“ Are We Architects or Construction Workers ? ” Re-Examining Teacher Autonomy and Turnover in Charter Schools

Charter school teachers nationwide expressed greater autonomy compared to traditional public school teachers at the turn of the century. But is this trend changing? The recent proliferation of Charter Management Organizations (CMOs), which often have prescriptive organizational models, has raised questions around how teachers perceive autonomy and control in these schools. Researchers consistently find that faculty input into decision-making has a strong influence on staff commitment and turnover. This exploratory, interview-based study critically examines why and how CMO teacher autonomy is linked to turnover. CMO teachers linked various concerns over autonomy, which was limited by their organization’s model, with decisions to leave their  jobs. Teachers with strong expectations for autonomy or ideas inconsistent with their school’s model experienced substantial conflict involving the discipline or socialization of students that affected their career decisions. Teachers’ desires to have a voice in issues surrounding students’ socialization ultimately hinged upon their skepticism that current practices were adequately preparing students for college and life.


Introduction
Autonomy is viewed as a cornerstone of the charter school movement.In theory, charter schools provide practitioners with increased autonomy from rules and regulations to enable the design of innovative programs and increase buy-in and commitment from school leaders, teachers, parents, and communities (Wohlstetter, Smith, & Farrell, 2013, p. 69).Related to autonomy is the idea that "local decision makers, mainly teachers, would become more involved in decisions related to schooling.This, in turn, would lead to greater teacher commitment to and ownership of the charter school" (Wohlstetter, Smith, & Farrell, 2013, p. 76).In this sense, teacher autonomy1 was an original goal behind the charter school concept because it was intended to increase teacher commitment and leverage their expertise as professionals (Kahlenberg, 2007).
Early case studies showed that teachers in start-up charter schools responded to increased responsibility with a sense of commitment and ownership --they were energized by the opportunity to help build and found schools (Vasudeva & Gratzik, 2002).Using data from 1999-2004, more recent studies found that charter school teachers perceive having more autonomy compared to traditional public school teachers (Ni, 2012;Renzulli, Parrott, & Beattie, 2011).However, Charter Management Organizations (CMOs) or organizations that manage multiple charter schools with a home office offering centralized support have proliferated in the last decade and now account for nearly a third of all charter schools (Farrell, Wohlstetter, & Smith, 2012).CMOs operate networks of charter schools and can be highly prescriptive about school policies and practices (Bulkley, 2005;Lake et al., 2010), causing some to question how CMO teachers perceive and experience autonomy (Scott & DiMartino, 2011).Teacher autonomy in charter schools is important to re-examine because of the strong relationship between teacher control over decision-making and teacher turnover (Ingersoll, 2003), the expansion of CMOs, and concerns about higher than average teacher turnover rates in CMOs (Torres, 2014).
This study uses interviews with CMO and non-CMO charter teachers to explain what areas they desired autonomy over, why, and how these preferences affected their job satisfaction and career decisions.I find that most CMO teachers expressed varying preferences for autonomy or control in relation to their satisfaction and decisions to leave their school.Some formed initial expectations that they would have the autonomy to determine school policies because they were recruited as "founding teachers" for new charter schools and based on their previous expertise or experience teaching.Teachers also had or developed differing ideas and preferences over how students should be socialized, disciplined, or taught that often clashed with their management organization's "model," leading to substantial internal and external conflict (e.g.disagreements with school leaders).I describe how these conflicts were associated with teachers' decisions to leave their positions in CMO charter schools.By contrast, teachers in non-CMO charter schools did not cite limited autonomy as a challenge.However, many described being left alone and in need of feedback and support around student discipline and classroom instruction, citing these challenges as a reason for leaving their teaching positions.This study adds to the knowledge base by describing why and how teacher autonomy is related to turnover in organizations that prescribe practices for local level practitioners.While most studies suggest that teacher burnout is a primary reason for leaving CMOs, particularly those with a "No Excuses" culture (Lake et al., 2010), my findings suggest that the limits on teacher autonomy that result from highly prescriptive educational models is a factor contributing to high turnover in CMOs.
To contextualize my findings, I first review what is known about how charter school teachers perceive autonomy, teacher turnover in charters and CMOs, and the relationship between autonomy and turnover.I also discuss relevant organizational and teacher characteristics that influence teachers' perceptions of autonomy, and then explain the study design, methods, and findings.I conclude by discussing the findings in relation to existing literature as well as implications for future research, practitioners, and policymakers.

Literature Review Teacher Autonomy in Charter Schools
Charter school teachers reported feeling more autonomous than traditional public school (TPS) teachers at the turn of the century.Compared to TPS teachers, charter school teachers in a nationally representative sample from the 1999-2000 Schools and Staffing Survey (SASS) were more satisfied with their jobs and this higher satisfaction was explained by increased autonomy over decisions in their classrooms and schools (Renzulli, Parrott, & Beattie, 2011).
One study using [2003][2004] SASS data that matched charter schools with similar traditional public schools revealed that charter teachers perceived similar levels of classroom autonomy, but felt they had significantly more influence over school policies and a heavier workload compared to their traditional public school counterparts (Ni, 2012).Both studies define school autonomy as faculty influence over school policies in seven key areas (setting student performance standards, curriculum, content of in-service programs, teacher evaluation, hiring teachers, school discipline policies, deciding spending of budget) and classroom autonomy in six areas (selecting instructional materials, content, topics and skills to be taught, teaching techniques, student evaluation, homework policy, and discipline of students in one's classroom).These studies aggregate these facets of autonomy, and are therefore limited in their ability to determine which aspects of school or classroom autonomy matter most to particular teachers in specific contexts, and why.
Qualitative studies on charter school teacher autonomy have focused on how teachers perceive and experience some of these aspects of autonomy, but not how these perceptions relate to teachers' career decisions.One study of forty teachers in six urban charter schools found that the majority of teachers enjoyed and possessed the freedom and autonomy to make individual decisions about instructional strategies, writing curriculum, purchasing needed materials, identifying professional development, and selecting other staff (Malloy & Wohlstetter, 2003).In case studies of four different charter schools, using interviews with forty teachers, Gawlik (2007) found that teachers who had prior experience working in traditional public schools all described having more autonomy in a charter school over the school budget, curriculum, and hiring and firing of personnel.While these teachers perceived having autonomy over school budget and personnel issues, curricular freedom varied based on an existing relationship with a school district in one case and a CMO that had strong control over curricular materials in another.
In case studies with six charter schools from three different for-profit Management Organizations, Bulkley (2005) found variation between all three companies in the degree of freedom and input afforded to the individual schools to determine their educational programs (defined as the school mission, curriculum, assessment, instruction/pedagogy, professional development, and internal school governance).In the company with the highest organizational influence, core decisions about the educational program were decided outside of the school building, but even in the company with the least influence, Bulkley (2005) argued: "there [were still] more decisions about the educational program made outside the school than for a charter school that does not work with a management company" (p.227).In an effort to build a clear and consistent mission and high quality schooling experiences across schools, these organizations necessarily grappled with the question of how and to what extent they should involve practitioners at the local level in decisionmaking.This dilemma is becoming increasingly salient in a rapidly evolving charter school landscape.Although by some estimates MOs only represented six percent of all charter schools in 2000 (Miron, 2010), they now represent about one-third of all charter schools (Farrell, Wohlstetter, & Smith, 2012).The tension between the degree of centralized decision-making at CMOs and teacher autonomy that charter practitioners are known to enjoy has potentially significant implications for turnover.
While most research on charter school teacher autonomy focuses primarily on teachers' ability to make decisions about the school's academic program (Bulkley, 2005;Gawlik, 2007;Malloy & Wohlstetter, 2003), compelling evidence demonstrates that the level of teacher control in decisions about how students are socialized may be a much better predictor of conflict and teacher turnover in schools more generally.Richard Ingersoll (2003) analyzed national longitudinal data and conducted case studies with four different high schools and found that high teacher control over social (i.e.school discipline policies) and administrative (i.e.scheduling) decisions are associated with much lower teacher turnover compared to schools that have low teacher control over the same decisions.Teacher control over student discipline policies and the socialization of students showed the strongest associations with school conflict, teacher engagement, and teacher turnover across schools.This finding has important implications for CMO models that often prescribe school-wide behavior policies and systems that all teachers in a school are responsible for implementing (see Woodworth et al., 2008;Lake et al., 2010;and Whitman, 2008 for descriptions).While school-wide behavioral systems are essential to achieving order and consistency, especially in urban schools, teachers must be involved in decisions about these systems to reduce the likelihood of conflict (Ingersoll, 2003).

Control and Turnover in CMOs
Nationally representative data from the 2008-2009 Teacher Follow Up Survey (TFS) showed that average teacher turnover was about 15% in traditional public schools compared to 24% in charter schools (Keigher, 2010).Available data show similarly high turnover rates in Charter Management Organizations (CMOs) and No Excuses charter schools (Furgeson et al., 2011;Merseth, 2009;Tuttle et al., 2013;Lake et al., 2010).The Knowledge is Power Program (KIPP), which currently has 141 schools in 20 states (KIPP.org,2013) is one of the most prominent CMOs.In the 2010-2011 school year, 27% of teachers in KIPP schools nationwide left their classroom teaching positions (KIPP Foundation, 2012) and 32% left in school year 2011-2012(KIPP Foundation, 2013).High turnover in KIPP schools is primarily attributed to a heavy workload for KIPP teachers and principals (Ash, 2013;Woodworth et al., 2008).Compounding this issue is the fact that 86% of KIPP middle school principals report teacher vacancies that are difficult to fill -citing insufficient qualifications, candidates not being a good fit for their school cultures or goals, and vacancies in high-need areas (Tuttle et al., 2013, p. 30).High turnover and a low supply of ideal teachers cause some to question the long-term sustainability of CMOs like KIPP and their ability to expand (Ash, 2013;Brill, 2011;Torres, 2014;Wohlstetter, Smith, & Farrell, 2013).Although it is likely that teacher autonomy (or the lack thereof) contributes to turnover, we know little about teacher autonomy in CMOs.
Limited faculty input into decision-making at one's school is associated with higher degrees of teacher turnover across a wide variety of the nation's schools and teachers (Ingersoll, 2001;Ingersoll & May, 2011).This is because teachers, especially as they gain teaching experience, often except to exercise their professional autonomy to feel efficacious.In a book about the careers of 1,000 graduates from UCLA's Center X Urban Teacher Preparation program, Karen Quartz and colleagues (2010) argued: "teachers who consider themselves professionals expect corresponding levels of autonomy in their daily work" (p.31).In particular, teachers who are in their "midcareer" (i.e. 4 th -8 th years of teaching) acquire greater mastery in their craft and thus expect a higher degree of professional autonomy to remain "challenged, energized, and useful" (Quartz et al., 2010, p. 33).Teachers who expect professional autonomy -which many consider a mark of competence, status, and respect -and do not receive it are more likely to leave teaching (Quartz et al., 2010;Rosenholtz & Simpson, 1990;Weiss, 1999).At the same time, this generation of teachers are also interested in pursuing roles that allow them to have greater influence on students, colleagues, and schools (Donaldson et al., 2008).

Teacher and CMO Characteristics that Influence Autonomy
Charter schools tend to hire teachers who may prefer or expect professional autonomy.Relative to the labor market in traditional public schools, CMOs tend to prefer and hire high proportions of Teach for America (TFA) corps members and alumni (Chadwick & Kowal, 2011) and teachers with higher than average academic ability who graduate from prestigious institutions (Wilson, 2009).In an analysis of high-performing Massachusetts charter schools, Steven Wilson (2009) wrote about the importance of having teachers implement best practices and highly scripted curriculum with fidelity.He hypothesized: "Even TFA-style teachers, while embracing accountability and uniform standards of excellence, may greet such [schools] not as an asset to their practice but rather as discordant with the individualistic challenges in which they excelled in high school and college" (p. 6).Charter teachers are more likely than TPS teachers to choose their schools because they expect to have classroom autonomy, be involved in decision-making or governance, and be able to influence school policies (Cannata & Penaloza, 2012).Dissatisfaction is a likely result when teachers who expect autonomy do not experience it.
Yet many CMOs prescribe practices to some degree.Using surveys from 43 CMOs and site visits to 10 CMO home offices and 20 CMO schools, Robin Lake and colleagues (2010) found that CMOs are most prescriptive in their decision-making approach around support for struggling students, teacher evaluation and compensation, and education and discipline programs (Lake et al., 2010, p. 30).While schools vary within and across CMOs, 84% of 43 CMOs were categorized as moderately to highly prescriptive in terms of decision-making (Lake et al., 2010).For example, 95% of principals at about 300 CMO schools reported implementing a school-wide behavior strategy (e.g. a merit-demerit system) and 74% required parents to sign "responsibility agreements," or contracts agreeing on explicit expectations for parents, students, and teachers (Furgeson et al., 2011, p. xxiv).
Subsets of these CMOs, particularly the higher-performing CMOs, adhere to a "No Excuses" charter school (NECS2 ) model or philosophy (Lake et al., 2010;Peyser, 2011).NECSs are becoming increasingly popular and some see them as a replicable reform standard (Dobbie & Fryer, 2011;Merseth, 2009;Whitman, 2008) because they are associated with higher than average student achievement for low-income students of color (Angrist et al., 2011;Fryer, 2011;Tuttle et al., 2013).NECS proponents contend that lessons and best practices should be implemented consistently across new schools (Whitman, 2008;Wilson, 2009), which puts these schools in tension with the original charter school principle of teacher autonomy.At the same time, teachers in these schools may be accustomed to success and independence (Carter, Amrein-Beardsley, & Hansen, 2011;Kretchmar, Sondel, & Ferrare, 2014) that puts them at odds with the requirements of educational models that rely on high external control.
In summary, many teachers expect some professional autonomy, especially as they gain expertise.These preferences and expectations may be especially important to charter and CMO teachers, particularly those who come to their schools with the expectation that they will have more autonomy than they would or did in traditional public schools.Little research on charter school teacher autonomy has addressed the proliferation of CMO models in the last decade, how these models influence teachers' perceptions of autonomy and their career decisions, or which aspects of autonomy matter most to teachers' satisfaction and career decisions and why.Therefore, this study uses 20 interviews with CMO and non-CMO teachers to address the following research questions: 1. What areas of school life did these teachers desire autonomy or influence over and why? 2. How did these preferences for autonomy affect teachers' satisfaction and career decisions?

Participants
This exploratory study was conducted with a purposive sample of 20 teachers who taught in and recently left their classroom teaching position in a New York City (NYC) charter school.NYC contains a large number of both CMO and non-CMO standalone charter schools (Scott & DiMartino, 2011, p. 186;Woodworth & Raymond, 2013).New York and NYC also have particularly high performing CMO and non-CMO charter schools compared to CMOs and non-CMO charter schools in 20 other states (Woodworth & Raymond, 2013, p. 10).NYC is home to several prominent, high performing CMO schools affiliated with Uncommon Schools and Success Academies (for evidence of success, see: Lake et al., 2012;Woodworth & Raymond, 2013), including CMOs like KIPP and Democracy Prep that have received significant federal funding for their "proven records of success" (Ed.gov, 2012).
I identified potential participants, any teacher who left or was planning to leave a full time teaching position at a CMO school within a year's time, by recruiting through a Teach for America (TFA) NYC-based newsletter and by emailing educators who might know participants who fit the criteria.I sampled based on some known characteristics of charter school and CMO teachers.For instance, different CMOs hire between 15-50% Teach for America (TFA) teachers (Chadwick & Kowal, 2011), and these teachers tend to be in their 20s (Wilson, 2009).Sixty percent of the teachers in the sample are white, female, and affiliated with TFA.I sought variation in the sample based on teacher and school characteristics to add conceptual richness to categories and themes (Corbin & Strauss, 2007).There is a great degree of variation in the sample in terms of age (between 24-35 years), years of teaching experience, subject and grade taught, and age of the charter school.Finally, I sought to recruit teachers with varying perceived teaching effectiveness.Two teachers were asked to leave but seven in the sample were either offered a salary increase, held a teacher-leader position, or were offered a leadership position to stay (see Table 1, Appendix).
I initially sampled for CMO teachers, and they represent 13 of the 20 participants.I decided to broaden recruitment criteria to include teachers at standalone schools, or charter schools unaffiliated with a management organization.This allowed me to compare how their experiences were similar or different from CMO teachers on the basis of emerging categories and themes (Corbin & Strauss, 2007).This was important since teachers at start-up or standalone charter schools are likely to have differing experiences with autonomy and control because they are not affiliated with a management organization (Bulkley, 2005;Gawlik, 2007).

Data Collection and Analysis
Between December 2010 and June 2011, each teacher participated in an hour-long, semistructured in-person interview.Interviews were audiotaped and transcribed.My interview protocol contained questions about the participant's general career decisions, their initial expectations, their school leaders, and the benefits and challenges of working at the charter school they recently left.During the planning stage, I was aware that recruiting teachers who volunteered to talk about their decisions to leave would result in oversampling teachers who may be especially dissatisfied with their school experience.Therefore, I interviewed until I had a wide variety of experiences represented.For instance, I interviewed two teachers who quit mid-year and were very dissatisfied with their experiences, but also teachers who were much more satisfied with their schools or had very positive things to say about their experience, and still others who were not satisfied at one school but left to teach at a different CMO school that they felt was a better fit.It was important to sample for teachers whose experiences differed in this way because overly disgruntled participants may be more likely to omit details that portray the school favorably or themselves unfavorably.Second, I asked participants to talk about the positive aspects of certain experiences, events, or perspectives, not just challenges they experienced.Third, I probed for highly specific examples and asked participants to describe events in detail when they explained their perspectives.Focusing my analysis on examples and stories of participants' lived experience allowed me to form a potentially less biased interpretation than taking a person's description of their feelings at face value.
In the first cycle of coding, I used an "initial coding" method, developed categories and subcategories from codes and iteratively revised codes/categories, then "pattern coded" in the second cycle to develop initial themes (Saldana, 2009).During and after reading and coding transcripts, I summarized information and analytic thoughts in memos.From this process, three initial themes affecting teachers' career decisions emerged from the analysis (conflict over teachers' professional autonomy, high and implicit expectations for teachers, and the relationship between disciplinary systems / expectations for students and teachers' sense of success).This paper is based on one of those themes and presentation of data is generally limited to experiences of teachers who described significant conflict with their school models or school leaders around issues of professional autonomy.I tested initial themes by going back to my categories, and re-coding the data based on these thematic ideas.Throughout the analysis, I followed an iterative process of writing memos, reviewing the literature, building and revising theories, using new understandings to inform future data collection, and refining my understanding of participants' experiences.I identified autonomy as a central theme by analyzing codes that indicated significant internal and external conflict teachers experienced around how to best educate, discipline, and build relationships with students.Furthermore, the significant absence of these kinds of conflicts for non-CMO teachers primarily, particularly as it related to their challenges and career decisions, helped solidify the theme.The teachers profiled in the findings were chosen to portray conceptual variation (Corbin & Strauss, 2007).For example, teachers with varied levels of teaching experience, expertise, and different kinds of lived experiences helped me to better understand similar and different facets of autonomy and turnover.I focus my findings on teachers' preferences for autonomy, why teachers have these preferences, and how preferences influenced career decisions.

Findings
CMO teachers described having substantial autonomy to write their curricula, but wanted a voice in larger school policies they felt would positively impact students' socialization and success beyond the realm of academics.Their ideas often clashed with their organization's previously established "model."This led to substantial dissatisfaction and (in some cases) decisions to leave, especially in cases when teachers felt students were not being served well.In particular, teachers experienced conflict around disciplinary and academic expectations they felt were inconsistent with the need to prepare and socialize students for college and life.The first section that follows describes "founding teachers'" expectations for involvement in school-wide decisions and the conflict between these expectations and their CMO's model.

"
We were hired to be architects, but we were construction workers." Teachers who were starting a new charter school associated with a CMO came with strong expectations to have a voice in important school-wide decisions.Three teachers -Brenda, Jason, and Natalie -came to their CMO schools with several years of teaching experience and were recruited with the promise of becoming "founding" teachers.Each of these teachers expected to exercise their expertise in their "founding" roles and wanted a say in decisions such as the development of school structures and the fundamental direction of the school.
Jason left his teaching position at a private school to work specifically with urban schools as an administrator.After finishing an MA in Educational Leadership, Jason first decided to look for teaching positions in an urban school.While many of his colleagues in the MA program got leadership positions after finishing graduate school, Jason explained: "I didn't feel like I should get an urban leadership position without teaching in an urban area."He added: "[I thought] it would be fun to live in NY," and decided to work at a charter school because he "wasn't certified" and knew "[charters] were a little more lenient on certification." Like several other teachers, Jason was given support and time to collaborate with colleagues to shape the curriculum during two to three weeks of planning time before the school year started.Unlike other teachers, Jason "couldn't stand it."He stated: "Sitting down and writing out a [middle school] reading and writing curriculum is terrible, in my mind."Rather, he preferred to have a say in larger school policies outside of his classroom.He explained: "I guess the word "founding" felt like founding in the sense that I've got to spend three weeks of my life writing a curriculum, rather than, I'm not really founding the discipline measures or the structures."He expected the term founding to mean being "privy to and contributing to those fundamental decisions that are being made."Jason voiced concerns to his principal and was told: "This is how we do things.That's just [the organization's] way."As he put it, "there wasn't a whole lot of transparency between how decisions were being made on a school-wide level, particularly on a disciplinary level, and teachers.So you've got people who are all founding teachers of a school, and yet you've got two people who are like, they're higher-ups making the decisions and handing them down."He expected to have a voice in decisions on a "school-wide level" and felt frustrated at only having the classroom autonomy to write curriculum.
After teaching for six years at a CMO charter school in a large Midwestern city, Natalie moved to NYC for personal reasons and chose to teach at a CMO school for the opportunity to help build a new school.Natalie did not express dissatisfaction with writing curriculum as Jason did, but further clarified why she felt writing curriculum was not meaningful to her.Like Jason, she was frustrated that "founding" did not mean more than creating curriculum.She explained: I signed on thinking I was gonna be a founding teacher, so that meant for me not just creating curriculum, because I've been creating curriculum since I started teaching, but building the school.What are the cultural pieces of the school?What traditions do we start creating for these kids?What are the discipline systems?How do we grow when we have 200, 300 kids?What are our growth trajectories?And it became very quickly clear that upper management and the principal had decided on what the school was gonna be in its trajectory and were not receptive to hearing change in a year, even if it could have been beneficial for the students.Similar to Jason, Natalie felt "upper management and the principal," who were not interested in hearing new ideas, had already decided the trajectory of the new school.Both Natalie and Jason had a form of classroom autonomy in decisions made about curricula but expected faculty to have influence over broader aspects of school culture and matters related to students' socialization or success because they were hired as "founding teachers."While they perceived writing curriculum as an important aspect of their work, they wanted a voice in determining school traditions, deciding how students would be disciplined, and school growth.
Brenda had a similar experience but was given an informative explanation for why her ideas could not be acted upon.A Teach for America (TFA) teacher who came with four years of prior teaching experience in a large NYC traditional public school, Brenda looked at various positions outside the classroom but wanted to continue teaching.Specifically, she wanted a place that would support her in her efforts to become a better teacher and she "intentionally chose to work [at a new CMO high school] because of the opportunity to build."Like Natalie and Jason, she expected to have a voice in decision-making about school policies and structures.A self-proclaimed teacher of color and proponent of "culturally responsive" pedagogy, she wanted to structure the school's student advisory periods to have "really frank conversations with students around issues that ninth graders were facing."According to Brenda, however, her "manager" responded to her ideas by saying "academics is all we have time for."She recalled saying to her manager: "You know, I think the only way that I can talk about this is with an analogy.I think we all as a team thought we were hired to be architects, but in actuality we're just construction workers, and that's what's hard."She recalled her manager's response as follows: In all honesty, we developed a model for what the 9th grade should look like before you got here, and I was instructed as your manager to ensure that there would be as little deviation from that particular model as possible.What we had seen in 6th through 8th grade worked, and we wanted to deviate as little as possible from that in this launching of the 9 th grade.Brenda's experience illustrates one reason why CMOs and CMO leaders are reluctant to relinquish control to teachers: the desire to implement a previously established model of best practices.This is consistent with broader policy rhetoric about the ability of CMOs to scale up their efforts while sustaining academic success across schools (Farrell, Wohlsetter, & Smith, 2012).For example, in discussing the success of the highest performing charter school models, Steven Wilson (2009) wrote: "we know that precise adherence to a coherent school design is essential to achieving consistently strong results" (p. 6).Brenda's proposal for advisory periods was at odds with this particular model that "worked."This rationalization became a form of control that she spurned.She explained: "This conversation, I believe, was towards the spring, February or March.That would have been really helpful to know, from the beginning."In the end, limited autonomy combined with increasing conflict with her manager (e.g.two-hour meetings over the accuracy of instructional feedback and being told that she was not yet adjusted to a charter school environment) led to "mounting frustration" for Brenda.As a result, she began looking for jobs by "October-November" and left after one year in spite of being offered a leadership position and salary increase to stay and feeling extremely guilty about leaving, as she passionately stated: "I felt like I didn't want to leave those kids [in tears]." Natalie, Jason, and Brenda wanted to have a say in creating structures they felt could be beneficial to students.These structures were related to how they believed students should be socialized, a point I explore in greater depth in a later section.Teachers also desired autonomy to make their own decisions about what was best for students when they disagreed with some of the CMO's existing policies and expectations for students.

"This is How We Do Things:" Conflict Over Academic & Disciplinary Aspects of The Model
Teachers' ideas about what was best for students sometimes clashed with their organization's model, leading to internal or external conflict that influenced their decisions to leave.Jason actively voiced his concerns to his principal and felt strongly about what he called the "disciplinary piece" of charter schools.For example, he expressed initial reservations about charter school discipline during his job search after seeing, in one school, "a student standing with his nose against the wall, half an inch against the wall, as punishment in the hallway."Of the incident he noted: "I don't know what he had done.That really got to me."While he felt the school he chose to work in had a positive climate for students and he had a "great relationship" with his school leader in spite of occasional disagreements, he continued to have issues with school-wide disciplinary expectations, especially what he felt was an absence of dialogue around them.Jason gave a specific example that highlighted his frustration with charter school "structures."He directly connected this incident to his decision to leave, saying, "I knew for a fact, with the green ink incident, I was like, 'I've got to get the hell out of here.'That was probably a year and a half in."He explained: There was one [example] I remember vividly, there was this awesome kid, he was one of these ADHD kids who had a ton of energy, never mal-intentioned, but he would have detention every day because he just couldn't sit still in class, very much like I was as a basket-case [middle-schooler].He had a green pen one day, and [the school] supplies [the pens]."Here are your pens, here's your everything."So everybody's prepared.You have no excuse not to be prepared.
[He] had his green pen and he was walkin' up to lunch from tutoring and I don't know if he threw the pen or if it fell out of his hand, or he was probably throwing it up and trying toanyway, green ink got on the wall of the school.Release the hounds.So people are outraged.There's, like, a school-wide-and I'm like, What is goin' on?Where are our priorities here?You can't even say, "OK, there's a little bit of ink on the wall?"Let's hold it together."Sweating the small stuff," is a common tenet of "No Excuses" charter schools (Whitman, 2008).This entails having clear expectations for students (i.e.prescribing exactly what a student is expected to do and holding them accountable to it with consequences and rewards), and spending as much time as possible on instruction.In this case, students are provided with pens so they have, as he put it, "no excuse not to be prepared."It is a potentially time saving strategy --if the school provides the pens, teachers would not need to waste their time arguing with the student over the matter, or finding him a pen to use.In this case, the student had a green pen instead of the required black or blue pen.When green ink got on the wall, Jason felt that staff reactions were excessive.For example, according to Jason, the dean of discipline explained in a meeting: "We're strictly a black and blue school.There's going to be no more green ink [in our school].They're gonna get detention."After the meeting, teachers needed to have "students pull out their pencil boxes and [writing utensils] because if they had [a different color] it was gonna be an automatic detention."Jason complied for the sake of consistency with the school's policies, but felt unable to voice his concerns or change disciplinary policies he resisted.He concluded: "That's when I knew it was time to go." Cara was a TFA teacher fresh out of college who knew nothing about charter schools and was placed in a CMO elementary school.She left the charter school to teach at a TPS after her third year.After her first year teaching, she grew increasingly troubled about particular values and practices.She perceived her principal's decision making as constrained by her CMOs "model," as she explained: "I really can't say, 'She did this well and she didn't do this well,' because that's the way it was… she wasn't choosing this for us, it was already decided… the model is the model." Cara came to disagree with several academic and disciplinary expectations of what she perceived to be her school's non-negotiable model.She acknowledged that frequent assessments were important to helping her determine instructional areas to focus on but grew to hate the "constant testing," and the disciplinary practices and structures of the school model, saying, "[this] is so far from anything I would ever want for my child."She explained: There were times [when we gave assessments], it might be three weeks straight at a time, which was very overwhelming, I think for small children.I actually had a child tell me he wanted to kill himself because he just couldn't take it anymore, which I thought, at seven years old?That should not be your educational experience, being tested until you can't take it anymore.If I had a child, I would never want them to be in such a strict, structured environment from 7:15 in the morning until 4:15 in the evening.It was physically and emotionally draining, not only for the teachers, but for the children.There wasn't enough time to just be a child.We weren't educating the whole child.We weren't thinking about their development and what was ageappropriate.We were doing so much test prep that I felt like their childhood was being lost.Cara came to believe that certain core elements of a "No Excuses" charter school model such as frequent interim assessments and a longer school day (Whitman, 2008) were inappropriate for young children.She cried as she described "silent lunches" for "five year olds" and an example of an older child being placed in her classroom as punishment even though the child's mother had just passed away, calling their treatment "inhumane" and "unfair."She also equated what she called a strict, structured, "No Excuses kind of mantra" with a "military environment" that, to her, seemed diametrically opposed to the college culture the school espoused: It was almost to me like a military environment, walking with your hands by your sides, eyes forward, right behind the person in front of you.When I clap, you clap!You look, I look!We turn this way!It was ridiculous to me.And it was very much a school where they were promoting, "We want to send your kids to the best colleges in the U.S." But I had never seen anyone on the Yale campus or the University of Michigan campus walk in a straight line with their eyes straight, head forward, hands by their sides.Cara's concerns were rooted in how students were being socialized and treated in a "No Excuses" environment.When asked whether she shared her concerns with anyone, Cara explained: "There were a lot of conversations teacher to teacher about this kind of thing but I would say overall, it was pretty much, if you know that this is not the way you want the management to be, that wasn't the place for you."Feeling unable or unwilling to directly address her deep concerns around the school model, she looked exclusively for jobs within the traditional public school system in spite of district hiring freezes and added: "I wasn't interested in any charter school in any part of the world."Interestingly, she saw her choice to work in a traditional public school instead of a charter school as a way to increase her autonomy.She explained: "[I am] actually given the freedom to try what I think works best in my classroom." Jessica, a kindergarten teacher who stayed at her CMO charter school for three years before leaving, was also frustrated with what she perceived as a lack of control to structure her classroom the way she wanted and inability to have a voice in certain school policies.Unlike Cara she expressed a positive view of her school and said, "I think it is an awesome school."She called her former colleagues "incredibly humane and talented."Yet Jessica still felt that "a lot of things were top-down."She gave an example of an "all hands on deck" policy for teachers to monitor student breakfast then walk up at 7:50 with their classes to begin teaching.She described how teachers would try to offer suggestions to change this policy: "[We said], couldn't the system be such that we can switch off when it comes to breakfast?And we would suggest it, teachers would say, 'That charter school down the street's doing that.They have parent volunteers.They switch on and off.One co-teachers on one week.'And [school leadership] wouldn't give.It was frustrating."Jessica noted that "big curricular and structural decisions" like this were not open for discussion or teacher input.She added: "It was kind of hard to have a voice.It didn't really feel like a democracy." Her school's model also restricted her sense of autonomy over her classroom, which felt "demeaning" to Jessica.She discussed a "special consensus" of K-2 teachers who believed that "there wasn't enough time for play, for center-based activities, I guess science and social studies were supposed to be a hands-on part of the day, and math, but it didn't have a developmentally appropriate feel to it, the behavioral systems and things like that."According to Jessica, these practices were a result of a book she described as the school's "bible" in which "they talk about having desks in rows and how that's important because all eyes are on the teacher… and that was the standard.[So they told us], 'we are having desks in rows.'"She added: But when you're talking about kindergarten, first and second grade?Arguably other grades as well, but especially those early grades, the social curriculum is as important as the academic curriculum, so it's so important to have the kid have the opportunity to be at a table or desks configured as a table so they can talk together.While administrators seemed to prefer order and consistency across classrooms, Jessica wanted to arrange her kindergarten classroom in a developmentally appropriate manner that would aid students' ability to grow socially.Ultimately, Jessica and her co-teacher arranged the desks the way they wanted.She explained, "and they let it slide, but it wasn't without a fight, and it was only in kindergarten.To not be able to arrange the desks the way you wanted felt so demeaning… like my knowledge of teaching and the way I want to organize my classroom and conduct my classroom doesn't really matter."Similar to others, Jessica felt that the limits placed on teachers' school and classroom autonomy were demeaning.While she expressed dissatisfaction about this situation it seemed to play less of a role in her decision to leave.In contrast with Cara, she believed that her school was a positive place for students.Instead, Jessica talked about the early, long hours and the intensity required to do her job well, and concluded: "Definitely I started feeling burned out.It was time to go."

Student Socialization and The Limits of Classroom and Curricular Autonomy
Teachers' desires to have a voice in issues surrounding students' socialization ultimately hinged upon their skepticism that their schools' current practices were adequately preparing students for college and life.Natalie explained: "Schools have to be more than good classrooms… I think going to college, students seem to need social capital as much as they need academic capital.We need to provide that for them."Similar to Cara, Jason expressed skepticism about preparing students for "college readiness" in light of his school's disciplinary system, saying, "We created these little robots who do things or don't do things explicitly so that they either receive a merit or don't receive a demerit.That's not how life works."Natalie further explained how her limited influence affected her decision to leave: There were other voices that were more important than ours, I think.And when you're trying to create something, you have to feel like your voice is important.And if other people's voices are really important, then they should be part of the discussion.It felt like we were compartmentalized, we weren't allowed to be a part of the bigger picture.I felt like as a founding teacher, it could only get worse for the next set of teachers coming in.And so I think that's when I stopped believing, when I realized that even if I say something, there's somebody else whose opinion will always supersede mine, without a discussion with me about it.Teachers felt their voices were primarily limited to the realm of academic and curricular decisions but they wanted to have larger conversations about building "social capital" and more effective disciplinary practices.Feeling "compartmentalized" to academics in terms of autonomy and time also affected teachers' perceptions of their relationships with students.When asked about his decision to leave, Jason explained how the school's academic emphasis impacted him: That's what I really love, kids.Give me kids, man.I love it.That lack of relationship building… it left me feelin' a little vacuous.I was spendin' so much time writing curriculum, creating workshops, filling out detention slips, I had so little time to sit and get to know kids and what was happening with them.So I think I felt empty in that sense.Jason realized that "relationship building" is what he loved.He felt unable to build the relationships he wanted with his students because of what he perceived as an overwhelming focus on academics.

Left Alone But In Need of Support
The CMO teachers profiled described experiencing conflict around top-down decisionmaking processes they attributed to inflexible models.In comparison, newer teachers in several standalone charter schools were often left alone but wanted more direction and support.Consistent with the theory that expectations for professional autonomy increase as teachers gain expertise (Quartz et al., 2010), those teachers with more expertise had more strongly formed opinions about what would be best for students and those with less expertise wanted more guidance around what to do and how to do it.For example, CMO teachers often preferred the structure of a school model in their first year, as Jessica explained: "[As a new teacher I felt like] 'listen, you tell me what to do.I have no idea what I'm doing.But the second and third year, I was like, I kind of wish that what I said made a difference.'"Non-CMO teachers were more likely to discuss challenges related to the absence of a strong school model guiding discipline and instruction.Their experiences emphasize the tension between teacher autonomy, expertise, and school effectiveness.
Brand new teachers like Jasmine, a TFA corps member who left teaching after her two-year placement, described her standalone charter school as "chaotic" and deliberately contrasted it with what she knew about other CMOs: [The school] was very chaotic.It was very clear that this wasn't-this was a standalone charter school.It's not an Achievement First network.It wasn't a KIPP network.It was a standalone charter school.It seemed like there was no curriculum in place.It seemed like there was no-they had expectations in place for the children to learn, but there was no guideline as to, "What's your long-term plan?What's your short-term plan?What's your daily plan?"They left that pretty much up to me.I felt very well supported by Teach for America, so I leaned heavily on them for support in the first year, because I would say that the priority during my first year was making sure that the classes were peaceful and working, [considering the] discipline problems.Jasmine said that most decisions were left to her, but this obviously resulted in chaos.New teachers need substantial support and mentoring as a necessary condition for feeling successful and wanting to stay in teaching (Johnson et al., 2004).She felt the biggest challenge the school faced was student behavior, and wanted a "school-wide system of discipline."One administrator attempted to establish this type of system.But according to Jasmine, the principal was pushed out of the position because of conflict with the founder who felt the system philosophically contradicted the mission of the school.
Allan left a standalone charter school after his first two years for a teaching position in a CMO school.Like several others, he left looking for a school that could support him to become a better teacher.He described his biggest challenge as "not getting feedback.
[laughs] It felt very isolating, which teaching naturally is in some ways, but it doesn't have to be as isolated as it often becomes.One of the reasons I ended up leaving [the school] was, I [was struggling with student behavior and instruction and] wasn't getting enough feedback."In contrast with other schools where teachers described being observed and given feedback weekly, Allan was almost never observed.He trusted his principal because he felt respected and listened to but also felt that "he should have been more active in the classroom, having a say, trying to improve instruction." These teachers' experiences lend deeper insight into the advantages and also the limitations of school models or practices that can help improve teachers' craft, particularly newer teachers in need of guidance, while at the same time potentially competing with more experienced teachers' ideas about what is best for students.Clear systems, procedures, and expectations around instruction and discipline are necessary to the support and success of teachers.But in order to increase the satisfaction and commitment of educators who expect professional autonomy, organizations have to be flexible and responsive to the competing needs and ideas of teachers and students.

Discussion
This study reinforces, extends, and challenges existing research on charter school teachers' perceptions of autonomy.Previous research shows that autonomy is more likely to matter to charter school teachers in their job search, and that they perceive a greater degree of school autonomy compared to TPS teachers.Autonomy was indeed important to charter school teachers in their job search --those who joined new CMO schools with the explicit expectation to help found a school cited this opportunity as a key reason for choosing to teach there.While earlier case studies documented the ways in which charter school teachers were both energized and challenged in their efforts to create standalone charter schools (Vasudeva & Gratzik, 2002), the CMO teachers in this study found it difficult to have a voice in important school decisions.Compared to new standalone schools with unique founders and missions, many CMOs are at a stage in their organizational trajectories where they can draw upon the best practices that previous schools in their network have already found to be successful.Based on these teachers' experiences, the drive for organizational consistency and implementing what is viewed by leaders within the network as working can limit the decision-making power of teachers and principals at individual schools.In turn, teachers who wish to exercise their expertise for the benefit of the school or who expect to be granted the professional autonomy to decide what works best for students may experience conflict when these ideas are inconsistent with a previously established organizational model.
Teachers' expectations and satisfaction are also mediated by teachers' buy-in or fit with their organization's model or practices.Jason was uncomfortable with disciplinary practices even as he looked for charter school teaching positions, and he continued to disagree with certain practices at his school.Brenda, Cara, Natalie, and Jessica each started at their schools without strong ideas of what it meant to teach in their CMOs.They either came to disagree with school practices or wanted to create their own school initiatives without realizing how their proposals would conflict with previously established practices.Teachers who come with clearer ideas about what to expect either through their own initiative or through better communication by the school leader at the interview stage may experience less conflict once hired.
Previous research has called for a more fine-grained understanding of variations within the charter sector (Cannata & Pennazola, 2012), an important point to note considering the growth of CMO models in recent years.While much of the literature on charter school teacher autonomy examines stand-alone charter schools, this study demonstrates the possible tensions and conflicts between teacher autonomy and CMO control, and grounds this tension in the experiences and career decisions of teachers.Recent studies examined charter school teachers' perceptions of control over the educational programs at their school while largely ignoring control over decisions related to the socialization of students, an aspect of teacher autonomy known to be a powerful predictor of conflict and turnover (Ingersoll, 2003).Indeed, teachers' struggles for voice and control mainly revolved around beliefs about how students should be socialized.Although curricular freedom and autonomy to control decisions over how to teach academic content are clearly important to charter school teachers (Gawlik, 2007;Malloy & Wohlstetter, 2003), conflict over student socialization was the most challenging and relevant aspect of autonomy for the CMO teachers in this study, especially as it related to their decisions to leave.This is important to acknowledge given the disciplinary structures and expectations of many CMOs, particularly those that have a "No Excuses" culture.
The findings in this study raise important considerations for autonomy and turnover in the CMO sector.Since expectations for student behavior are one of the most prominent and highly controlled features of "no excuses" charter schools and CMOs (Lake et al., 2010;Whitman, 2008), finding ways to listen to or incorporate teachers' voices into decision-making processes in ways that feel meaningful to them may be especially important to increasing teacher satisfaction, commitment, and retention.As teachers gain more experience and become increasingly opinionated and competent in their practice (Rosenholtz & Simpson, 1990), finding ways to extend teachers' professional autonomy may be an important strategy for retaining experienced teachers.This is a critical consideration for CMOs in light of the tremendous resources devoted to developing new teachers (Rich, 2013;Lake et al., 2012), the limited supply of teachers who are a good fit with their organizations (Tuttle et al., 2013), and increases in teacher effectiveness over time (Harris & Sass, 2011;Rivkin et al., 2005;Rockoff, 2004).
In schools where teachers are unsure about or disagree with their organization's model, teacher turnover due to limited teacher autonomy may be especially high.Understanding of this problem is limited, but there is some evidence that teacher buy-in was a problem in San Francisco KIPP schools, where 34% of surveyed staff reported they were not comfortable with KIPP disciplinary policies (Woodworth et al., 2008, p. 53).One response might be to say that departing teachers are clearly not a good fit for the school, and would undermine the school's mission if they stayed.While this may be the case in some schools or with certain teachers, this response is problematic because a narrow definition of an organizationally ideal teacher exacerbates a situation where teacher supply is low and demand is high (Tuttle et al., 2013).Additionally, this problem may be in part attributable to leadership (e.g. a leader who is not competent at communicating and developing trust with teachers) rather than an individual teacher not being a fit for the school.Either way, the issue of "fit" combined with the problem of high annual turnover in these settings makes it much more difficult to envision the widespread replication of CMO and No Excuses schools.Stated differently, if there are too few teachers who are a fit, and the definition of fit is the cause of recruitment and turnover problems, then can we reasonably expect to expand these schools and their practices into the wider system?Finally, autonomy in the traditional sense of being left alone (Little, 1990) is obviously not desirable and can quickly lead to turnover, as the non-CMO teachers who needed support illustrated.Strong instructional and disciplinary support systems to promote consistency and help struggling teachers develop are a key feature of academically successful CMOs (Lake et al., 2012;Rich, 2013;Tuttle et al., 2013).These systems may not be as prevalent in standalone charter schools whose founders have a unique vision or that rely on teachers and leaders who may not have the necessary time or expertise to develop effective support systems.While outside the scope of the study, future research should examine differences in autonomy between standalone and CMO teachers and how these differences may be associated with school effectiveness.

Implications and Limitations
Although this is a small study and appropriate caution is advised in generalizing these results, findings are consistent with theory and other studies on CMOs and teacher autonomy.Current evidence suggests that CMOs have a moderate or high degree of control over local practices (i.e., student discipline), which has clear implications for teacher autonomy and turnover.While this study informs the knowledge base on how and why teacher autonomy is related to turnover, it does not address to what extent these kinds of issues are experienced in the rapidly shifting charter school and CMO landscape.More research is needed on the practices, strengths, and challenges of CMOs and "No Excuses" charter schools as well as how these compare with standalone or conversion charters.In particular, large-scale quantitative studies should differentiate between types of charter schools and examine differences within and across them since there is a great deal of variability in CMOs and the charter school sector.This could help us better understand possible differences in charter school autonomy within and across these growing school types and to what degree this is associated with teacher satisfaction and turnover.
This study is based on a small, purposive sample that does not include a comparison group (e.g., teachers who choose to stay that are also in the same school).Talking to "stayers" within the same school could be particularly important to understanding whether a teacher's individual views are shared by other teachers and would allow for a more nuanced and richer understanding of a school's setting and working conditions, which was outside the scope of this study.My intent was to understand individual teachers' career decisions at a deep level, and how they perceive their own experiences is what matters to teachers' individual career decisions.While it cannot be assumed that self-reports represent how events objectively played out in practice, the teachers' perceptions of these events influenced their satisfaction and career decisions.Therefore, self-reports are vital forms of data in terms of understanding teachers' career decisions (Ingersoll, 2001).Additionally, the interview data focused on the general experiences and career decisions of charter school teachers and did not have questions related to specific aspects of autonomy that might be important to teachers.Future studies should investigate how charter school teachers experience various aspects of school and classroom autonomy, which is important because of how rapidly the charter school landscape is evolving.
This study raises several implications for practitioners.Improving a principal's ability to understand and influence staff perceptions through effective communication is one solution to managing the conflict discussed here.Principals have the power to mediate teachers' expectations and build trust and commitment regardless of the organizational context (Bryk et al., 2010).Future studies should investigate whether and how CMO teachers experience greater autonomy as well as how school leaders manage conflict and disagreement within the context of the their organizational models.Clearly, not all CMO teachers will disagree with the values and practices of their school's "model" -the teachers in this study simply give us insight into how and why disagreement may occur in ways that influence teacher satisfaction and turnover.
Finally, it is important to be aware of broader implications for the teaching profession as reformers promote the expansion of these school models and practices.On one hand, certain charter networks can be adept at acculturating novices to quickly become more effective (Rich, 2013).In this respect, rigidly defined systems, practices, or policies can serve to reduce ambiguity and uncertainty by showing young, novice teachers what to do and how to do it.As illustrated in this study, these systems may become polarizing for certain teachers as they gain expertise and desire greater autonomy or a voice in decision-making.If limited teacher input contributes to the problem of high turnover, this may increase the likelihood that relative novices, not experienced teachers, are the norm for staffing these schools.In turn, policymakers might respond by expanding alternative certification models such as Teach for America to increase the supply of new, enthusiastic young teachers instead of considering policy levers to concurrently increase teacher retention and the longevity of the teaching profession while simultaneously focusing on increasing teacher effectiveness.Autonomy has long been an indicator of professional work, status, and respect in a profession (Gorman & Sandefur, 2011).Promoting policies and practices that may limit teacher autonomy reinforces the view that teaching is a "semi-profession" with limited status, a shallow knowledge base (Etzioni, 1969), and not a viable long-term career option for young professionals with a variety of job options will consider doing for more than two to five years (Rich, 2013).At a minimum, the experiences the teachers in this study highlight the importance of considering whether and why teaching in some charter schools is more like being a construction worker than an architect.