeholders in school-related decisions (Buyukgoze et al., 2022; Lin, 2022) and related studies on inclusive and shared leadership outside the educational sector (Javed et al., 2018, 2019, 2020; Liu et al., 2021a; Svensson & Kang, 2019).

We also found support for the assumption that in addition to DL practices among teachers LMX as representation of vital formal leadership practices enhances teachers’ IWB because the uncertainty of innovation processes can be more easily managed if the relationship between teachers and their immediate supervisor is characterized by loyalty and professional respect and they are provided with adequate job latitude, resources, support, and access to information (Kim & Koo, 2017). This positive relationship aligns with previous studies in teachers’ work context (Vermeulen et al., 2022; Zeng & Xu, 2020).

Moreover, the outcomes of the present investigation shed light on BPNS as a mediator in the relation between leadership factors and teachers’ IWB. First, as far as the leadership–BPNS link was concerned, the current findings support the assumption that DL enables BPNS because the activities involved in enacting DL, such as initiating temporary leadership behavior based on one’s expertise and granting each other such a leadership role, allows teachers to engage in autonomous self-management, to gain a sense of competence through the acknowledgement by others, and to experience vital social connections at work (Coun et al., 2019; Liu et al., 2021a). These results align with prior research in which a relationship between DL and the perceptions of autonomy and competence of teachers (Aas & Paulsen, 2019; Lin, 2022; Liu et al., 2021b) and between shared leadership and perceived relatedness (Coun et al., 2019) was established. By contrast, our findings did not consistently support the assumption that LMX fuels teachers’ BPNS. That is, although we found a robust connection between LMX and perceived autonomy we only found a weak connection between LMX and perceived relatedness and no connection between LMX and perceived competence which contradicts previous evidence from studies outside the educational sector which reported positive relationships between LMX and all three basic psychological needs (Dose et al., 2019; Graves & Luciano, 2013). On the one hand, concerning perceived autonomy our findings are in line with our theoretical assumption that LMX enhances teachers’ perceived autonomy as this offers teachers with the necessary job discretion for making decisions at work (Martin et al., 2016). On the other hand, our findings suggest in contradiction to our assumption that the psychological encouragement and safety that is immanent to LMX contributes to teachers’ sense of relatedness only to a limited extent. A possible explanation could be that the measurement items for perceived relatedness focused on colleagues and not the supervisor. Also in opposition to our expectations, LMX did not bolster teachers’ perceived competence by being a source of active help when problems occur and by containing feedback that facilitates effective performance. This may be explained by the fact that teachers operate primarily autonomously in the classroom and, while it can be helpful to some extent to get help with problems and to receive feedback from one’s supervisor, ultimately teachers need to figure it out on their own in the classroom. These findings and the already discussed significant relationship between DL and perceived competence and relatedness suggest that teachers are likely to be closer to colleagues than to their supervisor.

Second, as far as the BPNS–innovation link was concerned, our findings with few exceptions are in accordance with self-determination theory (Deci & Ryan, 2000) and with earlier research (Klaeijsen, 2015) and support the assumption that teachers’ BPNS provides them with the motivational energy that is needed for initiating and enduring the development of innovative solutions at school. That is, if teachers perceive themselves as autonomous, competent and, socially related they seem to be motivated enough to deal with the uncertainties of innovation processes and, thus, perform more IWB (Bowers & Khorakian, 2014; Janssen, 2003). In accordance with the study of Messmann et al. (2022), our findings showed that the perception of competence was the most important factor of the three basic psychological needs. Conversely, perceived autonomy was less important which may be due to the fact that, as indicated, teachers are relatively used to experiencing freedom and choice in teaching and, as a consequence, differences in perceived autonomy do not matter as strongly with respect to starting an innovation process than does the feeling of competence. In a similar vein, perceived relatedness was less important which may also be due to the nature of teachers’ work which mainly consists of their individual classroom work, which in turn may have caused that in relation to innovative contributions they do not so much rely on their social work environment and more strongly on their own feelings of competence.

Third, as a consequence of these direct links between the leadership factors DL and LMX and BPNS and between BPNS and IWB, the current findings support the assumption that DL impacts teachers’ IWB through BPNS. Thus, it may be concluded that DL appears to be a powerful lever for facilitating teachers’ IWB through the motivational effects it exerts on the perceived need for competence and to lesser degrees their perceived needs for autonomy (see also Lin, 2022) and relatedness. Conversely, the assumption that LMX also exerts its impact through BPNS was hardly supported as only the perception of autonomy played a role for the link between LMX and IWB. Rather, our findings lend support to the conclusion that the connection between LMX and IWB is of a more direct nature.

In terms of practical implications, our study highlights that for fostering teachers’ IWB school leaders should not solely focus on the quality of their relationship with teachers but also create a culture in schools in which teachers feel confident to use their expertise and take responsibility for leadership in the social groups they are involved in at work as well as grant such leadership responsibilities to other group members. In this regard, previous research showed that being open to teachers’ ideas, showing trust in their expertise, and encouraging and enabling teachers to develop new ideas is the way for school leaders to achieve this goal (Hulsbos et al., 2016; Sijbom, 2013). Our study also implies that the work environment at school is important for enabling teachers to make experiences that contribute to BPNS, especially their feeling of competence as this may enhance their readiness for IWB. For example, school leaders could give positive feedback on teachers’ ideas and approaches for implementing these ideas. This in turn may help them to feel more confident about what they can accomplish and how to enhance their involvement in the development of innovative processes and products (Messmann et al. 2022).

All these practical implications are not only relevant for schools but for all kinds of profit and non-profit organizations in which vital leader–member interactions, distributing leadership practices, and the development of innovative solutions may occur. In accordance with the general relevance of the constructs tackled in the current study, the measures employed for capturing these constructs addressed interactions, needs, and behaviors from a general work-related perspective that would be applicable in any work context and not just in the teaching profession. Accordingly, our implications have meaning for researchers and practitioners in the wider industrial organizational community. Nevertheless, additional studies in other (non-profit) organizations are needed to corroborate our findings.

Limitations and implications for future research

Based on our study some points have to be considered in future research. First, our study design which involved a time-lag design to prevent common method bias (Podsakoff et al., 2003) has to be considered. While it was possible to collect data at two points of measurement and include a repeated measurement of BPNS, future studies should involve three points of measurement for all variables in this study. By this means, the causal order of the relationships between the constructs (i.e., leadership factors, BPNS, and IWB) could be ideally reflected, the reciprocity and stability of effects could be more fully examined, and potential changes could be investigated. The latter aspect is particularly important as IWB is a construct that may exert long-term effects on intra-individual developmental changes, such as careers (Schalk et al., 2011).

Second, our results concerning LMX suggest that other motivational factors than BPNS as well as cognitive factors could be important for the relationship between LMX and teachers’ IWB. As an illustration, one such cognitive factor may be the construct inquiry habit of mind which encompasses aspects such as valuing a deep understanding, reserving judgement, tolerating ambiguity, and posing increasingly focused questions from different perspectives (Vermeulen et al., 2022). Additionally, a valuable perspective could be to complement self-determination theory with the action regulation theory (Zacher & Frese, 2018) which is targeted at the management of goal-directed behavior in the workplace and which suggests that as part of this regulatory process individuals search the social work environment for action-relevant information or signs like support from one’s immediate supervisor for presenting and applying innovative ideas (Messmann et al., 2022; Zacher & Frese, 2018). Building on this perspective, it would be relevant to examine the role of social perceptions, such as participative safety (Messmann et al., 2017), in the link between leadership factors and IWB.

Finally, regarding generalizability the context of the current study needs to be considered. While the focus of our study was on secondary education, future research may complement this focus by investigating whether our findings also apply to other educational contexts such as primary, vocational, and higher education, to the work context of educators in other countries and across cultures, and to other professions.