Reforming for “ All ” or for “ Some ” : Misalignment in the Discourses of Education Reformers and Implementers

The ways in which the language of reformers intersects with and informs reform implementation is important to our understanding of how education policy impacts practice. To explore this issue, we employed critical discourse analysis (CDA) to analyze the language used by a 21 century skills-focused reform organization to promote its program alongside the language that local actors used to explain its implementation. We examined source materials, field notes, interview data, and publicly available organizational data collected over a five-year period to critically examine how discourse 1) illustrated alignment between the stated and implicit audience for the school reform program and 2) Reforming for “All” or for “Some” 2 shaped subsequent implementation. Analyses suggest the reform organization promoted itself through a discourse that all students in all reform schools were being prepared for college, career, and civic life. There was a significant misalignment, however, in the discourses regarding the appropriate student audience for the reform. Local actors questioned whether the reform program 1) was suitable for all students and 2) provided necessary supports for all students in all schools. This misalignment led to uneven implementation and resulted in some educators dismissing the goals of the program as unrealistic. Given that educational agencies have considerable freedom to choose among diverse reform programs, our analysis suggests it is important to understand the discourses through which reform organizations advertise models, implementers justify adoption, and educators respond.


Introduction
Starting in the 1980s with A Nation at Risk, and crystallizing in policy with the 2001 No Child Left Behind Act, the language used to promote education reform has fostered an idealized policy discourse with broad appeal across the political spectrum (Mehta, 2013).Education reform organizations (e.g., Teach for America), charter school management organizations (e.g., KIPP), and policy names (e.g., iterations of the federal Elementary and Secondary Education Act) represent principles of equality, possibility, and universality that epitomize this discourse.Phrases like "one day, all children," "no excuses," "no child left behind," and "every student succeeds" promote a vision of education as the mechanism through which the American dream can be realized.These phrases have emerged out of and continue to embody the policy discourse that all schools can ensure that all students achieve success in school on the path toward economic prosperity.This discourse has united many policymakers and education leaders around market-based and marketoriented school reforms that are politically problematic to challenge (Hursh, 2005;Lahann & Reagan, 2011) because they are grounded in theories of student choice, agency, and performance.The simplicity inherent within the message that all schools can prepare all students for success in college and career, though, papers over the complexities involved in successful implementation of reforms intended to do just that.
Despite the popularity of this idealized policy discourse, student performance has shown that no educational initiative has ever truly served all students well (Deschenes, Tyack, & Cuban, 2001).Even highly "evidence-based" or "proven" programs in the What Works Clearinghouse result in many students performing the same as or below what would have been expected without the program treatment.The children most often left behind in any educational innovation are those the U.S. educational system has never consistently prepared well: students of color, low-income children, English language learners, and students with disabilities.Many policies and reform initiatives that contribute to the idealized policy discourse do indeed produce positive results for many students.However, the potential problem with the idealized discourse is that it may not be shared by those responsible for implementing reforms.Consequently, there may be implications for how reforms are enacted, how useful they are in supporting changes in practice, and, ultimately, the extent to which they are able to improve educational outcomes for children.These implications are especially important for low-performing schools, which are common targets for reforms claiming to realize the "all students" discourse.Because competing discourses may affect implementation, we approached this work from the critical perspective that the language used to frame policy raises implications for who holds power-and responsibility-in policy implementation.
The idealized discourse described here has motivated numerous 21 st century educational reform organizations to develop reform programs that claim to prepare all students for an array of post-secondary and workforce opportunities.Yet, there is a lack of evidence on whether the actors who adopt and implement these reforms also understand that these policies are intended for all students to be able to pursue college and career.In fact, research on school reform in lowperforming schools suggests that educators may hold deficit views of their students which could undermine external efforts to improve their practice (Berman & Chambliss, 2000;Berman, Chambliss, & Geiser, 1999;Valencia, 1997).Moreover, the policy literature suggests that misalignments in how the discourses promoting 21 st century school reforms are interpreted may adversely affect implementation, resulting in poor or inconsistent results across sites (Chase, 2014;Coburn, 2001;Datnow, 2007;Datnow, Hubbard, & Mehan, 1998).
To build on the policy implementation literature, we use data from a multi-year case study of a national 21 st century school reform program to analyze the discourses of reformers alongside reform implementers.The analysis focuses on the following questions: 1) Do reformers and implementers use similar discourses to describe the audience for a 21 st century school reform program?2) Do reformer and implementer discourses about the audience for a 21 st century school reform program impact implementation?In responding to these questions, we first review how policy implementation has the potential to vary across different discourse communities.Next, we explain how we used critical discourse analysis, or CDA, to understand how the discourses used to promote the reform program influenced how local actors decided to adopt and implement the program.Then, we discuss the implications of power asymmetries in relation to policy adoption, implementation, and the potential impact of reform on students, particularly those who have been traditionally underserved in the U.S. educational system.

Implementation Across Discourse Communities
Ideas expressed in educational policy face many hurdles on route to implementation.In part, this is because discourses of education policy are not translated in a linear way through the policy adoption and implementation process.Lipsky (1997) argues that the decisions, routines, and processes acted out by street-level bureaucrats-in this case, teachers and school leaderseffectively become policy.Street-level bureaucrats constantly negotiate their choices within the conditions in which they work, including the policies meant to change or improve their practice.An important component in this negotiation is the extent to which public servants are capable of acting on their commitments to service work.Lipsky (1997) suggests that, although public service employees enter their professions with idealized notions of the power of their work, the realities they face often undermine these beliefs.
Educators, therefore, may enter the profession holding idealized beliefs about what schools can accomplish for all children.It is in the daily enactment of their roles that they must confront how to realistically achieve those ideals given the practical constraints of working in challenging school settings.This can manifest in deficit thinking, wherein educators place the responsibility for student performance on families, society, or students themselves, rather than taking ownership for the power that they have to influence student outcomes (Valencia, 1997).This phenomenon can be common in low-performing schools, particularly when teachers and students come from different cultural traditions (Delpit, 1995).Deficit thinking hence may be one way in which educators explain the gap between idealized notions of their work and the realities they face enacting it.Rather than placing the responsibility-and, in turn, the blame for failure-on themselves, educators may place the responsibility for success on students.
If the idealized policy discourse is not fully shared by practicing educators, then successful implementation likely becomes difficult.As Hill (2006) observes, different discourse communities make meaning out of policy language in different ways, and varied professional and personal experiences shape policy implementation.Although the gap in discourses between reformers and educators itself may be problematic (if unsurprising), the ways in which different kinds of policy discourses potentially affect policy implementation are less clear.For instance, discourses reflecting policy intent may be especially important for implementation given that they set the stage for understanding, accepting, and then carrying out policy in practice (Yanow, 1996).How educators translate policy messages about intent, perceive those messages through the lenses of their own prior experiences, and then negotiate them within the context of program implementation is germane to questions of policy implementation and impact.It is not just that implementers have preconceived discourses about policy intent that they enact during implementation.Rather, it is also that implementers' beliefs and discourses shift through the process of implementation in a perpetual negotiation of understanding what policy expects of them, what happens when they enact it, and how they use that information to understand future policy goals.
The discourses that educators bring to bear as they implement policy interact with idealized "all children" policy discourses.Though idealized discourses likely appeal to many educators, they are complicated by individual professional experiences.Not only must teachers make sense of policy as written, they must also implement policy across the gulf that separates the idealized discourse of reform from the discourses that have emerged out of their own lived experiences.These ideals are difficult to realize, and success depends upon myriad factors such as leadership, working conditions, collaboration, professional training, student preparedness, and community support (Ross & Bruce, 2007;Ross & Gray, 2006).Questions about the intent of school reform can emerge in implementation when educators attempt to bridge this divide (Figure 1).As educators implement reforms, they confront gaps between what reformers expect of them and what they are able to accomplish.Discourses about reform, then, offer explanations for those gaps.
The question of audience ("for whom is this program intended?") is fundamental in the enactment of school reform.The purpose of a particular school reform implies a set of organizational structures to support it (Lenhoff, 2013;Mehta & Fine, 2015).Misalignment in discourses about the purpose of reform may indicate weaknesses in organizational and human capital infrastructure that would be best suited to advance reform goals.Likewise, the target audience for school reform programs may affect educators' self-efficacy and implementation strategies.If educators' street-level discourses suggest that only some students will benefit from a particular reform, their understanding of whether an "all students" policy discourse is achievable or even worthy of pursing is an important question.As Deschenes, Tyack, and Cuban (2001) write, the current reform discourse of "for all students" will produce some students who fail to meet reform expectations.Tensions between policy and street-level discourses may be even more pronounced, then, in schools that have already been identified as "failing" by state or federal accountability policies, such as School Improvement Grant (SIG) schools.How this failure is foregrounded and shaped by the discourses of policymakers and educators helps to explain policy implementation, fidelity, and outcomes.

Critical Discourse Analysis
To investigate how discourse influences policy development and adoption, we turn to critical discourse analysis (CDA) as a form of critical policy analysis.CDA consists of a loosely organized school of thought (Wodak, 2004) that offers a wide array of theoretical and methodological options.Taken together, varying approaches to CDA work at the intersection of language and social theory (Rogers et al., 2005) to identify what discourses produce (Wodak, 2004).What holds across differences is a shared commitment to social theory, inequality, ideology, and the distribution of power.These interests often cause critical discourse scholars to gravitate toward power imbalances-or asymmetries-particularly as they relate to political discourse, ideology, racism, economics, advertising, media, institutions, and education (Blommaert & Bulcaen, 2000).Although CDA operates across a number of scholarly disciplines, it is particularly useful from a policy perspective in "documenting multiple and competing discourses in policy texts, in highlighting marginalised and hybrid discourses, and in documenting discursive shifts in policy implementation processes" (Taylor, 2004, p. 433).As such, we use CDA to attend to promotional discourse (Bhatia, 2005) and the ways in which discourse reproduces and challenges sociopolitical structures.
It is important to observe, however, that 'discourse' is perhaps not the unifying concept that it might appear to be, especially given differences in how, for example, political scientists and linguists approach the term (Wodak & Chilton, 2005).To clarify our use of the term in this study, we theorize discourse within macro-analytic traditions.Macro perspectives view discourse as a series of statements that join together to create discursive formations, or social conditions under which particular statements can be made (Foucault, 2002; see also Fairclough, 2013).Discursive formations lead to the invisible, unspoken rules of society by which institutions function, including education.Notably, the identification of discursive statements and formations departs from other variants of CDA, particularly those that focus on micro-analytic techniques.In contrast, macro-analytic CDA involves an iterative, deductive process in which "the analysis of the statement and that of the formation are established correlatively" (Foucault, 2002, p. 130).Being that the unit of analysis within macro-level studies is the statement (rather than the specifics of syntax or grammar), macro perspectives on discourse are more closely aligned with political science than linguistics.
This study, therefore, is a CDA in that it critically examines discourse in the nexus of theory, methodology, politics, and language (per Meyer, 2001).By examining participants' statements within broader historical-cultural, socio-political contexts, we generate critical understandings of what the discourses of "for whom" produce in education.Because we draw from traditions of CDA that are explicitly political in aim and scope (Lazar, 2005), we suggest that a critical agenda potentially complicates overly simplistic, top-down discourses of "all children" while problematizing one-sizefits-all approaches to education reform.
A macro-analytic approach to critical discourse analysis, though, is not without challenge or debate.In particular, macro perspectives often lend themselves to what Antaki, Billig, Edwards, and Potter (2003) describe as a circular discovery of discourses and mental constructs.As Antaki et al. suggest, this has the potential to occur when 1) a set of utterances are provided as evidence for a particular discourse and 2) the same discourse, in turn, is justified by the presence of those same utterances.From a strictly Foucauldian perspective, this is less problematic.Being that discursive statements and formations are co-constitutive, "When the time finally comes to found a theory, it will have to define a deductive order" (Foucault, 2002, p. 130).Yet, as Antaki et al. observe, analyses that run the risk of circularity can be counterbalanced by critically questioning statements at each step along the way.In other words, statements-particularly those used as supporting evidenceshould be examined from multiple perspectives to determine if, in fact, the statement does indeed support a particular discursive formation.Moreover, statements can be considered within broader, competing political contexts.Situating language within larger historical, social, and political contexts also serves as a validity measure.
As such, we discuss broader political contexts throughout the manuscript (Wodak, 2001) to generate trustworthy findings within a CDA frame.We further continue data collection and analysis through saturation, or what Jäger (2001) refers to as 'completeness' of results, and triangulate by collecting diverse forms of data and analyzing data through multiple perspectives (see also Scollon, 2001;Wodak, 2001).Taken together, these validity measures respond to the critiques of Antaki et al. (2003) while following more general calls for rigorous qualitative research that uses "sufficient, abundant, appropriate, and complex" theories, contexts, samples, data collection, data analysis, and time in the field (Tracy, 2010, p. 840).

Site Description
The school reform organization examined in this study, which we call Transforming Schooling, is a non-profit that has been in operation for nearly two decades.The organization began with a project-based reform initiative in one U.S. high school before expanding a comprehensive reform program for schools in all levels of K-12 education, including elementary and middle schools.The organization now supports implementation of its project-based instructional model in more than 150 schools across the US and several schools outside of the US.Transforming Schooling was acquired and later spun off from a larger school innovation organization that currently supports multiple reform initiatives.It supports a small central staff at its headquarters, with most of its employees working remotely across the globe to provide services directly to schools.Over the last decade, Transforming Schooling has received substantial support from federal and philanthropic grants.
As it grew, Transforming Schooling broadened its reach into different types of schools.At first, the organization largely worked with well-functioning schools with middle-or upper-income student populations.Around 2010, however, it rapidly expanded to a wider variety of schools, served more diverse student populations, and began to provide a greater range of educator supports and trainings.Throughout, the Transforming Schooling reform program has remained grounded in several 21 st century principles of education and learning: student-centered, skills-based, technologyfocused, and adaptable in different settings.Affiliated schools have been named as "exemplars of 21 st century learning" by the Partnership for 21 st Century Skills, and a search of the Transforming Schooling website produces 235 references to "21 st Century" schooling.The organization's reform program includes changes to school culture, pedagogy, course configuration, staffing, technology, and academic goals.Although this study primarily focuses on the enactment of Transforming Schooling's academic philosophy and pedagogical orientation (project-based learning), our dataset also provides evidence of discourses related to school culture, structure, and technology, as these were central program components.
This study examines the discourses that emerged in three high schools.These schools were selected in 2010 because they were entering their first year of implementation of the Transforming Schooling reform program in the same state, with the same reform coach, which would allow for comparisons across different local contexts.School sites represented three variations on Transforming Schooling's implementation strategy, including one new school in a small city, one school-within-a-school in a midsize suburb, and one whole-school turnaround in a large inner ring suburb of a major metropolitan area.These schools also varied by student performance, with School 1 performing among the top quarter of schools in the state, School 2 performing about at the median, and School 3 performing among the bottom 5% of schools in the state.After failing to meet adequate yearly progress (AYP), School 3 had partnered with Transforming Schooling as its external service provider for its School Improvement Grant.In addition, the student demographics of the schools varied, with School 1 serving mostly white, middle-income students and Schools 2 and 3 serving greater proportions of students of color and low-income students.Table 1 provides descriptions of the three school sites.

Data Collection
Data were collected in two phases, which facilitated comparisons over time, as shown in Table 2.As part of an ethnographic comparative case study, the first author investigated the adoption and implementation of the Transforming Schooling reform program in three Midwestern high schools from 2010 to 2012.This phase of the study was designed to investigate how the reform organization established supports for significant instructional and cultural reform in three high schools with substantially different student populations, prior performance, and staff capacity to undertake change.Building from literature on reform program adoption, sensemaking, and implementation (Coburn, 2001;Datnow, Hubbard, & Mehan, 1998;McLaughlin, 1976;Spillane, 2004), the first author sought to understand how and why the particular reform program was taken up by local actors and the ways in which the reform organization's characteristics interacted with the local context to impact implementation.Follow-up materials were collected in 2015 to examine changes in Transforming Schooling's outward-facing discourse over time; these materials were collected from the organization's homepage and include web pages, videos, press releases, infographics, and other online informational and promotional materials created between 2010 and 2015.More specifically, the paper is informed by data collected thorough semi-structured interviews (N = 20); participant observations (N = 37); organization documents such as instructional guidance, internal surveys and rubrics, and descriptions of program products and services (N = 44); and publicly available materials such as archived website information, blogs, reports, and press releases (N = 6).In-depth semi-structured interviews spanned across Transforming Schooling and three school districts.Interview participants included teachers, instructional coaches, district coordinators, and school and district administrators.Questions were designed to explore the participants' views on Transforming Schooling and its reform program, environmental context, and program implementation, particularly of instructional reforms (see the Appendix for sample interview questions).Participant observations generated more than 120 hours of data, including multiple half-day site visits, observations of organizational trainings and administrators' meetings, and multi-day training sessions.

Data Analysis
In performing a CDA, we began by examining external materials produced by the reform organization to identify key discourses.After reading and re-reading the interview transcripts, participant observation notes, and documents, we identified examples of power differentials, or asymmetrical discourses.Asymmetrical discourses are discourses that become imbalanced when elite institutions and members of society have disproportionate influence; they manifest when powerful gatekeepers have undue influence in setting the standards of acceptability for others (Bhatia, 2006;van Dijk, 1993).In this sense, we observed differences in how stakeholders described which students the program supposedly was "for."Evidence for asymmetrical discourses next was considered from multiple perspectives.Questions about intended audience were then situated within broader policy contexts, including 21 st century education discourses, and, as we have explained, idealized policy discourses that claim to be "for all" students.This iterative analytic process, wherein both authors would read, summarize observations, share with each other, and then re-read the data, led to the identification of discourses.Taken together, repeated readings produced analyses regarding how critical differences in discourse potentially influence implementation.

Analysis
By critically analyzing the discourses of the Transforming Schooling reform program and the local actors who adopted and implemented it, we explored how alignment between the stated and implicit audience for 21 st century school reform shaped policy implementation.When reading the data set through Transforming Schooling's discourses alongside the discursive statements of program implementers (superintendents, principals, and teachers), we identified a significant power asymmetry regarding "for whom" the program is intended.In short, a program advertised as being for all children in all schools is described by implementers as only appropriate for some.As such, we organize our analysis of this asymmetry into the following sections: "for all," "for some," and "for some, but not ours."Examples of language that illustrates these discourses can be found in Table 3.

For All
Transforming Schooling promoted its reform program as being "for" all types of students, in all types of schools.Evidence of this discourse was present in conference sessions, in-person interactions with organizational staff, and, in particular, web-based promotional materials.Online materials often situated program objectives in aspirational visions of the future of public education in America.For example, a 2015 report from the Transforming Schooling website expressed a vision in which every public school is able to realize every student's full potential.Infographics in this document (such as charts, maps, and pull-quotes) demonstrated an organizational commitment to college and career readiness for students in schools with a range of socio-economic, racial, and performance contexts.A complex flow chart then illustrated a potential path to success for students.This path first flowed through student empowerment and engaged teaching, then moved through transformed learning and professional development, and finally resulted in career-and collegereadiness.The infographic supported the report as a whole, which sold the program as being beneficial "for all" students.To promote the discourse of "for all" students, a Transforming Schooling webpage dedicated to professional learning claimed that the reform program design was flexible enough to succeed in a range of different school contexts and that it "easily integrates" nontraditional curriculum programs, such as courses that combine biology and literature.The words "easily integrates" suggest that the Transforming Schooling program is simple to incorporate into the varied high schools that have adopted it, as it was created with "all" in mind.This creates a potential conundrum.If the program already is "for all" students, then there is no immediate need to differentiate curriculum based on abilities, interests, or goals.In other words, Transforming Schooling went out of its way to emphasize in web-based materials that even though students are diverse in demographics, abilities, interests, and life goals, a singular programmatic curriculum can comprehensively address all students' academic needs.
To demonstrate how the reform program serves "all students," Transforming Schooling approved several reform schools as models.Model reform program schools, which served as tour sites for recruiting new schools, ranged from those serving mostly white middle-to upper-middle income students to Title 1 schools that served mostly low-income students of color.The range of school contexts in the model reform schools also was evident in the three study schools.Yet, at the time of the initial data collection, Transforming Schooling did not differentiate its support to accommodate the differing needs of schools.The organization advertised its program as being "for all" students, and the language used by the reform organization and its proponents indicated that providing additional training, supports, or differentiated resources depending on school contexts would undermine the core message that the program already was designed for every student.
The Transforming Schooling coach for the three schools also described the program as being "for all" during on-site observations.As the coach explained, "If a school works towards building its fidelity…no matter where you are, what school you are, what the demographics are of the student body, [it] can be a successful school."During several of the coach's interactions with reform school staff, we observed tensions involving the gap between the idealized discourse of the reform organization and the discourse of local actors.This was particularly the case in the SIG school, where only some of the reform components had been implemented and essential elements, such as one-to-one student computing, had been ignored.The coach expressed disappointment that the teachers in this school did not wholly accept the "for all" discourse, but the coach also conceded that leaders at Transforming Schooling had not been as thoughtful as they should have been in considering which schools would be a "good fit."In fact, the coach described how one Transforming Schooling staffer was dismissed, in part because the staffer had been too much of a "salesman," enlisting any school that would sign on, rather than being strategic about schools that might be best positioned to benefit from the program.In contrast to comments from the coach in the first year of implementation, the coach reported in the second year of implementation that the Transforming Schooling coaching staff was considering how to differentiate instruction depending on the schools and teachers they work with.
The discourse of "for all" permeated many of the documents Transforming Schooling used to promote itself and recruit new schools.For instance, the document used to describe the conditions necessary for a school to be successful does not mention students in any way.This omission could be translated to mean that schools with any student body could successfully implement the program.In addition, Transforming Schooling described student outcomes in broad, universal terms, implying that all students who participate will be rewarded with positive results; the organization further claimed that participating students demonstrated high levels of academic proficiency and readiness for successful college and career experiences.Rather than an aspirational goal, this statement from the Transforming Schooling website was positioned in a section on "results," indicating that the organization measured success based on the standard of all students achieving high levels of academic proficiency.As part of a discourse for recruitment, it is compelling.As part of a discourse educators receive from reformers as they attempt to implement the program, however, the statement becomes an expectation for success with all students and an implication that not meeting those expectations is a failure of the educator, not the program.In a stark illustration of the gap between expectations and local implementation, a graph located below this statement on the website indicated that only 65% of Transforming Schooling graduates thought they were prepared or extremely prepared for college-level math.Educators serving some of the 35% of students who do not think they are prepared, then, may have to construct a new discourse to explain the gap between "for all" and the reality they experience in implementation.As the next section illustrates, a program intended "for all" was, perhaps more accurately, thought appropriate by implementers only "for some."

For Some
In our analysis of interview and observation data from the first two years of implementation in three reform schools, we found a discourse of "for some" that contrasted with Transforming Schooling's discourse of "for all."Nine educators across each of the three schools indicated that the reform program was appropriate for some, but not all, students.This discourse emerged in two ways: 1) as educators described differences between students in their own schools and students in other reform schools and 2) as educators described problems in implementation that were not adequately addressed by Transforming Schooling.
As evidence of misalignment between reformers and implementers, the language superintendents used to describe why they adopted the reform program revealed that they believed the Transforming Schooling program was for some students-namely their students-but not necessarily all students in all schools.The superintendent of School 1 stated: "By no means is it a model that is perfect for everybody."This language indicated that the district-level implementer did not take up the "for all" discourse being promoted by the reform organization, despite buying into the reform for his own students.This same superintendent went on to say that one of the challenges of the Transforming Schooling model schools was that they did not match the demographics of his suburban, high-performing school: "[They] were very urban districts; again, kind of broken districts.They brought Transforming Schooling in as a way to fix the problems that they were having."In contrast, this superintendent had adopted the 21 st century reform program not as a solution to problems, but as an add-on to an already successful program.The principal of School 2 in the study described the program in the same way: "Our high school is not a failing high school.We have many, many great things going on there.So we were looking to just enhance the current setting."For these reform implementers, Transforming Schooling was "for" high-performing students who needed an extra push toward career-and college-readiness.
But some implementers of Transforming Schooling did see the reform program as a solution to the problem of failing students and expressed disappointment when they realized the program was not everything they had expected.As a teacher in School 2 described: "We saw this as being far more effective and far more [of a] panacea for the students that were struggling than it really was."Even some of the most enthusiastic implementers were disappointed with subsequent outcomes, which influenced how they thought about and employed the reform principles in their practice.
One educator in School 1, who indicated throughout his interview that he was excited about the reform program and liked the impact it was having upon students, admitted to differences between the students who are enrolled in the reform program and those who are not.Because School 1 was a school-within-a-school, students could opt in or out, unlike the students in School 3.This humanities teacher said, "on paper our students look the same as all the other students…[but] one difference is that students in the Transforming Schooling school, at least, somebody agreed to let them come here."Another educator in School 1 recalled a conversation with a teacher at a model reform school that led her to believe that the reform program may not be appropriate for all students.The science teacher at School 1 said that "their demographic is really tough because they have a much lower socio-economic demographic."As this teacher continued, "there are still too many absences," which disrupts group work.In this case, even though this particular teacher did not describe the Transforming Schooling program as problematic for her students, she recognized that there were some students in some schools that may have a more difficult time with the reform demands, particularly where the program was interrupted by contextual factors such as attendance.
In addition, some educators indicated that Transforming Schooling was more appropriate for students in other schools, such as those in the model school sites.For instance, a veteran science teacher at School 3 described the students and the conditions of the model school as being drastically different from those in his school, particularly because the model was a magnet school that students opted in to: "The kids understand if they don't meet the standard, they're just going to go back to their local school, and none of them wanted to go back."In this regard, the implementers were more sensitive to the effects of local context than were blanket promotional materials and discourses regarding the one-size-fits-all program.Although the "for all" discourse was useful in inspiring school leaders to join the reform organization, this discourse was less useful in describing, and providing a basis for supporting, local implementation.The differences between model schools and the schools in this study prompted educators to question whether the reform program had the tools to support all students to be successful in their specific school contexts.For example, a math teacher at School 3 suggested that, although the reform coach assigned to the school was helpful, the school "would need somebody that met our specific needs" to support a "challenging" school with unsupportive district and school leadership.
Other teachers interviewed at School 3 also expressed a gap between what the reform organization expected of them and what the school was able to accomplish, given the leadership void at the district and school levels.One teacher spoke highly of the support from Transforming Schooling, but also allowed that there were many challenges to implementation that it either could not or would not address.For example, the teacher said the administration would not change the school schedule to accommodate the collaborative project planning that Transforming Schooling required and would not prioritize training for reform components, such as technology usage.A veteran teacher at School 3 also suggested that Transforming Schooling should have been doing more to support implementation given the school context and difficult administration.He said that he wished the reform organization would "force the issue on professional development," on "teachers' time in the class," and on "one-to-one technology."Though each of these elements were listed as necessary conditions for success by Transforming Schooling, its staff did not hold schools accountable for them in School 3. Taken together, descriptions of the gap between what School 3 needed for successful implementation and what Transforming Schooling actually provided laid the foundation for an implementer discourse that indicated the reform program might work for some children, but not for all, and certainly not for all schools.Our analysis suggests that the teachers in this study did not begin implementation with this discourse.Rather, it emerged out of their experiences negotiating the reform discourse with the realities of their school's implementation.The most salient example of this negotiation emerged as a sub-discourse of "for some, but not our" students, as described in the next section.

For Some, but Not Ours
A sub-discourse of "for some" could be summarized as "for some, but not ours," particularly in School 3, where teachers described how the implementation of the Transforming Schooling program in their own school did not resemble that in the model schools.For instance, several teachers described the need to modify the reform program in order to better meet the needs of their own students.One teacher copied a project from Transforming Schooling's bank of curricular resources, but then "modified it to fit our students."Other teachers discussed the ambitiousness of the 21 st century learning that the reform organization asked them to facilitate, and compared that to what they thought were more limited possibilities with their students: "I just think with our students' culture it would have been best to have a separate program.Because we literally threw these kids in with no knowledge."Comments like this revealed the discourse that the reform program may be better for some students, rather than the case school students.These comments often indicated teachers' beliefs that the program itself was just not appropriate for the student populations served, perhaps particularly for low-income students of color who struggled to meet basic proficiency levels on standardized tests.
In particular, educators at School 3 struggled with how to manage large class sizes of students with new technology, new curricula, new instructional approaches, and what they viewed as an unsupportive leadership team.A teacher at School 3 reported, for instance, that the leadership team would not go "all in" for the reform: "[The administration says,] 'hey, got you a Smartboard!' Alright, great, but you still got 55, 60 kids in this room for a tiny Smartboard.…We got those put in for example in November, I went through my training in March."Even though the discourse of implementers was broadly similar across schools in that most teachers described Transforming Schooling as being "for some," teachers in School 3 more often described the program as not being "for" the students in their school.With a student population of predominantly low-income students of color, the implementation effect of this discourse was that the students who have been typically left behind in the U.S. school system also were left behind in implementation of this reform program.
Teachers at School 3 in particular suggested that "our students" had deficiencies that would have made any reform program unsuccessful.We identified several examples of what could be described as "deficit thinking," wherein teachers indicated they believed that their students' backgrounds would prevent them from succeeding (Valencia, 1997).For example, another teacher in School 3 said: "Our students, they lack coping ability, they lack patience, they know nothing about persistence.They're only persistent about arguing.That's the only persistence they have."This same teacher also organized his students into two groups: "students that really want to learn" and "the other half that don't want to learn."Similarly, during a Transforming Schooling training that included school representatives from all of the reform schools in the state, several teachers from different schools brainstormed about how to deal with "deadweight" students.One strategy that the group endorsed was to put all such students in the same project groups and let them "sink or swim."When the "for some" discourse was expressed in these ways in open meetings, Transforming Schooling staff did not object.This conveyed that, although Transforming Schooling would maintain the "for all" discourse for purposes of promoting the organization, it would not hold implementers to this discourse in implementation.
In another illustrative example, a Transforming Schooling coach described an interaction with a teacher in School 3 who the coach felt was not doing everything possible for students.When the coach asked what the teacher could do to support students in the reform program, the teacher replied, "Well, I can't do that.Have you seen our kids?"When the coach asked what technologybased tools the teacher had tried to use, the teacher said, "I can't use those.They're too dangerous.We have 60 kids."The implication-unchallenged by Transforming Schooling staff-in this discourse is that certain students are not worthy of the educational innovations offered by the Transforming Schooling program.
In sum, throughout the first two years of Transforming Schooling implementation in our case schools, there were repeated examples of the ways in which the organization's discourse of "for all" did not align with implementers' discourse of "for some."This manifested in two related forms.The first was a general discourse that the reform program was appropriate for some students, but not all.The second was a more specific discourse that the reform program was not appropriate for "our students," or the students in the schools under investigation here.Although not all schoolbased staff expressed this view, there were examples of this "for some, but not ours" discourse in each of the three schools.The sub-discourse emerged most strongly in School 3, which had adopted the program in its school turnaround effort.
The misalignment between the "for all" and "for some" discourses resulted in repeated conflicts among Transforming Schooling staff and school-based staff, as they worked to implement the reform program in varied settings, with vastly different student and staff populations.Though the reform organization discourse of "for all" often was characterized in aspirational sentiments for public relations purposes, the gap between those sentiments and what many teachers experienced on the ground seemed to create disappointment, resentment, and mistrust of the reform organization.This, in turn, led teachers to describe uneven implementation across school sites.
The resulting effect of the asymmetry between reformer and implementer discourses is that both groups began to consider why the idealized discourse was not realized in practice.For reformers in Transforming Schooling, this meant that teachers who did not adopt the "for all" discourse were characterized as responsible for poor implementation.For implementers, this meant that students themselves were often blamed for not reaching the expected goals of the program.Indeed, throughout our analysis, we did not find evidence that reformers or implementers understood their own roles in creating the contexts in which implementation quality was uneven.Ultimately, the two reform schools in this study from which most of the "for some" discourses came withdrew from the Transforming Schooling program.Only School 1, with a predominantly white and higher income student population, continues to implement the reform program.Nevertheless, the program has continued to promote itself online as being "for all" students.

Discussion
When comparing the language used by reformers to implementers, there were clear differences in descriptions of whom the Transforming Schooling program was intended to support.These misalignments emerged out of implementation and, in turn, informed educators' future behavior and attitudes in the reform program.As a result, misalignment in the idealized discourse of 21 st century school reformers and educators in reform schools produced foils that educators used against the reform program as they attempted to implement it.For instance, the simplified "for all" discourse promoted by the reform organization could be easily dismissed by educators in difficult school contexts who were struggling with implementation in the first two years.Considering the reform organization did not build in mechanisms for supporting changes in teachers' mindsets, the discourses they brought from their personal experiences with reform implementation ultimately undermined any potential power of the idealized discourse.Consequently, this analysis raises implications for school reform organizations, school reform adopters, and policy itself.
School reform organizations may benefit from critically examining how (and if) the discourses used in their promotional materials are supported by practical tools for implementation.The prevalence of deficit thinking in the discourses of some reform school teachers potentially indicates a need for the reform organization to provide support and development opportunities to better align with the discourse that 21 st century school reform is "for all" and that all children can succeed in it.Without these tools, the "for all" discourse gets lost in translation in challenging school environments with few examples of how to achieve success.Implementation is likely to be further muddled, if not undermined, by a lack of alignment between the discourses they promote and the discourses of schools.
School reform adopters seeking programs to support the work of school improvement may benefit from critically examining whether reform organizations are structured to execute the discourses they promote.Research on reform implementation has demonstrated the difficulty external programs have had in supporting consistent local implementation (Berman & McLaughlin, 1975, 1978;Firestone & Corbett, 1988;Mirel, 2001).Without evidence of success in similar schools, mechanisms for incorporating modifications from implementation sites, and tools to support aspirational goals, the fees for this and other school reform programs may be of uncertain value in already strapped school budgets.Like many other organizations, Transforming Schooling continues to expand its reach to academically failing schools with large populations of low-income students, yet has not been subject to a rigorous impact evaluation of its outcomes.Consequently, wellintentioned adopters seeking a high-impact, cost-effective reform program are forced to rely on rhetoric rather than evidence when making decisions.Misalignments between the needs of schools and the services that reform organizations can actually provide, therefore, may create a series of unintended, unwanted costs.These costs are not only financial.For struggling students, misdirected resources also may cost time, learning, and educational success.
As such, our analysis additionally raises two important implications for policy.First, the ways in which promoters frame and describe policy subsequently influence how policy is adopted and implemented.In this case, the idealized discourse that the program could be successful for "all students" actually undermined that goal.Implementers were forced to rationalize why they were not seeing success for all in their own schools, and part of the way some implementers did that was through discourses representing deficit views of their students.In other words, when students were not immediately successful in a program that purportedly had been designed for "all students," then some educators attributed that lack of success not to the program or classroom implementation, but to perceived academic or cultural deficiencies among students.The emergence of deficit discourses within programs designed to serve all students is particularly unfortunate.
This leads directly into the second implication for policy, which is that policy framed in terms of supporting all students should include the adequate supports to be successful for all students in all contexts.It is not that discourses of "all children" are misguided and that we should not be striving for such aims.This is a worthy and important goal.However, it is not enough to create policies or reform programs that simply repeat aspirational statements.For "all children" to learn will involve an enormous effort-one in which pedagogical resources are carefully crafted to meet the needs of local contexts and individual students.Like the ongoing debate about the difference between equality and equity, supporting all students does not mean that all students need the same supports.As such, our analysis indicates that an "all students" policy discourse limited the reform organization's willingness to express how different school contexts may need different supports, ultimately hindering reform in these and potentially other low-performing schools.

Conclusion
Local and state education actors now have an ever-expanding collection of organizations, programs, and consultants with whom to contract in order to advance methods of 21 st century schooling.The emphasis on student-centered, technology-enhanced, and career-and college-ready school reform has all but taken over the national dialogue about how to improve schools through student learning (e.g.Race to the Top; Common Core; real-world learning; hands-on instruction).But two decades of research on implementation of school improvement policy and reform suggests that the ways in which school-based staff interpret, make sense of, and then act on concepts of reform significantly influence the impact it is likely to have on educator practice and, in turn, student outcomes (Coburn, 2001;Cohen, Peurach, Glazer, Gates, & Goldin, 2014;Datnow, 2007;Datnow, Hubbard, & Mehan, 1998;Elmore, 2004;McLaughlin, 1976;Spillane, 2004).We posit there may be a need to extend the literature on policy implementation to the discourses used to promote and justify adoption and implementation of policy.
We suggest that the discourses of school reform-particularly those of the school reform organization and reform program implementers-are a fruitful area of future research on understanding education policy implementation and outcomes.The discourses we found in one 21st century school reform organization may have implications for understanding 21st century school reform writ large.The ambiguous language about "for whom" the program was actually intended may be representative of broader school reform discourses that set aside structural inequalities to promote the concept that all students can succeed if given the right school-based inputs.The lack of evidence on effectiveness, in combination with the misalignment of program discourses and supports for those discourses in practice, raises significant questions regarding the ways in which education reform models may be replicating social stratification (Schmidt, Burroughs, Zoido, & Houang, 2015) within public schools.Given that educational agencies have considerable freedom to choose among diverse reform programs, it remains important to understand the discourses through which reform organizations advertise models, implementers justify adoption, and educators respond.
policies.Rachael's current projects investigate supports for adolescent literacy, disciplinary literacy, state policies related to reading instruction and tools for teacher evaluation.

Table 1
Descriptive Data on Three Transforming Schooling High Schools (2010) Notes.Student enrollment data are provided in ranges to protect the identities of the schools.Estimates are based on teacher self-report and district data.School 3 is a School Improvement Grant site.

Table 2
Data Sources for Study of Reform Organization Organizational documents have been grouped into descriptive categories.Planning documents include meeting and conference agendas.Performance indicators include rubrics, priority lists, and School Improvement Grant guidance.Operational documents include descriptions of program roles, master schedules, teacher certification applications, internal teacher interview protocols, and business industry contacts. Notes.

Table 3
For Whom?The Discourses of "All" and "Some" Students