Teacher Leadership, Wellbeing, and Intent to Leave in US Rural Schools: Evidence from the 2020–21 National Teacher and Principal Survey

Abstract

Leadership in rural schools is well studied. However, few studies have investigated how teacher leadership influences rural teacher job satisfaction, burnout, and attrition. This study aimed to fill the research gap by analyzing data from the 2020–21 National Teacher and Principal Survey (NTPS), focusing on U.S. rural teachers. Using a sample of approximately 8910 rural teachers selected through a two-stage sampling process, this research examines the associations between teacher leadership, teacher wellbeing, and their intent to leave. The survey, administered by the National Center for Educational Statistics (NCES), included measures of teacher leadership (instructional and non-instructional), job satisfaction, burnout, and intent to leave. The authors found that (a) rural teachers with higher levels of teacher leadership reported higher levels of job satisfaction and lower levels of burnout, (b) rural teachers with higher levels of job satisfaction or lower levels of burnout are less likely to leave or move, and (c) instructional and particularly non-instructional teacher leadership are indirectly linked to teacher intent to leave or move through job satisfaction and burnout. The discussions, conclusions, and implications for school policy and school leadership are presented.

1. Introduction

Ensuring a consistent presence of qualified teachers is crucial for both the immediate and sustained success of schools. In light of the prevailing teacher shortages in U.S. schools, particularly pronounced in rural areas, school leaders must adapt their management approaches to mitigate teacher burnout (TB) and uphold teacher job satisfaction (TJS) [1]. Rural schools, grappling with unique challenges such as geographical isolation, limited access to teacher preparation programs, and professional seclusion, face heightened difficulties in recruiting and retaining highly qualified staff [2]. Administrators in rural schools shoulder an extensive array of responsibilities, including policy oversight, budget management, and staff maintenance, often compounded by the lack of administrative support. The strain on these administrators could be alleviated by exploring shared leadership models, particularly with teaching staff, in smaller schools where administrators might be the sole authority in a building or district.

This study seeks to address existing gaps in research by examining the impact of teacher leadership on TJS, TB, and teacher intent to leave or move (TILM) in the context of rural schools. Delegating essential tasks becomes imperative in smaller administrative settings [3], and this study explores both instructional and non-instructional teacher leadership dimensions, acknowledging the potential benefits teachers gain from participating in leadership opportunities.

Given the persistent teacher shortages in rural areas [4], it becomes crucial for rural schools to explore initiatives that enhance teacher retention. This study aims to investigate whether teacher leadership practices influence teacher satisfaction and burnout in rural schools, subsequently impacting teacher attrition through satisfaction and burnout pathways. If established, this influence could guide school administrators in deciding whether to invest in promoting teacher leadership in rural schools.

The study is guided by two research questions:

• To what extent do instructional and non-instructional teacher leadership influence TJS and TB in the rural school setting?
• To what extent do instructional and non-instructional teacher leadership affect TILM through TJS and TB in the rural school setting?

2. Literature Review and Conceptual Framework

2.1. Teaching in Rural Schools

Rural schools have been referred to as the forgotten minority in US education [5]. This may be attributed to the trend of urbanization in modern American history and shrinking rural populations [6]. Nonetheless, rural teachers have a direct instructional responsibility to one in five American students [7]. While only 20% of students are found in rural schools, the expectations for schools to maintain high-quality teaching staff and meet academic standards remain consistent with the demands placed upon schools in urban settings.

An analysis of the existing literature describes the need to understand the feeling of ‘place’ when discussing rural teachers and rural education [5]. Rural places are defined in many ways, often classified based on isolation, whether isolated from urban areas, by the agricultural land surrounding a community, or by an idyllic idea of land in contrast to the urban environment [8]. This isolation marks the community in many ways, but it usually means a greater sense of community and belonging between residents, which can be appealing for rural teachers [9]. Other scholarships have examined the social construction of rural places and defined “rural” in relation to other spaces whether urban and rural or near and far [10]. Place in terms of rural education has come to be understood more as a localized approach to educating students with awareness as to where they are both geographically and culturally.

These competing descriptions leave the definition of rural unclear, or at least very broad, but Brown and Schafft [8] attempt to clarify some research-based general characteristics of rural communities as follows. Generally, rural communities have smaller and less dense populations than non-rural communities. Economics of rural communities are heavily dependent on agriculture, raw materials, and other blue-collar occupations, and the school is often the leading employer. Rural communities generally have fewer institutional options for religious organizations, clubs, activities, and school choice, meaning each has more influence than in non-rural communities.

The US Census Bureau defines rural broadly as “what is not urban” [11] (p. 3) and further by distance from an urban center—fringe, distant, and remote [11]. In a report for the US Department of Education, Geverdt asserts that these location distinctions are based on “urban footprint [and] have been defined using measures based primarily on population counts and residential population density” [11] (p. 2) and “the term ‘rural’ encompasses all population, housing, and territory not included within an urban area” [11] (p. 3). Thus, any data collected by the US government or US Department of Education specifically define rural as a location which is apart from an urban area. This location-based approach is how rurality is defined for this study.

An area of focus for modern rural education is preparing teachers for service in rural schools. Barter points out the need to dispel misconceptions about teaching in rural areas as part of preservice training for teachers [12]. Threats of consolidation, insecurity in teacher allocations, and other factors are cited as issues that preservice teachers see in accepting roles as teachers in rural schools. In light of studies carried out on place and misconceptions of rural preservice teacher candidates, Hardre’s work showing greater comfort in rural teachers who were, themselves, raised in a rural community or the same rural community in which they teach is not surprising [2].

Resource scarcity and a lack of opportunities to grow or develop are also well cataloged in the literature. Rural teacher salaries are generally lower than non-rural salaries [13], putting rural schools at a hiring disadvantage relative to non-rural schools. Rural communities are often professionally tight-knit and culturally unique from one another [2]. The decision of a rural teacher to live in the community where they practice or not is a significant choice, as this will either increase or inhibit the ability of the teacher to live in and understand the unspoken culture of the community [9].

In the rural setting, one consequence of a smaller population is lower student and teacher census as compared to urbanized areas. Teaching in rural schools often leads teachers to feel a deep sense of connection and responsibility to the community in which they teach [14]. One study of rural teachers showed many teachers cited teaching groups from multiple ages/classes as a necessity given the small nature of the school [14]. There is also a challenge in small communities to gain ‘community knowledge’. As an outsider to a rural community, a teacher who is not from the same rural area where they are employed may face the added burden of gaining the trust of locals who have deep knowledge and understanding of the community [15]. The community knowledge aspect of rurality is echoed by research of the culture of rurality being more than just a population statistic but instead a culture with norms and values specific to the rural setting [16].

2.2. Teacher Job Satisfaction

TJS refers to the single dimensional measure of whether a teacher has a sense of satisfaction with their job [17]. According to American surveys, TJS hits an all-time low of around 12% in 2022 [18]. While it is possible that teachers may be able to make certain changes within their job that allow them greater enjoyment, many factors related to job satisfaction in education such as compensation, job duties, and management styles are outside the individual teacher’s control [17].

Both TJS and school/community population are common items found on most educator surveys. Understanding the setting of schools and the satisfaction of teachers are important drivers for research and innovation in education. Teacher workload, collaboration, and student discipline are three commonly cited factors for TJS [19]. Another common theme in research into TJS is self-efficacy. Klassen and Chiu found that higher levels of self-efficacy in both instruction and classroom management led to higher levels of TJS [20], and a meta-analysis of international data from 50 countries showed a unanimous positive and significant relationship between self-efficacy and TJS [21].

Teacher demographics and teaching experience can also affect TJS. Teachers with opportunities to continue professional development or complete a formal teacher induction program have significantly higher job satisfaction than those who do not [22]. Additionally, it has been shown that certain demographic variables may affect TJS. White teachers have been found to be significantly less satisfied than African American and Latino teachers in certain school environments [23]. Male teachers showed lesser satisfaction than their female counterparts [22].

Several indicators have been linked to TJS including many extrinsic motivators, such as compensation [24,25], and the availability of job resources, which has been shown with STEM teachers [26]. Though teachers can choose where they work to a certain extent, teacher–student ethno-racial mismatch negatively affects TJS, especially for White teachers [23]. The working environment also contributes to teachers’ feelings of job satisfaction; this includes feelings of autonomy, administration, and support from colleagues [27].

2.3. Teacher Burnout

Burnout is a well-studied topic in social sciences. In their work, Maslach and Leiter describe burnout as a psychological condition that develops over time due to persistent stressors in the workplace [28]. To measure burnout, Maslach and Jackson developed the Maslach Burnout Inventory (MBI) that measures burnout across three dimensions: emotional exhaustion, depersonalization, and lack of accomplishment [29]. Emotional exhaustion is characterized by feelings of depletion and emotional overextension due to work demands, leading to fatigue, debilitation, and loss of energy. Depersonalization is manifested as negativity and cynicism towards people, reflecting a shift away from empathy and care. Lack of accomplishment represents feelings of inadequacy and incompetence, diminishing their sense of professional effectiveness and achievement.

With a focus on teachers, Madigan and Kim identified three symptoms closely associated with TB: emotional exhaustion—individuals experience emotional overextension and exhaustion due to work-related stress; cynicism—cynical and impersonal reaction toward those around them; and reduced efficacy—a diminished sense of competence and success at work [17].

Distinct from TJS, burnout relates more to a teacher’s exhaustion and perception of no longer being able to physically or mentally complete the duties demanded of them. Recent research has found that burnout makes a larger difference in teacher intent to quit than TJS [17]. This phenomenon poses a significant threat to the education landscape, driving teachers out of the profession at alarming rates. Compared to job satisfaction, burnout plays a far more prominent role in influencing teachers’ intent to leave, ultimately impacting student learning and outcomes [17,30].

Many research articles have identified what could be considered as common strings in TB. Included in these are logistical challenges, red tape, disruptive students, lack of supervisor support, increased demands and duties placed upon teachers, dealing with colleagues, low self-esteem, role conflict, and poor working conditions [28,31,32]. Lack of support is of particular interest as this is a variable directly related to leadership within the school. Kyriacou suggests teacher leaders and administrators give thought to the way unnecessary stress may be created through poor management and consider a management style where “management decisions are based on consultation” [32] (p. 31).

2.4. Teacher Attrition

Employee retention is important for any organization. For an organization to be successful, it needs the ability to retain important personnel to complete key jobs. Employee retention has been linked to a variety of factors including job satisfaction, compensation, security, support of supervisors, participation in decision making, and more [33]. Hughes notes several associations with teacher turnover, including the prevalence of early-career teachers to leave the profession and for teachers with higher college entrance exam scores to be twice as likely to leave the profession [34]. These associations lead to knowledge and expertise gaps among teachers and highlight how much of the teacher turnover literature has focused on associations, not causes. Hanushek extended the evidence that less-tenured and ineffective teachers often leave the profession, but also showed that teacher turnover affects instruction in many schools even after accounting for other factors [35]. Thus, teacher retention in education is paramount as the teacher shortage continues to affect American schools.

Teacher turnover can be a confusing subject, with teachers leaving a particular job for other positions in the same or different district, family reasons, health reasons, etc.; a myriad of reasons can cause a teacher to leave. During the global COVID-19 pandemic, one longitudinal study surveying US teachers showed thoughts of turnover decreasing during the Fall 2020 semester, just as the effects of the pandemic were peaking within American schools [36]. Another study found that monetary incentives can cause qualified candidates to leave [37]. Thus, several factors can affect teachers’ intentions to move or leave, so here, we demonstrate a few connections that research has found, but it is in no way exhaustive.

Several school, student, and teacher factors affect teacher retention and teachers’ decisions to move or leave their building. Carver-Thomas and Darling-Hammond found that higher turnover rates are associated with everything from a lack of administrative support to teacher salaries, as well as alternative certification routes [38]. Rural schools generally see less teacher turnover than non-rural schools, but this is not the case for rural, predominately Black Georgia schools [13]. A study using national data found that the South is impacted most by teacher turnover in America [38].

Studies of teachers in multiple contexts have shown job satisfaction as a primary indicator and predictor of intent to leave in the job setting [39,40]. A quantitative analysis found job satisfaction as the primary and majority indicator of retention of Missouri elementary teachers [39]. Exploiting this relationship by increasing TJS is challenging, as noted earlier in this chapter. A meta-analysis of studies comparing job satisfaction and burnout as they relate to teacher attrition found burnout to have a greater effect on attrition than job satisfaction [17]. Hence, minimizing the effects of felt burnout is of primary importance to retain teachers.

2.5. Teacher Leadership

Over the past four decades, the concept of teacher leadership has been a recurring theme in the educational literature. York-Barr and Dukes conducted a comprehensive review spanning the first two decades of research. They defined teacher leadership as “the process by which teachers, individually or collectively, influence their colleagues, principals, and other members of school communities to improve teaching and learning practices, with the aim of enhancing student learning and achievement” [41] (p. 288). Silva et al. further traced the evolution of teacher leadership through three distinct waves: Wave 1: Formal Roles—teacher leadership was initially associated with specific formal positions within schools; Wave 2: Instructional Expertise—the focus shifted to recognizing teachers’ expertise in instructional practices; and Wave 3: Re-Culturing Process—teacher leadership emerged as a collaborative effort to transform school culture [42].

Beyond this historical context, researchers have explored the multidimensionality of teacher leadership. For instance, York-Barr and Duke summarized seven dimensions of teacher leadership practices (see

Table 1. Instructional and non-instructional dimensions of teacher leadership.

	Instructional Dimensions	Non-Instructional Dimensions
York-Barr and Duke [41]	(b) School or district curriculum work.	(a) Coordination and management; (c) Professional development of colleagues; (d) Participation in school change/improvement; (e) Parent and community involvement; (f) Contributions to the profession; (g) Preservice teacher education.
Teacher Leadership Exploratory Consortium [43]	(d) Improvements in instruction and student learning; (e) Promoting the use of assessments and data for school and district improvement.	(a) Fostering a collaborative culture; (b) Accessing and using research to improve practice and student achievement; (c) Promoting professional learning; (f) Improving outreach and collaboration with families and community; (g) Advocating for student learning and the profession.
Shen et al. [44]	(c) Facilitating improvements in curriculum, instruction, and assessment.	(a) Promoting a shared school vision, mission and goals of student learning; (b) Coordinating and managing beyond the classroom; (d) Promoting teachers’ professional development; (e) Engaging in policy and school decision making; (f) Improving outreach and collaboration with families and communities; (g) Fostering a collaborative culture in school.

) [41]. The Teacher Leadership Exploratory Consortium (2011) identified teacher leadership practices across seven domains (see Table 1) [43]. Shen et al. conducted a meta-analysis, consolidating teacher leadership dimensions reported in the literature. They proposed a new categorization into seven dimensions (see Table 1) [44].

Interestingly, these categorizations predominantly emphasize non-instructional leadership opportunities that extend beyond the traditional role of classroom instruction. While curriculum and instruction form only one dimension, the remaining six dimensions encompass tasks such as professional learning, coordination, and collaboration. This suggests that when researchers advocate for teacher leadership, they primarily refer to non-instructional aspects. In the subsequent section, we delve deeper into the literature, examining both instructional and non-instructional facets of teacher leadership.

2.5.1. Instructional Teacher Leadership

Instructional teacher leadership, a concept introduced in the early 2000s, has been advocated as a mechanism for enhancing instruction and alleviating principal workload [45,46,47,48]. Neumerski takes a distributed lens and identifies three distinct bodies within the instructional leadership literature: principal instructional leadership, teacher instructional leadership, and coach instructional leadership [49]. Examining the teacher leadership literature, Neumerski observed variations in the teacher leader role, ranging from full-time teachers assuming additional leadership tasks to part-time teacher leaders and full-time out-of-classroom teacher leaders. Notably, only a subset of teacher leadership research focuses explicitly on its relationship to instruction [49,50].

The documented positive outcomes of distributing leadership to teachers on curriculum and instruction-related issues are noteworthy. Studies, such as the one conducted in South Florida, demonstrate accelerated instructional results in teachers with leadership roles in curriculum and instruction [51]. Beyond the instructional benefits, distributing curriculum leadership to teachers also alleviates the growing complexity of principals’ roles, as reported by three out of four principals [51].

Teachers’ active involvement in curriculum decision making has been associated with successful collaboration, particularly within professional learning communities (PLCs). Abundant research guides teachers in making curriculum decisions and stresses the need for teacher-led conversations on curriculum choice and implementation in the classroom [52,53]. Furthermore, teacher leadership in collaborative efforts, such as PLCs, significantly enhances teacher collective efficacy [54]. Despite the positive impact, only 43% of teachers feel they have quality opportunities for weekly collaboration [55].

The exploration of instructional teacher leadership in rural school settings remains a gap in educational research. Often referred to as the “forgotten minority” in education research, rural schools lack substantial exploration of teachers’ curriculum leadership [56]. A study on instructional leadership in rural Chinese schools reveals a perceived deficiency in instructional leadership among teachers, along with a lack of perceived ability and tools to oversee instructional decision making successfully [56].

2.5.2. Non-Instructional Teacher Leadership

Leadership in the school setting has been a subject of research for decades. In the past, the traditional “CEO”-type leader made all decisions, oversaw discipline, set policies, managed staff, and undertook various roles. If anyone wished to influence school decisions, they had to go through this singular leader. However, the narrative in leadership research has shifted with the advent of distributed leadership. In the school context, distributing leadership means delegating crucial non-instructional tasks to teachers, such as budgeting, hiring new staff, and setting discipline policies. Exploring innovative and effective leadership types benefits principals, teachers, and students alike, as leadership, as famously stated by Leithwood et al., is second only to instruction in its effect on student learning [57].

Non-instructional leadership among teachers is less prevalent than instructional leadership. Ingersoll et al. found that over 85% of teachers felt they had influence over instructional decisions, but only 6% felt the same about the school budget, 13% about hiring, and 36% about discipline policies [58]. This disparity in influence highlights a dichotomy between instructional and non-instructional teacher leadership. The scarcity of teacher influence in these non-instructional areas poses challenges for research on effectiveness. Therefore, further research on the impact of distributing non-instructional leadership duties to teachers is crucial to bridge this knowledge gap.

2.6. Conceptual Framework

This study explores the interplay between TL, TJS, TB, and TILM within the context of US rural schools. In this study, we operationalize TL as a teacher’s actual influence on school-level policies and decisions. This definition aligns with research suggesting that TL involves engagement in problem solving [59], participation in decision making [60], and influence on both school-level and broader policies [61]. Our measurement relies on the 2020–21 NTPS data, where teachers rated their influence on seven policy areas, including setting student performance standards, establishing curriculum, contributing to professional development programs, evaluating teachers, hiring new full-time teachers, setting discipline policy, and deciding the school budget.

Based on both literature analysis and factor analysis (details provided in the Methodology and Findings sections), we categorize these seven influence areas into two sub-constructs. The first one is instructional TL that focuses on policies directly impacting teaching and learning (e.g., standards, curriculum). The second one is non-instructional TL that encompasses broader school-wide policies (e.g., hiring, discipline, budget).

The literature examining the impact of TL suggests certain trends in the relationship between TL and TJS as well as TB. A comprehensive review conducted by York-Barr and Duke analyzed 100 TL studies published from 1980 to 2003 [41]. Their findings indicated a mixed impact of TL on teacher leaders, showing potential improvements in motivation and reduced isolation, but also highlighting the potential drawbacks such as role confusion and increased stress. However, it is crucial to note that the majority of the studies reviewed were small-scale and qualitative, leaving uncertainty about the extent of TL’s influence on TJS and TB.

Another extensive review by Wenner and Campbell focused on 72 TL studies published between 2004 and 2013 [61]. Their analysis categorized TL’s effects on teacher leaders into four main areas: stress, relationships with peers and administration, professional growth, and leadership capacity. The literature also points to a correlation between TJS, TB, and Teacher Intent to Leave the Profession (TILM). Notably, studies found that higher TJS was associated with lower anxiety [62] and a reduced inclination to leave the teaching profession [63].

Drawing from the existing literature, we propose our conceptual framework, illustrated in Figure 1. According to our conceptualization, the two TL constructs exert direct effects on TJS and TB, subsequently influencing TILM.

3. Methodology

3.1. Data Source and Sample

This study utilized data obtained from the 2020–21 National Teacher Principal Survey (NTPS) [64] sponsored by the National Center for Educational Statistics (NCES). In all, about 9900 public schools and their principals were sampled as the first stage for the 2020–21 NTPS. From the sampled schools, about 68,300 teachers were sampled as the second stage. From the sampled schools and teachers, this study included about 2400 rural schools and about 8910 rural teachers. Among them, about 3980 are female and 3560 are male; about 7990 are White and 920 are minority; and about 100 rural teachers hold associate degrees, about 3870 bachelor’s degrees, about 4220 Master’s degrees, about 650 specialist degrees, and about 80 doctoral degrees.

3.2. Measures and Latent Constructs

3.2.1. Dependent Variables

The three dependent latent variables (TJS, TB, and TILM) are based on seven survey items from the NTPS teacher survey (Appendix A) with the question stem asking, “To what extent do you agree or disagree with each of the following statements?” Teachers’ responses are based on a 4-point Likert scale where 1 is “Strongly disagree” and 4 is “Strongly agree”. The seven statements include the following: (a) the stress and disappointments involved in teaching at this school aren’t really worth it; (b) the teachers at this school like being here; I would describe us as a satisfied group; (c) I like the way things are run at this school; (d) if I could get a higher paying job I’d leave teaching as soon as possible; (e) I think about transferring to another school; (f) I don’t seem to have as much enthusiasm now as I did when I began teaching; (g) I think about staying home from school because I’m just too tired to go. The measure of TJS is based on items (b) and (c), the measure of TB is based on items (f) and (g), and TILM is measured by items (a), (d), and (e) [64].

3.2.2. Independent Variables

The two independent latent variables (ITL and NITL) are based on seven survey items from the NTPS teacher survey (Appendix A) with the question stem asking, “How much actual influence do you think teachers have over school policy AT THIS SCHOOL in each of the following areas?” Teachers’ responses are based on a 4-point Likert scale where 1 is “no influence” and 4 is “a great deal of influence”. The seven school policies include (a) setting performance standards for students at this school, (b) establishing curriculum, (c) determining the content of in-service professional development programs, (d) evaluating teachers, (e) hiring new full-time teachers, (f) setting discipline policy, and (g) deciding how the school budget will be spent [64]. In this study, ITL is measured by the first two items while NITL is measured by the last five items.

3.2.3. Control Variables

This study included four teacher and school background variables. The first is teacher gender. It is coded as 1 = female and 0 = male. The second is teacher race and it is coded as 1 = minority and 0 = White non-Hispanic. The third is a teacher’s highest degree with 1 = associate, 2 = bachelor, 3 = Master’s, 4 = educational specialist, and 5 = doctoral degree. The last background variable is school enrollment. These variables are included to remove some confounding variables’ influence on the analysis. This helps avoid bias and improve the internal validity of this study [65].

3.3. Analytical Procedures

The authors took two steps to analyze the data and answer the research questions. Step one is to develop all latent measures by applying both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Step two is to apply structural equation modeling (SEM) to analyze the structural effects between the latent variables. In total, three SEM models were developed. The first SEM model includes TILM as the dependent variable and the four control variables. The second SEM model adds TJS and TB to predict TILM. Finally, the third SEM model adds ITL and NITL to predict TJS and TB. The three SEM models allow us to answer the questions and estimate the proportion of variance explained in the dependent variables.

4. Results

4.1. Results of Measurement Models

4.1.1. Instructional and Non-Instructional Teacher Leadership

Using EFA, seven indicators relating to teacher influence on specific aspects of school policies were loaded to two latent constructs. This was confirmed by CFA where all standardized factor loadings for indicators in the CFA model were at least 0.642, indicating strong relationships between the indicators and the two latent constructs ITL and NITL. Model fit indices for this two-factor CFA model are reported in

Table 2. Model fit indices.

Index	χ2 (df)	CFI	TLI	RMSEA (90% CI)	SRMR
ITL/NITL	268.87 (13)	0.98	0.97	0.047 [0.042, 0.052]	0.02
TJS/TB/TILM	472.91(11)	0.96	0.93	0.069 [0.063, 0.074]	0.04
SEM Model 1	35.01 (8)	0.99	0.99	0.020 [0.013, 0.026]	0.01
SEM Model 2	1708.69 (28)	0.88	0.80	0.082 [0.079, 0.086]	0.10
SEM Model 3	2119.73 (114)	0.93	0.92	0.045 [0.043, 0.046]	0.05

. The model fit indices indicate that the data fit well (CFI = 0.98; TLI = 0.97; RMSEA = 0.047; and SRMR = 0.02).

4.1.2. Teacher Burnout, Teacher Job Satisfaction, and Teacher Intent to Leave or Move

Using EFA, seven indicators were loaded to three distinct latent constructs: TJS, TB, and TILM. This was confirmed by CFA where all standardized factor loadings were at least 0.618, indicating strong relationships between the indicators and the three latent constructs. Model fit indices for this model and are reported on Table 2. The model fit indices for this CFA model indicate that the data fit well (CFI = 0.96; TLI = 0.93; RMSEA = 0.069; and SRMR = 0.04).

4.2. Results of Research Questions

The model fit indices of the three SEM models are presented in Table 2. The final model (Model 3) presented a good fit (CFI = 0.93, TLI = 0.92, RMSEA = 0.045, SRMR = 0.05). The standardized estimates of all three models are presented in

Table 3. Standardized estimates for the SEM models.

Independent Variable	Model 1 Estimate [95% CI]	Model 2 Estimate [95% CI]	Model 3 Estimate [95% CI]
Dependent Variable: TILM
Teacher Gender	0.035 [0.033, 0.038]	0.118 [0.116, 0.120]	0.106 [0.104, 0.108]
Teacher Race/Ethn	0.043 [0.041, 0.046]	0.057 [0.055, 0.059]	0.055 [0.054, 0.057]
Teacher Degree	0.019 [0.017, 0.021]	−0.036 [−0.038, −0.035]	−0.032 [−0.034, −0.031]
School Size	0.045 [0.042, 0.047]	−0.009 [−0.011, −0.008]	−0.005 [−0.007, 0.003]
TJS		−0.641 [−0.643, −0.638]	−0.568 [−0.570, −0.565]
TB		0.639 [0.637, 0.642]	0.592 [0.589, 0.594]
Dependent Variable: TJS
Teacher Gender		0.000 [−0.003, 0.002]	−0.034 [−0.036, −0.032]
Teacher Race/Ethn		−0.016 [−0.019, −0.013]	−0.020 [−0.023, −0.018]
Teacher Degree		−0.047 [−0.049, −0.045]	−0.026 [−0.028, −0.024]
School Size		−0.040 [−0.042, −0.037]	0.021 [0.019, 0.024]
ITL			0.178 [0.173, 0.184]
NITL			0.360 [0.355, 0.365]
Dependent Variable: TB
Teacher Gender		−0.123 [−0.125, −0.121]	−0.090 [−0.092, −0.088]
Teacher Race/Ethn		−0.031 [−0.033, −0.028]	−0.031 [−0.034, −0.029]
Teacher Degree		0.040 [0.042, 0.044]	0.028 [0.026, 0.030]
School Size		0.049 [0.051, 0.053]	0.005 [0.003, 0.007]
ITL			−0.190 [−0.195, −0.184]
NITL			−0.221 [−0.227, −0.216]
Proportion of Variance Explained
TILM	0.006 [0.006, 0.006]	0.816 [0.813, 0.819]	0.806 [0.803, 0.808]
TJS		0.004 [0.004, 0.005]	0.259 [0.257, 0.261]
TB		0.020 [0.019, 0.021]	0.157 [0.156, 0.159]

. By comparing the proportion of variance explained, the authors found that by adding TJS and TB, Model 2 explained about 81% of additional variance in TILM beyond Model 1, and that by adding the two TL constructs ITL and NITL, Model 3 explained 25% of additional variance in TJS, and about 13% of additional variance in TB, beyond Model 2.

The results of Model 3 showed that the effects of ITL and NITL on TJS were β = 0.178 (95% CI [0.173, 0.184]) and β = 0.360 (95% CI [0.355, 0.365]), indicating that higher levels of teacher leadership are associated with higher levels of job satisfaction. The results also showed that the effects of ITL and NITL on TB were β = −0.190 (95% CI [−0.195, −0.184]) and β = −0.221 (95% CI [−0.227, −0.216]), indicating that higher levels of teacher leadership are associated with lower levels of burnout; that is, as teachers are delegated greater responsibilities with either instructional or non-instructional duties, the likelihood of burnout decreases. Model 3 showed that after controlling for TL and other variables, both TJS and TB presented an impressive effect on TILM (TJS: β = −0.568, 95% CI [−0.570, −0.565]; TB: β = 0.592, 95% CI [0.589, 0.594]). Figure 2 shows the Model 3 results omitting the control variables.

Based on the above direct effects, the authors also estimated the indirect effects of the two TL constructs on TILM. The results are presented in

Table 4. Standardized estimates for the indirect effects of Model 3.

Independent Variable	Total Indirect [95% CI]	Indirect Effect Mediated by TB [95% CI]	Indirect Effect Mediated by TJS [95% CI]
ITL	−0.213 [−0.220, −0.207]	−0.112 [−0.116, −0.109]	−0.101 [−0.104, −0.099]
NITL	−0.335 [−0.341, −0.330]	−0.131 [−0.134, −0.128]	−0.204 [−0.207, −0.202]

. The results showed that ITL had a total indirect effect on TILM (β = −0.213, 95% CI [−0.220, −0.207]) and NITL had a total indirect effect on TILM (β = −0.335, 95% CI [−0.341, −0.330]). This shows that both ITL and NITL make a difference in TILM. Moreover, the results show that NITL has a greater influence on TILM when compared to ITL.

5. Discussions

5.1. Does TL Influence TJS in Rural Schools?

This study found that TL could make a big difference in rural TJS. Specifically, both TL practices, ITL and NITL, presented an impressive and positive effect on TJS. The findings are consistent with Chew’s and Andrews’ finding that TL roles gave teachers a strong sense of purpose and satisfaction [66]. Dierks et al. also found that about 62% of interviewed teacher leaders reported that TL provided them with both extrinsic and intrinsic rewards including satisfaction [67]. Marcionetti and Castelli found that an increase in self-efficacy can increase TJS [68]. It is possible that increases in TL opportunities provide teachers with more work engagement and are a demonstration of organizational support, both of which mediate the effects of burnout on TJS for rural teachers [69]. However, none of the above studies were based on rural schools. In contrast, a study on urban schools found that higher levels of teacher participation in school decision making are not associated with higher levels of TJS [70]. Therefore, school rurality might make a difference in TL’s effects [71].

Between the two TL practices, this study found that NITL’s effect doubles the effect of ITL, indicating that where rural TJS is concerned, non-instructional TL opportunities beyond curriculum and instruction are particularly helpful. In the literature, the non-instructional aspect of TL components has been mentioned or discussed [59,60], but its effect has not been examined. This study is one of the first that examined both instructional and non-instructional aspects of TL in the context of rural schools.

5.2. Does TL Influence TB in Rural Schools?

This study found that TL could help reduce TB in rural schools. Both TL practices (ITL and NITL) presented an impressive and negative effect on TJS, indicating that higher levels of TL are associated with lower levels of TB in rural schools. In the literature, however, the effect of TL on TB is mixed. Smylie pointed out that TL initiatives could be stress-inducing and stress-reducing as well [72]. York-Barr and Duke’s review found that the effects of TL on teacher leaders are not all positive; it could improve motivation and reduce isolation but also create role confusion and cause stress [41]. Wenner and Campbell’s review found that the negative effects of TL on teacher leaders include stress and difficulties due to additional responsibilities [61]. It is noted that teacher leader roles are usually formalized in larger, non-rural schools, but are not formalized in small rural schools. Again, the contrasting findings require further research to look into some moderating factors such as school rurality.

Both ITL and NITL can be viewed as adding to teachers’ workloads, which has been shown to increase feelings of burnout [73]. It is possible that this increase in workload is perceived differently for TL than, for example, more students in the classroom, and thus we see a different effect on TB. Additionally, TL may produce higher feelings of self-efficacy, which is known to reduce feelings of burnout in teachers [74]. Between the two TL practices, NITL’s effect is larger than the effect of ITL, indicating that when rural TB is concerned, teachers’ leadership opportunities beyond curriculum and classroom teaching are more helpful than their traditional role in leading curriculum and instruction.

5.3. Do TJS and TB Influence Teacher Attrition in Rural Schools?

The authors found that after controlling TL practices, both TJS and TB presented two impressive effects on TILM. This is consistent with Madigan and Kim’s meta-analysis in which they found that TJS showed a negative relationship with teachers’ intentions to quit, and that all three dimensions of burnout are positively associated with teachers’ intentions to quit [17]. However, the two effects in this study are much larger than the two estimated in Madigan and Kim’s meta-analysis: TJS’ effects are −0.57 vs. −0.40, and TB’s effects are 0.59 vs. 0.41 [17]. The differences might indicate that TJS and TB play a more important role in teacher retention in rural schools than in schools in general.

Besides the opposite directions, this study found that TB presented a larger effect on TILM than TJB, showing that burnout is more concerning than satisfaction when teacher retention is concerned. This is also consistent with Madigan and Kim’s finding that burnout plays a larger role than job satisfaction in teachers’ intention to leave the profession [17].

5.4. Does TL Influence Teacher Attrition in Rural Schools?

The authors found that TL could also make a big difference in rural teacher attrition. The results showed that through the two mediators TJS and TB, both ITL and NITL presented two indirect and impressive effects on TILM, respectively. This tells us that higher levels of ITL and NITL are indirectly associated with lower levels of ITL. The findings align well with Wenner and Campbell’s belief that TL appears to be a possible solution for the problem of teacher attrition [61].

Among the four indirect paths, the path NITL→TJS→TILM presented the largest effect, showing that NITL is most helpful with TILM and it works more effectively through improving TJS than reducing TB. Given the evidence from this study and that of Madigan and Kim that burnout plays a larger role than job satisfaction in teachers’ intention to leave the profession, promoting NITL and TJS could be a solution to combat TB’s effect on attrition [17].

6. Conclusions and Implications

6.1. Conclusions

Based on the analyses, we can answer the two research questions with the following conclusions. For research question 1, both instructional and non-instructional TL practices advance TJS and reduce TB in rural schools. Also, non-instructional TL practices are more effective than instructional TL practices regarding their effects on TJS and TB in rural schools. For the second research question, both instructional and non-instructional TL practices help mitigate teacher attrition in rural schools through TJS and TB, and NITL→TJS→TILM is the most promising path among all examined paths.

6.2. Implications

This study has several important implications for rural school teachers, administrators, and school policies. First, based on findings, the authors suggest that policy makers and administrators take more initiatives to develop TL in rural schools since it helps advance TJS, reducing TB and ITL. Teachers are traditionally hired to teach in the classroom. New policy could encourage or incentivize teachers to take on more school- or district-level leadership roles in either instructional or non-instructional domains. Second, since non-instructional leadership presented a much larger effect on all three outcomes, school administrators and new policies should encourage teachers to go beyond the instructional domain and take on more non-instructional leadership responsibilities. These responsibilities could include (a) deciding the content of teacher professional development programs, (b) hiring new teachers, (c) evaluating teachers, (d) setting discipline policy, and (e) deciding school budget, among others.

Overall, both instructional teacher leadership and non-instructional teacher leadership increase TJS and reduce TB. That is to say, the more teachers feel trusted with leadership roles related to the curriculum and management of a school, the more satisfied they feel and the less burnout they experience. Intuitively, teachers are more likely to stay put in a job where they feel satisfied and experience less burnout. These promising results demonstrate the need for school leaders to seriously consider delegating leadership responsibilities for the well-being of the teachers they rely upon for successful instruction.

7. Future Research

Teacher leadership in this study was split into two distinct categories: instructional teacher leadership and non-instructional teacher leadership. In the current literature, there is ample evidence to support the practice of instructional teacher leadership, but the literature pertaining to the effectiveness and implementation of distributing non-instructional leadership duties to teachers is limited. Similarly, it is understood that TJS and TB are two distinct constructs, and recent research has treated them as such [17]. The large effect sizes of NITL merit further investigation into the overall effects of involving teachers in non-instructional leadership responsibilities. Future studies would examine the effect of implementing specific non-instructional leadership duties, such as setting discipline policy or overseeing the hiring of new teachers, on TJS and burnout.

Future research into teacher leadership could also benefit from a multilevel approach. This study used a single-level SEM approach observing data gathered from teachers at an individual level. Research indicates that principals benefit from delegating, but research into the specifics of principal job satisfaction and intent to leave as it relates to teacher leadership duties is needed. Future studies might consider a multilevel approach where teacher and principal data are considered simultaneously for analysis. This, in turn, might help us to understand the perceptions of principals in schools utilizing a distributed leadership framework.