Capacity Building and Districts’ Decision to Implement Coaching Initiatives

The United States has experienced tremendous growth in the development of coaching initiatives including professional training programs, state endorsements and resources for coaches. These developments bring attention to the potential for coaching to improve education. They also raise the question of how best to facilitate implementation in local districts. One approach is to build capacity for coaching in schools and districts. Capacity building, as a policy tool, can facilitate the skills and knowledge necessary to engage in new practices. To understand the influence of capacity building efforts on districts’ decision to implement coach initiatives, this study examines how one Regional School District (RSD) worked to build capacity for literacy coaching among its 20 constituent districts over three years. To analyze the data, I employed Weiss’ (1983) decision-making framework, which examines the influence of ideology, interests and information on policy positions and decision outcomes. Findings suggest that the RSD’s ongoing capacity building efforts influenced districts’ decision to implement coach roles, even in districts that initially expressed no interest in coaching. In addition, formal district leaders played an important role in shaping the interplay epaa aape Education Policy Analysis Archives Vol. 22 No. 56 2 between information, ideology, and interests. This study provides insights into the utility of capacity building for implementing coach roles and into the “politics” behind coaching initiatives.


Introduction
Coach roles are increasingly touted as an effective means for improving education. Research suggests that the implementation of coach initiatives is correlated with improvements in teaching (Cantrell & Hughes, 2008;Neuman & Cunningham, 2009) and increased student achievement outcomes as measured by standardized tests (Biancarosa, Bryk, & Dexter, 2010;Campbell & Malkus, 2011;Elish-Piper, 2011;Marsh et. al. 2008). The premise is that on-site coaches increase teachers' access to high quality professional development. As such, coaches are understood as working to facilitate teacher learning (Neufeld & Roper, 2003;Poglinco & Bach, 2004;Vanderburg & Stephens, 2010). Typical coaching activities might include modeling instructional strategies, observing and providing feedback, facilitating conversations about student work, or leading teachers in data analysis and problem solving (Killion & Harrison, 2006;Walpole & McKenna, 2004).
The potential benefits of coaching have led to a flurry of activities aimed at facilitating the implementation of coach roles, often under the banner of increasing teachers' opportunity to assume leadership roles. States have created endorsements and certificates that teachers can add to their teaching license (Hohenbrink, Stauffer, Zigler, & Uhlenhake, 2011;Shelton, 2011). Universities, professional associations and school partners have invested in the development of professional training programs (Mangin & Dunsmore, 2014;Berg, Bosch, Lessin-Joseph, & Souvanna, 2013;Bradley-Levine, 2011Bukowiecki, 2012;Paxton, Slattery, Baynum, 2012;Leonard, Petta, & Porter, 2012;Ross et al., 2011;Valli, van Zee, & Rennert-Ariev, 2006). Coaches can attend conferences and join online forums designed to guide and support their practice (e.g. Learning Forward Summer Conference on Teacher Leadership http://www.learningforward.org/ and The Center for Teaching Quality "Collaboratory" http://www.teachingquality.org/). In addition, professional standards such as the Teacher Leader Model Standards (Teacher Leadership Exploratory Consortium, 2011) and the Standards for Middle and High School Literacy Coaches (International Reading Association, 2006) offer guidance on how to design and enact these roles.
These developments bring attention to the potential for coaching to improve education. They also raise the question of how best to facilitate the implementation of coach roles in local districts. One approach is to build capacity for coaching in schools and districts. Capacity building, as a policy tool, can facilitate the skills and knowledge necessary to engage new practices (Schneider & Ingram, 1990;Weatherly and Lipsky, 1977). While capacity building may require significant financial investment up front, it is expected to yield long-term returns (McDonnell and Elmore, 1987). However, these returns are uncertain because capacity building alone is not necessarily predictive of the extent to which policy actors embody the will to reform (Firestone, 1989). Rather, policy actors' propensity to implement new policies is dependent, in part, on their ideology, interests, and access to information (Weiss, 1983). Thus, capacity building may be both necessary and insufficient as a mechanism for facilitating coaching.
To understand the influence of capacity building efforts on districts' decision to implement coach initiatives, this study examines how one Regional School District (RSD) worked to build capacity for literacy coaching among its 20 constituent districts. Absent the authority to mandate coach initiatives in the independent districts, the study asks: How did the regional district's capacity building efforts influence the 20 constituent districts' decision to implement coach initiatives? What influence did ideology, interests, and information have on the districts' decision outcomes? To investigate these questions, I examined the decision outcomes in all 20 districts over the course of three years. During this time the RSD provided districts with literacy coach training that was intended to build capacity for the implementation of coaching initiatives. To analyze the data, I employed Weiss' (1983) decision-making framework, which examines the influence of ideology, interests and information on policy positions and decision outcomes. Findings suggest that the RSD's ongoing capacity building efforts influenced districts' decision to implement coach roles, even in districts that initially expressed no interest in coaching as a school improvement strategy. In addition, formal district leaders played an important role in shaping the interplay between information, ideology, and interests. This study provides insights into the utility of capacity building as a way to influence districts' decision whether to implement coach roles. This study also provides insights into the "politics" of coaching initiatives.

Capacity Building and Decision Outcomes
In the sections that follow I provide a conceptual framework for understanding and analyzing district efforts to implement coach initiatives. First, I describe the potential benefits of capacity building as a policy implementation tool. Next, I present Weiss's decision-making framework as a lens for interpreting districts' responses to regional capacity building efforts. This framework provides an analytic lens for examining coach role implementation.

Capacity Building as an Approach to Policy Implementation
A common theme in the field of policymaking is the tension between autonomy and control, which is evident in ongoing shifts from centralized decision making to more localized discretion. Policies that emphasize the legitimacy of centralized control typically employ authority tools such as regulations, mandates, monitoring and accountability measures as a means to facilitate policy implementation (Elmore, 1983). "These tools assume agents and targets are responsive to the organizational structure of leader-follower relationships and that lower level agents usually will do as they are told" (Schneider and Ingram, 1990, p. 514). However, policy actors do not always perform as instructed, sometimes demonstrating a lack of will or interest (Firestone, 1989). In some cases, local contexts present unique conditions that lead to policy adaptations at the local level (McLaughlin, 1987(McLaughlin, , 1990. Alternately, lack of compliance may result from limited know-how or resources, suggesting a need for capacity building efforts (McDonnell & Elmore, 1987). As such, policies that rely primarily on mandates and inducements have found limited success (Liu, 2004, Strunk & McEachin, 2013).
An alternate approach to policy implementation involves greater autonomy for localities and increased discretion in decision-making (Elmore, 1983). In the context of local autonomy, capacity building is thought to increase adherence to policy goals. Capacity building can include a range of components including technical assistance, material resources, information distribution and increased opportunities to learn (McDonnell & Elmore, 1987;McDonnell, 1994).
Capacity programs assume that the target groups will have sufficient incentive or motivation to participate in the activity, or change their behavior, if they are properly informed and have the necessary resources. Thus, these programs assume individuals are free agents, able to make their own decisions, and do not need to be coerced through the passage of formal laws mandating certain behavior (Schneider and Ingram, 1990, p. 518).
Short-term investment in information, resources and assistance can produce actions at the local level that align with policy goals (McDonnell & Elmore, 1987). Capacity building may be especially important in the context of education, where changes to instruction require high levels of skill (Shulman, 1986;Spillane & Thompson, 1997). Research on capacity building approaches to policy implementation suggests that building skills and knowledge can facilitate reform (Cohen & Hill, 2001;Strunk & McEachin, 2013). At the same time, the long-term benefits of capacity building are uncertain because policy implementation also depends on the target population's will to reform (Firestone, 1989).

Decision Outcomes
Understanding the decision outcomes of policy actors requires an examination of the factors that influence decision-making. Here I draw on Weiss's decision-making framework (1983), which serves as a tool for examining the decisions made by policy actors (see figure 1). According to Weiss, three forces influence decision-making: ideology, interest, and information 1 . This confluence of forces affects the local level policy actor's policy position and decision outcome. Drawing directly from Weiss (1983), I offer a brief description of each of these three components.

Information Information
Interests Interests Ideology Ideology Following Weiss' definition, ideology encompasses "philosophy, principles, values, and political orientation" (p. 224). At the heart of ideology are personal beliefs that are moral and ethical in nature as well as political orientations that predispose policy actors to particular programs of action. Personal ideologies may be "well constructed and coherent" or "haphazard and makeshift" (p. 224). As such, ideology constitutes an "emotionally charged" stance from which policy actors take a political position. Interests refer primarily to self-interest, including but not limited to, one's desire to maintain or increase status, legitimacy, position, influence or material gain. The interplay of interests may occur publicly, in view of other actors and policymakers, or interests may be negotiated out of sight, lending to the notion of "playing politics." In some cases self-interest may conflict with organizational interest. Information refers to the knowledge that people bring to bear in their decision-making. Weiss' uses the term "information" to convey "the partial, biased, or invalid understandings" that we bring to bear on decision-making (p. 225). Information can include "scientific research, experienced judgment, folk wisdom, or gut feeling" (p. 226). Regardless of its quality, information contributes to the explanatory framework used for interpreting policy choices and actions.
Weiss explains that the policy decision-making process should be understood as a continuous interplay between these three forces. Moreover, this process is inherently influenced by policy actors' ability to leverage power. "The distribution of power determines WHOSE ideology, interests, and information will be dominant" (p. 239). Shifts in power can result in different combinations of ideology-interests-information receiving more or less attention in the decision-making process.
I present Weiss's decision-making framework as a lens for interpreting the decision outcomes of 20 districts exposed to a regional capacity building program aimed at increasing the prevalence of literacy coach initiatives. While capacity building can take numerous forms, this study examines the regional district's effort to build capacity by creating a Literacy Coaches Network (LCN): an extensive program of professional development intended to increase opportunities to learn about coaching. Creating opportunities to learn can facilitate policy actions at the local level that align with larger policy goals (McDonnell, 1994;Weatherly & Lipsky, 1977). As such, I treat the LCN as a source of information that intersects with other sources of information, as well as the ideologies and interests of key policy actors in each of the 20 districts. Using Weiss' I-I-I decisionmaking framework provides insights into the utility of capacity building as a means to influence coaching policies and practices.

Study Design
This qualitative study presents findings from a three-year investigation into one Regional School District's (RSD) efforts to build capacity for literacy coaching in their 20 constituent districts. Here I describe the study context; sample selection; data sources and collection procedures; data organization and analysis; limitations; and the terms used in this paper.

Study Context
The research for this qualitative study took place in one Regional School District (RSD) in one Midwestern state. RSDs are intermediary districts that may operate in a way similar to countylevel education offices in other US states. They are a structural mechanism for creating economies of scale and providing services such as special education, data processing, and professional development that might otherwise be too costly for individual districts. RSDs have an elected board and rely on millages to fund their operations. While RSDs do not hold authoritative power over their constituent districts, they commonly seek to influence educational improvement efforts by providing their constituent districts with technical assistance and resources for high-priority improvement goals and policies. Nevertheless, the districts remain autonomous and free to make independent decisions irrespective of the goals or initiatives advanced by the RSD. The Regional School District participating in this study (hereafter called "the RSD") included 20 constituent districts, each with it's own administrative structure, budget, operations, and decision-making authority.
During the time period for this study, 2006-2009, the RSD's work took place in the context of significant shifts in state education policy. These shifts were due, in part, to a state recession and growing belief in the need for a knowledge-based economy to supplement the declining manufacturing industry. In this context, the state department of education adopted new standardized curriculum, increased graduation requirements, and focused more attention on assessing student achievement outcomes. Shifting demographics also increased the importance of educational programs that would meet the needs of impoverished and underserved populations including English learners and transient families. In response to these practical and policy demands, both regional and local districts sought multiple ways to improve the quality of public education.
The RSD participating in this study had a reputation for being forward thinking and a leader in the state with regard to educational reform. In response to the educational challenges described above, they developed an ambitious systemic improvement initiative that aimed to increase educational capacity in five areas, one of which was literacy. A key component of the literacy initiative was to develop and implement literacy coach roles in all 20 of its independently operated constituent districts. Coaching was understood as a way to support and strengthen literacy teaching and student learning. Whereas the RSD could not mandate the implementation of coaching, they could direct resources toward that end and, in 2006, the RSD hired a team of consultants to develop and lead a literacy-focused professional development program intended to build capacity for coaching. Later that year, they launched the Literacy Coaches Network (LCN) with 90 participants from 19 2 of their 20 constituent districts. In subsequent years that number expanded to 115 participants from all 20 constituent districts. The aim of the Literacy Coaches Network was to provide aspiring coaches, primarily classroom teachers, with opportunities to interact, learn from one another, and build a network of colleagues from which to exchange ideas and experiences.
The design of the LCN in the first year included 8 day-long training sessions that took place at a centralized location at a cost of $75 per participant annually, which was paid by the sending district. This configuration altered slightly in subsequent years, to accommodate two cohorts: one for new participants and a second for "continuing" participants. The LCN sessions included opportunities to learn about the work that literacy coaches do, the content matter that coaches need to know, and procedures for communicating that content to teachers. The participants read three books about literacy coaching and networked with other aspiring coaches from within their districts and from other districts. Principals were invited to approximately half of the LCN sessions to learn about coaching alongside their participating teachers. Given the districts' status as autonomous agents, the RSD administrators and the LCN leaders recognized that the training outcomes would likely vary by district and even encouraged participants to develop coaching roles that would fit the needs of their local context.

Data Sources and Collection Procedures
The data reported here come primarily from annual interviews that were conducted in three consecutive years with assistant superintendents of curriculum (or their designees) from all 20 districts of the RSD. Typically, assistant superintendents supervise literacy coaches and/or control the budget for such instructionally-related personnel expenditures. In some districts, there was no assistant superintendent or someone other than the assistant superintendent was responsible for overseeing the literacy coach initiative. Thus, in the first year interviews were conducted with 14 assistant superintendents, 1 superintendent, 4 principals, and 3 program coordinators for a total of 22 interviews with district representatives (In two districts, responsibility for literacy coaching was shared by two people so I interviewed two people). A similar interview schedule was repeated in years two and three.
The semi-structured interview protocols included open-ended questions that focused on four areas of inquiry: (a) How would you describe the district? (b) What kinds of literacy-related roles currently exist in your district? (c) What factors facilitate the implementation of literacy coach roles? (d) What factors constrain the implementation of literacy coach roles? These broad inquiry fields elicited information about the districts' policy positions, decision outcomes and sources of influence without asking leading questions about individual interests, ideology or sources of information. Participants were asked to elaborate, explain, and provide illustrative examples.
In addition to annual interviews with the district-level leaders, each year I also interviewed the Assistant Superintendent of the RSD and the two lead consultants in charge of organizing the literacy coach training sessions. These semi-structured interviews focused on understanding (a) the organizational structure and purpose of the RSD, (b) the history, design, and purpose of the LCN and, (c) the perceived outcomes of the LCN and possible future plans. These interviews supplement the interviews with district-level leaders. All of the interviews were audiotaped and transcribed except in the case of one district-level administrator who declined to be recorded. Upon completion of each interview I wrote field notes to capture preliminary insights and emerging themes.
To gain information about the RSD's efforts to build capacity I observed 22 day-long Literacy Coaches Network sessions over the course of the three-year study. The purpose of the observations was to understand the delivery of the program content, to observe the kinds of interactions participants had with one another and to talk informally with the participants about their experiences. During these sessions I sat with the participants, took electronic field notes, and participated in some of the activities. In addition, I collected all of the training manuals and learning materials distributed as part of the LCN training program.

Data Organization and Analysis
Interview transcripts and field notes were organized and analyzed with the assistance of NVivo software. The analytic procedures included the development of descriptive and thematic codes (Creswell, 2013). In the first phase of coding, the principal investigator and a research assistant descriptively coded all the interviews, expanding and refining the codes as study participants revealed new information (Coffey & Atkinson, 1996). We worked collaboratively to reconcile any coding differences and reviewed the entire set of interviews to ensure that codes had been applied consistently across the data set. In the second phase of analysis, the principal investigator applied a deductive coding scheme using concepts derived from the literature on coaching and policy making (Miles & Huberman, 1994). Coded data were organized into matrices to facilitate a search for patterns across the data set as well as disconfirming evidence. In the process of analyzing the data I sought to understand the influence of information, ideology and interests on districts' decision outcomes in the context of regional capacity building.

Limitations
This study examines district decision-making, focusing on the district as the unit of analysis. While I recognize the challenge of making singular assertions about a large organization made up of diverse people, I nonetheless aimed to characterize districts' experiences distinct from the experiences of individuals. As such, I do not provide details about the interviewees' background, skills, or experiences. At the same time, I rely on individual administrators to convey information about their district, which presents the risk of interpreting a single individual's experience as the "district" experience. To address this challenge I asked the interviewees to speak as district representatives and to share the range of viewpoints prevalent in their district. All of the interviewees had regular contact with a wide range of district stakeholders and indicated that they were familiar with others' viewpoints.
Administrators' reports provided a means to learn about the ways in which ideology, interests and information influenced their decision outcomes related to the implementation of literacy coach initiatives. I relied exclusively on the perception of those administrators responsible for making decisions about allocating resources to instructional support personnel. The use of selfreports recognizes that the administrators' perceptions shape how they respond to reform pressures (Spillane, 1998;Spillane, Reiser, & Reimer, 2002). As such, I treat the administrators' reports as legitimate interpretations that can influence their decision outcomes. At the same time, the administrators' reports should not be interpreted as the sole perspective on coaching or coaching initiatives; rather, I give these reports credence because of the administrators' decision-making status, not because they accurately reflect the majority view. Furthermore, to the extent that this investigation relies on self-reports, the findings should be interpreted as exploratory and in need of further investigation.

Note on Terms
The districts in this study did not use common language to talk about coaching. Thus, districts with similar kinds of roles often used different terms to talk about these roles (e.g. academic support teachers, cognitive coaches, reading specialists, reading teachers, literacy consultants, etc….). Alternately, some districts used the same terms to talk about very different roles. For this reason I use the term "coach" to describe roles that focus on teachers' learning, whether those roles are fullor part-time positions. To further distinguish between different types of coaching roles I draw on a classification system that I developed in conjunction with an earlier analysis of coaching in these 20 districts (see Mangin, 2009). That analysis found three different orientations toward coaching: classic, modified and none (see Table 1: Literacy Coach Role Typology). These orientations recognize that not all districts aim to create "classic" roles but rather, may adapt "modified" roles to fit localized needs or may have legitimate reasons not to create coaching initiatives as part of their reform strategy. Table 1 Literacy Coach Role Typology

Role
Aim Example

Classic
To develop school-level positions that include release time for teachers to work with their colleagues on instructional improvement.
• Full-or part-time literacy coaches focus their work on: improving teachers' literacy instruction, adoption of new literacy programs, use of common literacy language, and/or incorporation of literacy into other subject areas.

Modified
To develop positions that do not include release time from teaching and/or that include tasks other than working with colleagues on instructional matters.
• Literacy coaches spend majority of their time working as: media specialist, reading specialist, classroom teacher, test coordinator, or other role.
• Classroom teacher is given literacy coach title; provides professional development to colleagues on in-service days, after school, during lunch and/or planning periods.

None
To maintain existing literacy practices.
• Existing system of paraprofessionals and reading specialists focus improvement efforts on students through pull-out or push-in assistance programs.
• No specialized roles for literacy instruction exist outside regular classroom teachers.

Findings and Discussion
I present the findings for this study in three sections. First, I describe the decision outcomes articulated by the district administrators, noting the ways in which districts shifted their policies over the course of the three-year study. Second, I report findings related to the intersection of information, ideology and interest, paying special attention to the influence of the RSD's capacity building efforts on districts' policy positions. Finally, I describe how leadership appeared to influence districts' policy agenda and decision outcomes.

Districts' Decision Outcomes
All 20 of the constituent districts participated in the Literacy Coaches Network yet the districts did not move forward with coaching initiatives at the same pace or in the same way. This variation is not surprising given that the districts had a high level of autonomy. Districts were not obligated to develop literacy coaching as part of their participation in the LCN. Indeed, the LCN leaders repeatedly stated during interviews that they did not have a singular vision for how coaching would be implemented or a particular model they aimed to promote.
Every school district needs to make those decisions based on their needs, personnel, and the human and financial resources. I think one of the successes of the literacy network is because we give them some sound principles about having a well-defined literacy vision, about having common language, about having shared leadership, about the importance of having your staff learning just like your students are learning. Then [the district] cuts out the dress so to speak, or the fabric you use for the dress, to make it fit (LCN Leader, Year 2, p. 11). This analogy likens district-level design of coaching initiatives to a tailor making different dress patterns from the same fabric. Thus, the LCN leaders treated variation in districts' decision outcomes as a logical and appropriate outcome. In this context, it is helpful to examine the kinds of roles that developed across the 20 districts (see Table 2). Prior to attending the LCN, five districts had already implemented coaching efforts. In these districts, educators with diverse titles (e.g. academic support teachers, reading specialists, reading itinerants) were responsible for providing teachers with professional learning, guidance, and support in much the same way a literacy coach works. At the conclusion of the first year of the LCN, nine additional districts had moved forward with coach-type roles although many of these were "modified," focusing on teachers' learning in addition to students' learning. At the conclusion of the third year of the LCN, an additional six districts had made progress with coach-type roles, creating new coach roles, improving upon existing roles, or moving toward "classic" roles focused exclusively on teachers' professional learning. Moreover, four districts indicated plans for further progression with coach roles in the upcoming school year.
Taken together, the capacity building efforts of the LCN appear to have tripled the coaching initiatives across the region. When the training program began, only a quarter of the districts had coaches. At the end of the three-year study, three-quarters of the districts had implemented some kind of coaching initiative aimed at increasing teachers' opportunities to learn. It would seem that the capacity building approach to implementation was successful. At the same time, individual school districts progressed unevenly and differentially. In the following section I describe how ideology and interests combined with the information flowing from the RSD and elsewhere to influence districts' policy positions and decision outcomes with regard to coaching.

The Influence of Ideology, Interests and Information
According to Weiss (1983), decision outcomes are the result of the interaction between policy actors' ideology, interests, and information. This confluence of forces affects the policy positions that actors take and their propensity to act in accordance with policy goals. Thus, examining the districts' policy positions in the context of Weiss' decision-making framework can provide insights into the uneven progress and differential decisions that typified the 20 autonomous districts.

Information
Weiss (1983) explains that information can be any source of knowledge that policy actors bring to bear in the decision-making process. This includes "scientific research, experienced judgment, folk wisdom, or gut feeling" (p. 226). Policy actors may be exposed to multiple sources of information simultaneously or influenced by some sources more so than others. While it is difficult to know the full range of information that may have influenced individual policy actors in this study, the participants indicated that information from the LCN was a key factor in their decision-making.
According to district administrators, the primary source of information about coaching came from the Literacy Coaches Network (LCN). Many described their participation in the LCN as focusing their attention on coaching as a strategy for improving literacy and positively influencing their decision to implement coaching initiatives. One Assistant Superintendent whose district developed classic coach roles explained, I couldn't have done it without the RSD. They have provided all of the professional development. They are the ones who continue to bring the folks together to give them the opportunities and network with other lit coaches from other districts. I mean I can't do it without our RSD. All those folks have been instrumental in helping this whole process. (District 19, Year 2, p. 16) Through the LCN, participants were exposed to many types of information including expert opinion on coaching strategies, research on literacy instruction, and experiential knowledge from other coaches.
Even districts that opted to implement modified coach roles described the LCN as a key source of information for designing their coaching initiatives. For example, the Director of Instruction and School Improvement in District 2 explained that the LCN increased the reading specialists' willingness to take on coaching responsibilities. He described the level of influence from the LCN as, Huge. I would say very, very huge. I would say the knowledge that the specialists gained and the experiences they had helped not only them but even me because I had an opportunity to visit some of the trainings a few times. It helped me see the difference between what a reading teacher does and what a literacy coach does and how important both pieces are and what an impact it can have on student achievement. So I would say it had a very, very large impact on our vision and on our direction. (District 2, Year 3, p. 8) Such claims about the utility of the LCN for building districts' capacity to implement coaching initiatives were common, yet the districts progressed unevenly in their development of coaching and exhibited differential responses. This variation is likely due to the influence of multiple sources of information but also to the interaction of information with ideology and interests. In some cases, the confluence of information, ideology and interests lead to increased willingness to implement coaching. Other times, they interacted to limit coaching initiatives despite three years of capacity building efforts by the RSD.

Ideology
As explained above, ideology encompasses "philosophy, principles, values, and political orientation" (Weiss, 1983, p. 224). In the context of schools, ideology can refer to beliefs about students, educators, the purpose of schools or any other aspect of education. Ideologies are understood as influencing decision outcomes by injecting emotion into the decision making process. Indeed, decision-making is never purely rational (Simon, 1986) and examining the underlying ideologies that influence policy positions can help us understand decision outcomes. Because I conducted interviews exclusively with administrators, the ideologies captured are either the administrators' beliefs and values or their perceptions of their colleagues' ideologies, that is, the beliefs and values held by other administrators, teachers or specialists.
The most pervasive ideology that emerged was a belief that educators' time should be spent working with students. This belief focused on the idea that students need learning supports and remediation more so than teachers. Stated crudely as a "fix the kids" mindset, this belief predisposed educators to retaining roles focused on providing supplemental services to children, such as paraprofessionals and reading specialists. Administrators described educators with this ideology as demanding assistance for students they viewed as being difficult to teach, behind grade level, or needing different learning strategies. This ideological stance is evident in the following passage in which the Reading Director from District 17 describes what would happen if reading specialist roles were shifted to include more time spent with teachers and less time working with children. She exclaimed, The teachers would be screaming. I'm telling you that right now. Especially the lower el[ementary] teachers. The second grade teachers weren't real hot on the idea that the specialists were just going to come in rather than pull [students] out… [Spending less time with students] would be a really, really bad move. (Year 1, p. 12) In this case, moving from a "pull-out" to a "push-in" reading intervention model had been a hurdle. Thus, the Reading Director felt certain that teachers would oppose altering the reading specialist role to include more time providing professional learning to teachers. Other districts expressed similar views about the shift to coaching despite participating in the RSD training program where they received information about the value of coaching.
Despite administrators' perception that some school-level educators embraced ideologies that were counter-productive to coaching, many of those same administrators expressed the importance of changing existing ideologies. Administrators described gradually working toward implementation of coaching practices, typically by changing the responsibilities associated with existing student-focused roles. One elementary principal who was in charge of overseeing the reading teachers described this challenging shift toward coaching; "Teachers had some built up animosity towards the new literacy coaches [formerly called reading teachers] because they weren't coming into their room and pulling kids out and fixing them" (District 4, Year 3, p. 4). He went on to describe how teachers came to embrace the literacy coach role over time: At the beginning of this year we began to do some real deep data digging and teachers actually began to understand how to use data to determine instructional practices. That was led by our literacy coaches and all of a sudden, the light went on, and people went "Oh my gosh. I didn't realize this is what DBLS score meant and you've given me several things I can do in practice to make it happen." And then …the relationship between teaching staff and literacy coaches took off. In December, the comment was made to me at a staff meeting, "we don't know how we ever got along without them." So the change was dramatic and now the literacy coaches are looked upon almost with reverence. (District 4, Year 3, p. 4) In this example, teachers' ideology shifted in response to new information about the utility of coaching, leading to new beliefs about the value and importance of coaching. This gradual shift toward coaching exemplified all of the districts described as having "modified" roles.

Interests
According to Weiss (1983), interests refer primarily to self-interest, including but not limited to, one's desire to maintain or increase status, legitimacy, position, influence or material gain. Administrators participating in this study commonly indicated that their policy position regarding coaching was directly associated with the interests of the persons who would be affected by the development of a coaching initiative. Thus, the self-interests of teachers, para-professionals, and reading specialists influenced the administrators' decision whether to develop a coaching initiative. The primary interest that emerged was the desire to maintain job status and job stability. Specialists sought to maintain their status as expert interventionists and resist changes to their role that would force them to learn new skills or, possibly, eliminate their positions all together. Likewise, teachers sought to retain specialists who provided either pull-out or push-in services that they perceived as alleviating their work load. Combined, the interests of teachers and specialists influenced administrators to develop policy positions that limited coaching.
For example, District 17 had dedicated extensive resources to the development of a longstanding reading specialist program. Specialists were regarded as reading experts and teachers relied on them to work with challenging students. Both teachers and specialists were vocal about their desire to retain the existing program. The influence of self-interest on decision making and the specialists' power to shape the decision-making process are conveyed in this interview excerpt: Reading Director: I honestly don't know where this will all go next year. We're going to discuss that at a meeting next week.
Interviewer: Who will participate in that discussion? Who gets the chance to weigh in?
Reading Director: They do, the reading specialists. They get a heavy weigh-in. They weigh more than me.… I have to get the okay from the curriculum assistant superintendent and I have to get the okay from the principals before I can do anything. They weigh heavy but the reading specialists weigh the heaviest, which is really the way it should be. They're doing the job. (District 17, Year 1, p. 7) In this case, it was in the director's self-interest to support the interests of the teachers and specialists. As such, her need to maintain status and job stability for herself and specialists appears to counteract the influence of the information received from the LCN.
While some administrators aligned their own interests with those of teachers and specialists to buffer against the shift toward coaching, other administrators actively worked to shift teachers' understanding of coaching and demonstrate how it could align with their self-interest. To do this, administrators often relied on data (see District 4 above), making the case that their existing strategies were not working and a different approach could be beneficial. This was especially evident in districts with an increase in students needing reading intervention, which made it difficult for the reading interventionist to meet all the students' needs. In District 11, where free and reduced-lunch rates had climbed to 70% by the second year of the study, the Assistant Superintendent of Instruction and Personnel described the importance of helping teachers understand how coaching could support the students' best interest and, relatedly, the teachers' interests as well.
We have taken a look at some of the research on literacy coaching and how that could work. And we have a growing concern for the number of kids that repeat the reading intervention process. In other words, [students who] had reading intervention in second grade and in third and again in fourth. "Well, what's going on?" So we started to take a look at the capacity of the classroom teacher in terms of literacy skills and ask, "how well can a classroom teacher support the student who's at risk in order to make sure that they don't necessarily have to go back in" (District 11, Year 1, p. 3). In some districts, growing populations of low-level readers or English-language learners combined with teachers' uncertainty about how to service these children in the regular classroom created a convincing argument that coaching may serve teachers' interests more effectively than an intervention program. Thus, teachers' self-interest provided additional incentive to positively influence coach implementation.
In the context of this study, Weiss' (1983) decision-making framework was useful for examining how ideology, interests and information interacted to influence districts' decision outcomes with regard to coaching. At the same time, my analysis and search for disconfirming evidence revealed an additional variable that is not accounted for in the I-I-I framework: leadership.

The Influence of Leadership
Over the course of this three-year study, the participants repeatedly pointed to changes in leadership as precipitating changes in their decision outcomes. The formal leaders played an important role in shaping the interplay between information, ideology, and interests. Research indicates that superintendents and principals lead by setting direction, developing others, and building effective structures (Hallinger & Heck, 2002;Leithwood, Louis, Anderson, & Wahlstrom, 2004). This was true of the administrators in this study, whose formal job responsibility was to perform these leadership tasks with the aim of improving instruction. Thus, the formal leaders were in a position of power, capable of prioritizing some information, ideologies, and interests over others. The end result was variable progress with regard to coaching as a strategy for improving instruction.
Given that all of the districts sent participants to the Literacy Coaches Network, it may be surprising to learn that not all district leaders supported the notion of literacy coaching. One leader admitted sending teachers to the LCN to "appear collegial" and two others gave their teachers permission to attend, knowing that there was little support for implementing actual coaching initiatives. The Assistant Superintendent of Instruction in District 5 explained that, counter to the goals of the RSD, her district was focused on math and science.
I didn't commit one way or another on the literacy coaching. through the reading specialists. (District 5, Year 1, p. 14) While this administrator seems to suggest that teachers' interests and ideology would weigh into the decision process, she also conveys that her own ideals favor a student-focused model such as improving literacy through the use of reading specialists. As such the leader appears to prioritize her ideology in the decision-making process.
In some districts, the loss of a supportive leader lead to stagnation of the coaching initiative (Districts 12 and 13) or a reversal away from the development of coach roles (District 18). Other times, gaining a supportive administrator could favorably shift the direction of the district's coaching initiative. For example, following the second year of the LCN, District 16 hired a new Director of Instruction with a Ph.D. in literacy. Up until that point, coaches in District 16 had two hours each day to coach their teacher colleagues. Under the new Director, coaching became a full-time position. In addition, the director met with the six coaches monthly, sent them to literacy conferences, and purchased materials for them. The Director's literacy knowledge, her beliefs about the value of coaching, her personal interest in making the coach roles successful, and her positional power as the director of instruction, afforded her a great deal of influence over the direction of the coaching initiative.
Other districts had similar experiences. In District 4, where the central administration had little interest in coaching, a school principal was assigned to the task of overseeing the coaching initiative. During the first two years of the LCN, the principal reported making little progress toward coaching despite his personal belief that coaching was important. He explained that individual principals had discretion over coach role design and none had shown interest in developing coaches in their buildings despite sending participants to the LCN. However, when the district hired a new superintendent, the outlook on coaching shifted dramatically: The major key to helping us become a more cohesive district is that we got a new superintendent. And luckily for me the new superintendent and I agree on instructional philosophies and curriculum and we both agree that data is a key to determining your instructional practices. …That has filtered from the central office, to the administrative staff, to classroom teachers, to our literacy coaches. And just the ability to look at data and make decisions based on evidence instead of feeling has made a dramatic impact on what our literacy coaches do (Principal, District 4, Year 3, p. 4). New leadership brought new focus to the district and greater attention to coaching as a means for improving literacy. As such, the interplay of information, interests and ideology was directly related to the distribution of decision-making power in the district and to the role that formal leaders play in setting direction, developing others, and building effective structures (Hallinger & Heck, 2002;Leithwood, Louis, Anderson, & Wahlstrom, 2004).

Conclusion
Evidence from this three-year study on districts' decision outcomes suggests that regional capacity building efforts can significantly increase the implementation of coach roles. In this case, the training provided by the Regional School District (RSD) appeared to triple the number of coaching initiatives across the region. This growth may be understood as the result of increasing educators' understanding of coaching and their human capacity to develop coaching initiatives. Thus, capacity building efforts at the regional or state level may be an effective policy tool for implementing coaching initiatives at district-and school-levels.
At the same time, policy implementation is not a singular task; rather, it is a process enacted over time and influenced by a confluence of factors. My analysis of the data using Weiss ' (1983) decision-making framework provides further insights into the influence of information, ideology, and interests on districts' decision outcomes. Although the districts reported being influenced by information from the RSD, they also indicated that educators' self-interests and ideological beliefs about how teachers should use their time influenced the school districts' policy positions and their decision whether to implement coaching initiatives. These findings support the notion that capacity building, as a policy tool, is both necessary and insufficient for reform. Moreover, districts should anticipate the affects of ideology and self-interest, or "politics," on policy implementation efforts, even in the context of rigorous capacity building efforts aimed at providing the information and resources necessary for reform.
Finally, district decision outcomes were shaped by the distribution of decision-making power, specifically, by the formal leaders' influence on the interplay of information, interests and ideology. Although Weiss's (1983) decision-making framework does not specifically address the influence of leadership on decision outcomes, she explains, "The distribution of power determines WHOSE ideology, interests, and information will be dominant" (p. 239). In this study, shifts in formal leadership directly influenced districts' policy positions and their propensity to build on the resources and learning opportunities provided as part of the RSD's capacity building initiative. This finding supports existing research on educational leadership and extends Weiss's (1983) decisionmaking framework to recognize the influence of formal leadership structures traditionally found in schools. Future research that utilizes micropolitics and informal leadership as lenses for understanding policy implementation could further deepen our understanding of the decision outcomes districts make related to coaching initiatives.

About the Author
Melinda M. Mangin Rutgers University melinda.mangin@gse.rutgers.edu Melinda M. Mangin is an associate professor of educational leadership at Rutgers, The State University of New Jersey. Her scholarly interests include instructional leadership, teacher leadership, professional development, and preparing teachers and administrators to be school leaders. She is the co-author, with Sara Ray Stoelinga, of Effective Teacher Leadership: Using Research to Inform and Reform (2008).

About the Guest Editor
Sarah Woulfin University of Connecticut Sarah.Woulfin@UConn.edu Sarah Woulfin is an assistant professor of Educational Leadership at the University of Connecticut, Storrs. She studies the relationship between education policy, leadership, and instructional reform. Using lenses from organizational sociology, she investigates how leaders influence teachers' responses to reform efforts. In her doctoral work at the University of California, Berkeley, she focused on institutional theory, policy implementation, and coaching. She has published in the American Educational Research Journal (AERJ) and Reading Research Quarterly. Currently, she is an associate editor for Educational Administration Quarterly (EAQ). She is also on the executive steering committee of the Districts in Research and Reform SIG at AERA. From 2009-2012, Dr. Woulfin served as the program chair for AERA's Organizational Theory Special Interest Group. As a former urban public school teacher and reading coach, she was dedicated to strengthening students' reading and writing skills to promote educational equity. As a scholar, her commitment to raising the quality of instruction motivates her research on how policy influences-and is influenced by-administrators and teachers. Readers are free to copy, display, and distribute this article, as long as the work is attributed to the author(s) and Education Policy Analysis Archives, it is distributed for non-education policy analysis archives editorial board