The Covert Mechanisms of Education Policy Discourse : Unmasking Policy Insiders ’ Discourses and Discursive Strategies in Upholding or Challenging Racism and Classism in Education

Policy insiders across party lines increasingly acknowledge educational “gaps,” yet they talk about this inequity in very different ways. Though some critique disparities through a structural lens, others use deficit discourse, blaming famil ies of color and working-class families for educational outcomes. This study examines how state policy insiders explain educational inequity, shedding light on the complex relationship between language and the maintenance of systemic racism and classism in education. Drawing upon Education Policy Analysis Archives Vol. 23 No. 93 2 a unique data set of interviews with 50 policy insiders in one state in the United States, we found three main discourses used to explain inequity in education, each of which cited a different cause: 1) structural inequity, 2) perceived deficits of families and communities, and 3) teachers unions and teacher seniority. Policy insiders used often-veiled discursive strategies to advance their discourses. For instance, those that used deficit discourse: 1) asserted that those most negatively impacted by inequity cause inequity; 2) strengthened deficit discourse by blending it with one or both of the other two discourses; and 3) made inequity appear natural through the use of several substrategies, including obscuring the identity of those harmed by inequity. These strategies allowed some policy insiders to strengthen deficit discourse, divert attention from structural issues, and characterize themselves positively while advancing racist and classist ideas. These findings have compelling implications in terms of possibilities for policy changes supportive of educational equity.


Introduction
Educational resources in the United States are not divvied up equitably and, instead, follow lines of race and social class.For instance, students of color and working-class students have less access to rigorous curriculum and school counselors on average and are more likely to face school overcrowding than their White and middle-class counterparts (Darling-Hammond, 2010;Freelon, Bertrand, & Rogers, 2012).In addition, inequity can be found in other aspects of the schooling experience, such as school discipline (Losen, Hodson, Keith II, Morrison, & Belway, 2015) and expectations for students (Jussim & Harber, 2005).Research suggests that these disparities play a significant role in shaping differential school outcomes (Darling-Hammond, 2010;Jussim & Harber, 2005;Losen et al., 2015).Regardless of schools', districts', and policy insiders' intentions, these disparities-the manifestations of systemic racism and classism in education-place students of color and working-class students at a disadvantage while benefiting White and middle-or upperclass students.
Policy insiders-those who have sway over policy, including elected officials, staffers, lobbyists, business and civic leaders, and policy experts-can choose to acknowledge the systemic inequity in education or overlook it.Those who acknowledge it may question the role of institutions in perpetuating inequity.Those who dismiss it may instead favor deficit views that characterize communities of color and working-class communities as responsible for inequitable educational outcomes.According to such views, families of color and working-class families devalue education, lack parenting skills, and fail to provide a stable home environment (Lawson, 2003;Pollack, 2012;Valencia & Black, 2002).Despite much evidence to the contrary (Lopez, 2001;Reynolds, 2014;Terriquez, 2011;Valencia & Black, 2002;Villenas, 2001), this narrative is commonly found in many facets of education (Howard, 2013;Jimenez, 2012;Matias & Liou, 2015;Oakes & Rogers, 2006;J. Rogers & Terriquez, 2009;Solórzano & Yosso, 2001;Souto-Manning & M., 2006).
How policy insiders explain inequity in education-both inputs and outcomes-is important because of the potential policy consequences.As Critical Discourse Analysis (CDA) theory and research indicate, discourse-language use around a particular socially and historically situated topic-helps to constitute society, either maintaining and warranting or transforming it (Fairclough, 2011;Reisigl & Wodak, 2009).As an example, deficit discourse about communities of color and working-class communities has been used to justify school segregation and rote, back-to-the-basics curriculum that narrows learning opportunities for students of color and working-class students (Gutiérrez, Morales, & Martinez, 2009;Valencia, 2010).In addition, deficit discourse can support arguments against public investment in tackling school inequality, as research on the media and public opinion has found (Chart & Kendall-Taylor, 2008;O'Neil, Haydon, & Remington-Bell, 2013).
The role and function of policy discourse may appear straightforward, but actually entails mechanisms that are often obscured.Specifically, discourses are accompanied by discursive strategies, or planned discursive activities toward a goal, varying in the degree of automation and intention (Reisigl & Wodak, 2009).Research has shown that discursive strategies can be tools to either maintain or challenge hierarchical relationships (Fozdar, 2008).In terms of maintaining the status quo, Whites, dominant groups, and individuals in positions of power may use a range of discursive strategies (van Dijk, 1993(van Dijk, , 1997;;Wodak & Reisigl, 1999;Wodak & van Dijk, 2000), such as characterizing immigrants and people of color as problems, while simultaneously presenting themselves as non-racists (Bonilla-Silva & Forman, 2000;van Dijk, 1997).
Though these powerful discursive strategies are commonplace, they are often subtle or covert (Reisigl & Wodak, 2001), pointing to the need for research that elucidates the relationship between strategies and discourses and their function in upholding or challenging educational inequity.However, few studies have examined the discourses and discursive strategies of United States policy insiders in their talk about inequity in education.Our research sheds light on this area, drawing upon a unique data set of interviews with 50 policy insiders in one state in the U.S.These anonymous interviews provided a space in which our participants could verbalize their views on K-12 education without the fear of being quoted in the media or suffering from political backlash.Our analysis unmasks the ways that some policy insiders used discourses and discursive strategies to explain inequity in ways that had important policy implications.

Literature Review
In this section we begin by contextualizing our study with research about the structural inequity that exists in education.Next we discuss literature on the discourses and strategies used to challenge inequity, before turning to a discussion of those that uphold the status quo.

Structural Inequity in Education
As discussed above, research has documented school inequity along race and social class lines in the United States, indicating that resources are often distributed accordingly.For instance, predominantly White and middle-class schools are less likely to employ unqualified and inexperienced teachers than schools serving students of color 1 or working-class students (Adamson & Darling-Hammond, 2011;Darling-Hammond, 2004a, 2004b;Fanelli, Bertrand, Rogers, Medina, & Freelon, 2010).Since schools serving working-class students or students of color are more likely to have novice teachers, they are disproportionately affected by teacher turnover, which can disrupt the continuity of instruction and stymie teacher development efforts (Loeb, Darling-Hammond, & Luczak, 2005).In addition, schools serving majorities of students of color or working-class students often fail to provide enough college preparatory classes (Darling-Hammond, 2004a, 2010;Freelon et al., 2012), while racially mixed schools often track students, with White students more likely to be tracked into advanced courses (Darling-Hammond, 2004a;Oakes, 1985).Finally, these resource 1 Some subgroups of students of color are less likely to experience these conditions.inequalities are often compounded by overcrowding (Adamson & Darling-Hammond, 2012;Fanelli et al., 2010;Freelon et al., 2012), low expectations of educators toward students of color and other marginalized student groups (Jussim & Harber, 2005;McKown & Weinstein, 2008;Oates, 2003), and inequity in discipline practices (Losen et al., 2015).These unequal learning opportunities have been linked to unequal academic outcomes, such as standardized test scores (Adelman, 2006;Darling-Hammond, 2004a;Ronfeldt, Loeb, & Wyckoff, 2013;Welsh, Coghlan, Fuller, & Dauter, 2012).
In sum, school inequity has been well documented through research, with a range of studies examining the inequitable distribution of specific educational resources.It is this inequity in structural inputs that our participants either acknowledged or overlooked.

Discourses and Discursive Strategies Used to Challenge Inequity
A body of research across disciplines has explored the discourses and discursive strategies used to challenge inequity in society (Fozdar, 2008).Though researchers do not necessarily use the same terms for similar phenomena, we were able to discern two general types of such discourses.One of these, which could be termed "liberal discourse," is grounded in the tradition of liberal humanism and advances universalizing ideals such as fairness, empathy, equality, freedom, and human rights (Lakoff, 2002;Verkuytena, de Jongb, & Masson, 1994).Liberal discourse presumes that inequity is caused by a lack of understanding of or commitment to these general principles.Another type of discourse-or, more accurately, group of discourses-is based upon a critical viewpoint which asserts that inequity results from oppressive power relationships that create structural and systemic barriers and belief systems that justify the underlying inequality.While scholarship has not identified a unitary critical discourse under one name, one scholar argues that a form of what she terms "social justice discourse" shifts the gaze from "them" to "us," thereby accounting for the privilege of those in power and the structures that maintain social hierarchies (Choules, 2007).Additional terms include "equity discourse" (Evans, 2013;Liu & Milman, 2014;Rusch, 2004), "critical discourse" (Schechter, 2007), "anti-racist discourse," and "resistance discourses" (Fozdar, 2008).Much research on critical discourses originates from disciplines beyond traditional discourse studies, examining, for instance, the language of social movements (Snow, 2004) and the resistance narratives-or counterstories-of people of color (Solórzano & Yosso, 2002).
Unlike the research on discourses, the research on discursive strategies mainly originates from the discipline of discourse studies.Some scholarship from this tradition has indicated that the discursive strategies that are used to contest inequity include many of those used to justify it (Wetherell & Potter, 1992).For instance, Fozdar's (2008) analysis of interviews with Maori and Pakeha (White New Zealanders) showed that speakers who convey anti-racist discourse, like those who express racist discourse, provide examples to support their arguments, present themselves as experts, use disclaimers, among other strategies.Other research has pointed to strategies that are unique to discourse(s) challenging inequity (Augoustinos, Hastie, & Wright, 2011;Fozdar, 2008;Hastie & Rimmington, 2014).A study of comments about a U.S. Supreme Court case revealed several strategies used to counter arguments that racial equity has been achieved, including connecting past to current inequality and defining group differences in terms of disadvantage and privilege (Hastie & Rimmington, 2014, p. 201).

Discourses and Discursive Strategies Used to Maintain or Justify Inequity
There is a much larger body of research about discourses and discursive strategies used to maintain or justify inequity, and we start this subsection by considering the discourses.Counterintuitively, one of the discourses that can be used to maintain or justify inequity is liberal discourse, discussed above as a potential tool to challenge inequity.In the context of the maintenance of racism, Bonilla-Silva (2014) refers to this as the "frame" of "abstract liberalism,"2 in which Whites draw upon notions of individual choice, equal opportunity, economic liberalism, and meritocracy to support systemic racism or draw attention away from it.A range of scholars have noted similar uses of liberal discourse (Augoustinos & Every, 2007;Augoustinos, Tuffin, & Every, 2005;Bonilla-Silva & Forman, 2000;Delmont, 2012;Wetherell & Potter, 1992).As an example, some who oppose affirmative action draw upon the rhetoric of civil rights and principles of universal fairness and equal treatment (Bonilla-Silva, 2014).
Another discourse, which has received much scholarly attention, is deficit discourse.This is the verbalized or textual form of deficit thinking, which Valencia (1997aValencia ( , 2010) ) defines as that which cites perceived internal deficits of working-class students and students of color and their families as the causes of school failure, while not acknowledging structural racism and economic inequity.This common discourse can be found at a variety of levels-from individuals to systems, and from classrooms, to school administration, to research, to national policy (García & Guerra, 2004;Howard, 2013;Jimenez, 2012;Matias & Liou, 2015;Oakes & Rogers, 2006;J. Rogers & Terriquez, 2009;Solórzano & Yosso, 2001;Souto-Manning & M., 2006;Trent, Artiles, & Englert, 1998).From the time of colonization and slavery through the middle of the 20 th century, deficit discourse explicitly cited supposed intellectual differences (Menchaca, 1997;Valencia, 1997a).By the 1960s, it had mostly shifted to a focus on supposed cultural deprivation (Flores, 2005;Foley, 1997;Valencia, 1997b;Valencia & Solórzano, 1997), when it was promoted by a well-known anthropologist (O.Lewis, 1959) and a much-cited government report (Moynihan, 1965).Bonilla-Silva, who refers to this discourse as the "frame" of "cultural racism," explains that it "presents their [people of color's] presumed cultural practices as fixed features…and uses that as the rationale for justifying racial inequality" (2014).A newer version of deficit discourse frames working-class children as inherently "at-risk."Rather than referring to particular challenges created by structural conditions, the at-risk label constructs these children and their families as generally lacking (Swadener, 2010).Because the label locates the source of educational failure within individuals, it follows the broader logic of deficit discourse.
Deficit discourse has been linked to education policies.Historically in the United States, Whites' belief in their intellectual supremacy provided the undergirding for de jure school segregation (Valencia, 1997a).Later, education policies sought to compensate for students' supposed cultural deficits.For instance, Congressional debates preceding the passage of the Elementary and Secondary Education Act (ESEA) in 1965 drew heavily on cultural deficit arguments, especially about Black students and working-class students, overlooking unequal schooling conditions (Stein, 2004).Currently, deficit discourse is subtly entangled in accountability and testing policies, which often ignore systemic inequalities, yet expect an equalization of educational outcomes (Darling-Hammond, 2007;Leonardo, 2007;Love, 2004).In the area of language policy, laws in several U.S. states frame English language learners as exclusively deficient in English, rather than as students who speak another language (Viesca, 2013), prompting the promulgation of narrow, simplified curriculum (Gutiérrez et al., 2009). 3n conjunction with deficit discourse, discursive strategies can help to uphold inequity.One overarching strategy is "positive self-presentation and negative other-presentation" (Augoustinos & Every, 2007;Augoustinos et al., 2005;de Cillia, Reisigl, & Wodak, 1999;Every & Augoustinos, 2007;Reisigl & Wodak, 2009;van Dijk, 1992van Dijk, , 1997van Dijk, , 2002;;Wodak, 2009), which allows dominant groups to reproduce inequitable power relations while simultaneously appearing to be fair and reasonable.Serving this strategy are many others.A very common strategy is to obscure the identity of those who are advantaged or disadvantaged in inequitable power relations (Reisigl & Wodak, 2001;Wodak, de Cillia, Reisigl, & Liebhart, 2009).For instance, Whites may use a range of discursive strategies to avoid directly mentioning race or racial groups, while simultaneously advancing racist ideas (Augoustinos & Every, 2007;Bertrand, 2010;Bonilla-Silva, 2002;Bonilla-Silva & Forman, 2000).Also, dominant groups may use metaphors to negatively characterize certain groups (Santa Ana, 1999).Working in tandem with these strategies is "naturalization," a strategy (or "frame") that "allows whites to explain away racial phenomena by suggesting they are natural occurrences" (Bonilla-Silva, 2014, p. 76). Van Dijk (1997, 2002) has found similar strategies, such as denying racism.
The studies described here have shed light on the ways that dominant groups discursively perpetuate or challenge systemic inequity.The insights of this literature could be fruitfully applied to the language of U.S. education policy insiders, perhaps revealing how dynamic discourses and discursive strategies may either uphold or challenge the racism and classism in education.An analysis of unscripted talk is especially conducive to illuminating this area in that the discursive strategies therein may be more automatic than those present in written documents or prepared speeches.However, at present, there has been little research of this nature.Our research begins to address this area.

Conceptual Framework
To better understand our participants' views on school inequity, we draw upon the theoretical component of CDA, which encompasses several approaches (R. Rogers, 2011).Our approach is informed by that of Wodak and her colleagues (Reisigl & Wodak, 2001, 2009;Wodak, 2009;Wodak et al., 2009) and Fairclough (1992aFairclough ( , 1995Fairclough ( , 2011)).The two scholars define "discourse" somewhat differently.Reisigl and Wodak (2009) define it as "a cluster of context-dependent semiotic practices that are situated within specific fields of social action," which is related to a macro-topic and multiple points of view (p.89).Similarly, Fairclough (2009) defines discourses as "semiotic ways of construing aspects of the world (physical, social or mental) which can generally be identified with different positions or perspectives of different groups of social actors" (p.164, emphasis in the original).Drawing from these definitions, we view discourse as language use around a particular socially and historically situated topic, which, as Fairclough (2009) would argue, corresponds to a particular position and perspective.
Discourses are deeply implicated in social structures, serving constitutive and legitimizing functions.Likewise, social structures also exert pressure on discourses, pointing to the dialectical relationships between micro-level language practices and macro-level forces (de Cillia et al., 1999;Fairclough, 1995;Wodak et al., 2009).Discourses can either legitimate and reproduce inequity or contribute to transforming it, which is why they are "worth struggling over" (Fairclough, Mulderrig, & Wodak, 2011, p. 370).However, the power of discourses to create change should not be overstated.People may use discourses to create a new vision of reality, but the potential of this vision to change society depends upon a range of factors (Fairclough, 2011).This means that the constitutive nature of discourses does not equate to a "regular cause-effect pattern" (Fairclough, 2011, p. 122).However, under certain conditions, discourses can exert influence upon 1) people's beliefs and knowledge, 2) identities, 3) social relations and the material world, and 4) representations of the world (Fairclough, 2011;Fairclough et al., 2011).
Discourses do not necessarily occur in isolation and may instead be blended together (Fairclough, 1992a).This is "interdiscursivity," in which one discourse is present within another (Fairclough, 1992b;C. Lewis & Ketter, 2011).For instance, "a discourse on climate change frequently refers to topics or subtopics of other discourses, such as finances or health" (Reisigl & Wodak, 2009, p. 90).Interdiscursivity can promote the rearticulation of what would have otherwise been a stable discourses (C.Lewis & Ketter, 2011).In this way, interdiscursivity can be associated with social change, but not necessarily the type that challenges inequity.Indeed, interdiscursivity can serve to obscure power relations and bolster the status quo (Nolan, 2014).An important aspect of a discourse is that it may be associated with certain discursive strategies.A strategy is: …a more or less intentional plan of practices (including discursive practices) adopted to achieve a particular social, political, psychological or linguistic goal.Discursive strategies are located at the different levels of linguistic organization and complexity.(Reisigl & Wodak, 2009, p. 94) As this definition suggests, discursive strategies may not necessarily be instrumental in nature or planned precisely, and, instead, may be used automatically (Wodak et al., 2009).The degree to which a strategy may be intentional depends on the genre of language use (Wodak et al., 2009, p. 32).Confidential interviews, such as those described in this article, may have a lower degree of intentionality on the part of the interviewee.
An understanding of the relationship between discourse and discursive strategies, on one hand, and the inequitable status quo, on the other, is especially relevant in light of the participants of our study.Researchers have shown that elites play a special role in the discursive reproduction or transformation of social hierarchies (van Dijk, 1993(van Dijk, , 1997;;Wodak & Reisigl, 1999;Wodak & van Dijk, 2000).These individuals-including policymakers, media representatives and representatives of corporations, among others-are influenced by range of discourses, some of which justify or maintain racism and classism.Political elites, such as the policy insiders in this study, may be swayed by public opinion, but their access to it is limited and usually indirect; instead, "politicians talk mostly to other elites, and what they read is written by elites" (p.34).In this way, political elites' discourse develops and persists with little exposure to non-elites.In addition, political elites help define social phenomena and form public opinion because of "their preferential access to, and control over, various forms of public discourse" (van Dijk, 1997, p. 34).In other words, they have the power to promulgate or challenge discourses.The outsized role of these individuals in wielding discourses and discursive strategies merits special attention, and our study explores this area.

Methods
Our study draws upon our unique data set: interviews with 50 individuals with influence in education policy.In alignment with our desire to shed light on policy insiders' discursive strategies, our research questions are as follows: 1. What discourses and discursive strategies do policy insiders use to challenge the inequitable status quo in education?2. What discourses and discursive strategies do policy insiders use to maintain or justify the inequitable status quo in education?These research questions reflect the advances in the field of Critical Discourse Analysis, which has demonstrated through empirical and theoretical scholarship (discussed above) that language helps to maintain and/or transform inequity in society.Hence, our research questions are based in this understanding.
As explained below, our participants were 50 individuals who were influential in education policy within a Democratic-leaning state in the United States.We employed a snowball sampling methodology in recruiting our participants.Researchers differentiate between two main types of snowball methodology (Goodman, 2011;Handcock & Gile, 2011;Heckathorn, 2011): one that is used to study the structure of social networks in populations that are not hard to reach and another one used in hard-to-reach populations.We used the latter form, which we will henceforth refer to simply as "snowball sampling."Our target population was hard-to-reach because we did not preemptively define its membership, meaning that it fit the definition of a "hidden population," for which (the latter form of) snowball methodology was devised (Heckathorn, 2011).We were interested in reaching individuals who were perceived by legislative staffers and each other as having influence in education policy within a given state.Hence, we could not create an a priori sample list.
Snowball sampling begins with a small convenience sample of individuals that serve as "seeds," who then recommend participants for the first wave.The first-wave participants, in turn, recommend participants for the second wave, and the sample grows from there (Goodman, 2011;Heckathorn, 2011).In our study, this methodology unfolded as follows: We began by identifying and recruiting 11 key civic leaders and education staffers of both Republican and Democratic legislators in the state (our convenience sample, or "seeds").We identified these individuals through one author's knowledge of the state's legislature and also through publicly available information on the Internet.At the close of each of these interviews, we asked the staffers or civic leaders to name individuals within the state who influenced their thinking on education.Using Excel, we tracked not only the names and contact information of these staffers and civic leaders but also documented the individuals they recommended.Any individual named more than once by the staffers and civic leaders became a potential first-wave participant of our study.We then reached out to these potential firstwave participants, interviewing those who agreed.As with the seeds, we kept track of the individuals who were recommended by the first-wave participants.As before, any individual named more than once became a potential participant of the study, regardless of whether she or he was named by seeds, first-wave participants, or both.The snowball sampling methodology continued from there, with the waves becoming less and less defined: When any given individual was recommended by more than one study participant, she or he automatically became a potential participant and was contacted to be part of the study.This process continued until we had a sample size of 50.
Whether with the initial convenience sample or the subsequent waves, we asked for recommendations of individuals who influenced a participant's thinking about education according to five categories, listed here as worded to participants in interviews: 1) legislators or state policymakers; 2) labor leaders; 3) business or industry leaders; 4) civic leaders and education equity advocates; and 5) education policy experts.By casting this broad net, we were acknowledging that influence in education policy originates from a range of sectors in society, not just the legislature.This is one reason that we use the term "policy insiders," which encompasses this multi-sited influence.Another reason is because these individuals defined their own membership into a population-those with influence over education policy in the state-which was co-constructed not only in the creation of the sample but also in countless phenomena predating the study that would lead a participant to consider an individual as influential.
The participants held a range of positions and differed by political party, gender, and race/ethnicity.Five of the 50 participants were state legislators and 12 were legislative staff or held policy positions within the state.Of the 17 individuals associated with the state legislation, nine were Democratic legislators or staffers of Democratic legislators.Eight of them were Republican legislators or staffers of Republican legislators.There were eight participants who represented business or industry associations and seven who were labor leaders.Six were civic leaders, including civil rights lawyers who worked for organizations that sought to promote educational equity.The additional participants worked for think tanks or served as lobbyists, including lobbyists representing organizations focused on career and technical education.One additional participant was the superintendent of a K-12 public school district.Of these 50 participants, 34 were men and 16 were women.We did not ask participants to identify themselves in terms of race or ethnicity.Instead, after the completion of the interviews, we searched the Internet for images of each participant and considered the presence or absence of a Spanish surname in categorizing participants by race/ethnicity.We acknowledge the flawed nature of this approach in that participants' racial/ethnic identity may not match our perceptions of their phenotypes and inferences based on surname.Nevertheless, we report the racial/ethnic categories of participants to give an impression of the racially skewed nature of the resulting sample.Our categorizations indicated that there were three African Americans, three Asian Americans, six Latina/os, 34 Whites, and four others for whom images were unavailable.When the race data was crossed with the gender data, we found that 27 participants were White males.
Our data collection period extended from April 2010 to March 2011, reflecting our difficulty in contacting and scheduling interviews with high-ranking, busy individuals.We conducted one interview each with the participants, generally over the phone.The interviews followed the same interview protocol, were audio recorded, and usually lasted between 30 and 40 minutes.Our interview protocol was designed to garner participants' views on the purpose of education and was not originally intended to study discourse or discursive strategies.We began each interview by asking participants to share a personal memory about a learning experience.From there, we asked participants about their views on the knowledge and skills high school graduates should master and whether the schools in the state provided equal learning opportunities.In addition, we asked whether schools should 3) prepare all students for college; 4) provide civic education; and 5) teach "21st century skills."Finally, we also posed a set of scenarios of four fictional students with different academic and social backgrounds and asked participants to explain what knowledge and skills each student needed for the future.For the examination discussed in this study, we focus on answers to one interview question in particular, which was worded as follows: "Do you think that all…[state] schools-no matter the neighborhood-provide the same opportunities for students to develop the knowledge, skills and understanding you believe are important?If no, what differences exist?Why do you think these differences exist?"We chose to analyze the responses to this question in particular because they revealed the ways in which policy insiders used discursive strategies to maintain, justify, or challenge the inequitable status quo in education.The participants' responses to this question made clear that they did not all have the same conception of the nature of the differences that exist.Though the question asked participants about differences in schools, and some discussed these, others discussed differences in student characteristics or student outcomes, for instance.
It is important to note that our participants knew that the interviews were confidential.In addition, the interviews included unorthodox questions, including the scenario questions and the one asking participants to describe a personal memory about a learning experience.In this way, the interviews in the research project differed from media interviews, to which many of the participants were accustomed.It is possible to that our interviews encouraged greater candor with more possibilities for straying from a partisan "script" than media interviews, considering the promise of confidentiality and the interview protocol.Or it is possible that, regardless of the unique nature of the interviews, participants nevertheless drew upon the talking points they often used to address the media and the public.We are unable to ascertain this information for any individual participant.What we would like to note, however, is that we do not view the participants' talk as a static and exact reflection of their consciousness or thinking.Instead, we understand their talk to be constructed in the moment within the context the interviews.
We first coded the transcribed interviews within Atlas.ti qualitative data analysis software, beginning with a network of a priori "structural" codes (Saldaña, 2009) designed to address the original focus of the study-policy insiders' views of the purpose of education-as reflected in the interview protocol.From there, we began to code nuances among participant responses by using "initial codes," which are inductive (Saldaña, 2009).At this point, we began to notice participants' discourses and discursive strategies in challenging or justifying inequity in education.These arose with only the participants who acknowledged that inequity existed in the first place, in either inputs or outcomes of education, through a "yes" response to our question about whether all state schools "provide the same opportunities for students."Of our 50 participants, 45 responded in the affirmative to this question. 4fter experimenting with different CDA approaches to examine these discourses and discursive strategies, we settled upon that of Wodak and her colleagues (Reisigl & Wodak, 2001, 2009;Wodak et al., 2009), which they refer to as the Discourse-Historical Approach (DHA).This approach involves considering the context of language and how discourses change over time, and specifically analyzes 1) the contents of discourses, 2) discursive strategies, and 3) the talk, text, and images that are vehicles for the first two categories.Our study does not allow us to examine changes in discourses over time, but we consider the context of the policy insiders' talk-specifically the race and class inequities in education.In addition, we focus on all three of the areas of analysis.Especially helpful for this study is DHA's five categories of discursive strategies: 1) nomination (how people and things are referred to); 2) predication (how people and things are characterized); 3) argumentation (the arguments used); 4) perspectivation (the perspective from which the first three strategies are expressed); and 5) intensification or mitigation (how the ideas are discursively intensified or mitigated).
We applied the five-category typology to the responses of the 45 participants, relying on the lists of sub-strategies published by Reisigl andWodak (2001, 2009), some of which are explained in the findings.Specifically, we analyzed each participant's response twice using this lens: once to get a sense of the discursive strategies at work and a second time to code instances of strategies that had become apparent in the initial analysis.We also considered the ideas presented by Bonilla-Silva (2014) (as described in the literature review), but through the lens of DHA.For instance, we considered his frame of naturalization to be a class of discursive strategies, or a meta-strategy.We kept track of information gained in this process in an Excel spreadsheet, with interviews as rows and strategies as columns.
We established trustworthiness and credibility through four main avenues: 1) triangulation of data sources within the same method; 2) testing rival explanations; 3) peer debriefing; and 4) negative case analysis (Creswell, 1998;Lincoln & Guba, 1985;Patton, 1999).There are two main limitations to our methodology.One of them, discussed above, is that we did not ask participants to identify themselves by race or ethnicity.The second is that we did not ask non-legislature-associated participants to provide their political affiliation.This information would have allowed us to further compare and contrast the responses along party lines.

Findings
Our analysis indicated that policy insiders' discursive strategies served to subtly advance their discourses.Some strategies provided structural inequity discourse with supporting evidence and rhetorical force.Other strategies allowed deficit discourse to appear innocuous, obscuring the way it entrenched inequitable power relations.Even though we found clusters of strategies that were predominantly associated with a certain discourse, we did not find a one-to-one link between any given set of discursive strategies and a discourse.
Overall, we found three main discourses, in which participants placed blame for inequity on 1) the social structure, 2) deficits of families or communities, or 3) teachers unions and teacher seniority.The first discourse challenged the status quo, and the other two helped to maintain it.The strategies most often associated with structural inequity discourse included: 1) asserting that the social structure causes inequity; 2) appealing for equity or equality; 3) listing disparities in schoolbased resources; and 4) contrasting the availability of resources between schools.The strategies most often associated with deficit discourse included: 1) asserting that those most negatively impacted by inequity cause inequity; 2) strengthening deficit discourse through interdiscursivity; and 3) naturalization, which involved three substrategies.The teachers union discourse was a special case, serving as a quasi-structural discourse that was usually employed in service of giving deficit discourse added muscle, through interdiscursivity.We do not analyze strategies that were unique to the discourse because there was only one instance of it occurring without deficit discourse.
Below we discuss the main discourses and strategies that participants used to challenge or maintain inequity.This is not meant to be an exhaustive account of all the discourses and strategies that the policy insiders used; instead we focus on those that were most common.

Discourses
Following DHA's recommendation to examine the contents of discourses (Reisigl & Wodak, 2009), we begin with a discussion of the characteristics of the three discourses mentioned above, which we call structural inequity discourse, deficit discourse, and teachers union discourse.In addition to these discourses, we identified others that were less common.For instance, a few policy insiders blamed inequity on technical, managerial, or administrative problems and others cited low adult expectations.However, those discourses were rare, so they are not included in the discussion here.

Structural inequity discourse.
The structural inequity discourse, as its name suggests, focused on the role of the inequitable social structure, often considering school resources and systemic factors in influencing achievement.This discourse was similar to that voiced in some segments of academia (e.g.Darling-Hammond, 2004a), and to what scholars have called "social justice discourse" and "equity discourse" (Choules, 2007;Evans, 2013;Liu & Milman, 2014).Overall, 23 of 45 participants voiced this discourse in explaining differences in opportunities between schools.Due to the way that we framed our question on inequality, this high representation of the structural inequity discourse is not surprising.We asked whether schools provide the same opportunities for students, regardless of the neighborhood.The causal agent in this question was the schools, and the focus was on what some scholars have termed the "opportunity gap" (Carter & Welner, 2013), which could have influenced policy insiders to respond in terms of school-based issues.5Also, of the 23 who used structural inequity discourse, only 16 of them did so without mixing it with deficit discourse.As discussed in a later section below, mixing structural inequity discourse with deficit discourse served to undermine the former and bolster the latter.
The structural inequity discourse often entailed explicit or implicit references to economic inequality and sometimes race or racism, in conjunction with disparities in access to educational resources.The commentary of a civil rights lawyer exemplified this type of discourse.He said: The better-off neighborhoods tend to have better-resourced schools and more well-rounded college and career curriculum.The teachers who are fully prepared in their subject matters, including through AP courses, more AP courses and higher level courses that are available, the support for students that are struggling.This explanation in response to a question about school inequity hearkens reports on educational opportunities published by two of the authors of this article (Fanelli et al., 2010;Freelon et al., 2012), in addition to a great deal of other scholarship (Darling-Hammond, 2004a, 2010).The participant listed school inequities along several axes, including curriculum, Advanced Placement course offerings, and teacher preparation.Also, he tied these resources to economic inequality, explaining that "better-off neighborhoods" have more access to these resources, but did not cite a comparison group.
As is evident by the quotation from the civil rights lawyer, structural inequity discourse challenged the status quo of racial and economic inequity.Following the conceptual insights of CDA (Fairclough, 2011), this discourse had the potential to disrupt the reproduction of inequitable relationships and structures.However, many policy insiders weakened the discourse by mixing it with others, such as deficit discourse, to which we now turn.
Deficit discourse.Of the 45 policy insiders who agreed that school inequality exists, 21 voiced deficit discourse, through which they directly or indirectly located the cause of inequitable schools with working-class and/or racialized students, families, and communities.Only seven of these 21 policy insiders also discussed structural conditions, as described above.We found that our participants' deficit discourse mapped onto scholarly findings about deficit thinking (Valencia, 1997a).This category of discourse was by no means monolithic; instead, there was a spectrum of deficit discourse, ranging from more explicit arguments that framed students of color and workingclass students as being culturally deprived (Flores, 2005;Foley, 1997) to less explicit arguments that described populations as being "at-risk" (Swadener, 2010;Valencia, 1997a).However, deficit discourse did not include instances when policy insiders considered general economic inequality in connection with the experiences of families and communities, but did not locate the problem with families and communities themselves.
Several policy insiders relied upon cultural deficit discourse, framing students', families', and communities' cultures (defined broadly) as inhibiting student success.For example, an education consultant for state Republican legislators explained: I mean, many parents don't provide their own children the same opportunities that another parent might provide.They may not get their child to school on time or even every day.They may not emphasize that education is important.They may not have any reading material in their home, so that the child can, you know, just become a learner and a reader, whether they really know it or not; they may not read to their child.This participant voiced a popular refrain with this form of deficit discourse: the idea that some parents do not value education, here connected with actions such as providing books, emphasizing the importance of education, and ensuring punctuality and attendance.Though these are discrete parent actions, together they are common points in discourse about the cultures of working-class people and people of color (Lopez, 2001;Solórzano & Yosso, 2001).However, this participant did not explicitly mention people of color or any other group.
In contrast, the "at-risk" form of deficit discourse implied that some communities are reducible to and characterized by economic privation.Unlike examples of cultural deficit discourse, this one was subtle and did not explicitly cast cultures in a negative light.An apt example of such discourse can be seen in the commentary of the top education advisor to a high-ranking Democratic legislator.He said: I don't think we've given or paid enough attention to the degree to which a student's personal circumstances affect their ability to be successful in school, that one of the most profound relationships of social science is the relationship between poverty and academic achievement, and it's persistent…and it is something that we don't, I think, pay enough attention to….And that kind of relation can never be an excuse for crappy teachers or bad curriculum or instruction, and there's no excuse.But, to the degree that we can, as government, do something to intervene and improve whatever those correlates are to poverty that we make it [possible] for students to be successful.So there certainly is variability in the quality of public schools around the state, but I suspect that family circumstances and from non-school factors probably play as big or perhaps even bigger role than that.
In his commentary, this legislative advisor concentrated on the relationship between poverty and achievement.We certainly do not disagree that student achievement, as measured by test scores, is correlated with economic inequality.Also, we do not claim that mentioning the link between poverty and schooling outcomes is tantamount to deficit discourse.Instead, we seek to point out the nuances of the commentary above that appear to align with the "at-risk" notions.The advisor focused on "personal circumstances" and "family circumstances," even though he also alluded to structural issues with the reference to the government's role in addressing poverty.In this sense, the "circumstances" are presented as part of the students' and families' identities-who they are, rather than how they are treated by social institutions.He did not explain these circumstances, but characterized them as inhibiting rather than promoting academic success.6Interestingly, this deficit discourse, informed by "at-risk" thinking, led the advisor to the same conclusion as the Republican consultant cited above: parents and families may play as big or a bigger role in student success as school quality.
For some participants, deficit discourse provided a rationale for deemphasizing the importance of the role of the government in equalizing educational opportunity.For instance, a Republican state legislator remarked: There's different levels of emphasis on higher education based on the backgrounds of the families.You know, I don't see it as strictly economic issue for the schools….The schools [that are] are failing have a lot more money than the ones that are accelerating.So I think really what the idea is that third-party environment that affects the students as much as the school system does.Similar to the education consultant quoted above, this legislator explained that different families emphasize higher education to differing degrees.Like the education consultant, this participant left unstated any specific details about what types of background characteristics of families would correspond to a relative emphasis or de-emphasis on higher education.This "thirdparty environment" created by the family, for him, was equally responsible for student outcomes as the school system.This expression of deficit discourse allowed him to legitimize a lack of emphasis on funding for "failing" schools.The legislator's comment, then, points to the ways that deficit discourse can undermine public investment in schools.Also, the comment illustrates what CDA theorists (Fairclough, 1995;Wodak et al., 2009) refer to as the constitutive nature of discourse, which, in this case, served to help legitimate and maintain inequity.
Teachers union discourse.The third type of discourse cited teachers unions or seniority hiring rules as the cause of educational inequality.This type of discourse did not include general references to the supply of well-trained teachers across the state or the capacity of high-and lowwealth districts to provide attractive salaries or supportive conditions for teaching and learning (Horng, 2009).Instead, this discourse entailed teachers unions or seniority policies being blamed for the inequitable distribution of teachers.This is a different assertion than one that is often circulated, which argues that seniority allows ineffective teachers to keep their positions.As mentioned above, we consider this discourse to be quasi-structural, in that, like structural inequity discourse, it considered broad social structures.However, while the structural inequity discourse focused on inequitable resource allocation, the teachers union discourse cited union rules and union power as a root cause.Unlike the other two strains of discourse, the teachers union discourse was less common, with eight of 45 policy insiders voicing it.
The leader of an educational non-profit organization employed this discourse in his discussion of school inequality: And so you end up having, in many, many cases…, seasoned teachers and/or more effective teachers that have seniority, move on, and they move to other, better finance than they maybe started out in the tough neighborhood.If they decide that they don't want their car keyed in the teachers' parking lot, …they can move somewhere else.Well, because we have mandatory collective bargaining rules in state law that provides for step in salary increases based on, amongst other things, …continuing education credits, by definition older teachers take money with them when they make that preference seniority assignment….The non-profit leader's argument, then, was that seniority rules fueled teacher mobility as veteran teachers leave schools in "the tough neighborhood."This policy insider made it clear that the teachers unions were behind the seniority rules through his reference to "mandatory collective bargaining rules."Later in our article we'll explore in more depth the interdiscursivity apparent in his argument, which served to bolster deficit discourse.
Though the teachers union discourse masquerades as a structural discourse, it actually directs attention away from structural inequity, as does deficit discourse.The non-profit leader alluded to funding differences between schools by mentioning that funding follows the more senior teachers, a trend encouraged by seniority rules.This is an inherently different argument than one focusing on the inequitable distribution of resources in and beyond schools.It should be noted that the nonprofit leader, prior to the quote printed here, mentioned funding disparities between schools that were the result of "well-intended formulas that produce ridiculous results because they haven't been updated in 35 years."So the non-profit leader didn't blame all funding differences on seniority and did consider other systemic issues.However, we do not seek to make an argument about whether the non-profit leader himself directed attention away from structural inequity throughout all his commentary, but, instead, illustrate the nature of the teachers union discourse as a quasi-structural discourse.
Also, as used by our participants, this was not a full-fledged, stand-alone discourse, and, instead, served as a crutch for deficit discourse (as described below).There was only one instance in which this discourse was voiced on its own.In contrast, deficit discourse was voiced alone nine out of 21 times, and structural inequity discourse was voiced alone 16 out of 23 times.
We now turn to discussions of the discursive strategies associated with structural inequity discourse and deficit discourse.Since teachers union discourse was not a stand-alone discourse, we do not include a section on strategies associated with it.

Discursive Strategies Often Associated with Structural Inequity Discourse
In this section we describe the strategies most often associated with structural discourse.As mentioned above, we found four main strategies: 1) asserting that the social structure causes inequity; 2) appealing for equity or equality; 3) listing disparities in school-based resources; and 4) contrasting the availability of resources between schools.Only one of these strategies was present across all instances of structural inequity discourse (and one was occasionally present in conjunction with deficit discourse).As such, the presence or absence of these strategies, as a constellation, did not define structural inequity discourse.The strategies could be viewed as fitting within one or more of the five categories of discursive strategies outlined by Wodak and colleagues (Reisigl & Wodak, 2001, 2009;Wodak, 2009;Wodak et al., 2009): 1) nomination; 2) predication; 3) argumentation; 4) perspectivation; and 5) intensification or mitigation.We mention these categories below and explain other DHA concepts as needed.
1) Asserting that the social structure causes inequity.The one strategy that was present across all instances, either explicitly or implicitly, was the assertion that the social structure causes inequity in education, or, more specifically, that elements of the social structure are the reasons that not all schools offer the same opportunities.Since this strategy of argumentation mapped onto the essence of the discourse, its number of instances was the same as for the discourse: 23.As explained in the Methods section, we asked participants whether they thought all schools provided the same opportunities and why.If participants answered these questions, they, by necessity, created causal arguments.Commonly, policy insiders mentioned policies and school finance systems as the causes of educational inequity.Fewer of them, such as the director of an advocacy organization, made more complex arguments that involved chains of causation.After commenting that funding was the reason why schools in "low-income communities" lack "adequate resources and quality teaching," she explained: Well, I think one problem is that more and more people of privilege, more White and more middle-income have their kids in, outside of the public school system, so the public schools are increasingly Black and Brown and low-income.And yet they're funded by taxpayer dollars.They will have taxpayers, many of whom are not sending their kids those schools….
And then you have [a state tax policy], which has made it impossible to raise the level of revenue needed to adequately fund our schools, or not impossible but extremely difficult since property taxes are traditionally the main way of funding schools.And so there's the racism both on behalf of taxpayers and on behalf of elected officials.One interpretation of this fairly complex argument can be represented as follows (with arrows representing causation): "racism both on behalf of taxpayers and on behalf of elected officials"  lower taxation rates and middle-class Whites keeping children out of the public school system  inequitable funding  a lack of adequate resources and quality teaching in schools serving "Black and Brown and low-income" students.Another interpretation is that the phenomena following the first arrow-lower taxation rates and middle-class Whites keeping their children out of schools-amount to racism, and are not necessarily the result of it.Regardless of this point, the quotation shows that the discursive strategy of asserting that social structures cause inequity in education can entail complex chains of causation.
2) Appealing for equity or equality.Before stating the quote above, the director of the advocacy organization said, "When we say 'equitable' funding, it doesn't necessarily mean equal.We need more funding in communities or in schools where there is a higher level of challenges that need to be addressed."In this way, the director used a strategy of argumentation, appealing for equity, which was often found in structural inequity discourse.This strategy, used by eight policy makers in conjunction with structural inequity discourse, was similar to what Wodak and colleagues, drawing on argumentation theory, call the topos of justice (Reisigl & Wodak, 2001).A topos (a singular noun) is a conventionalized part of an argument that connects the argument to the conclusion (Reisigl & Wodak, 2009;Wodak et al., 2009).Topoi (plural) are also called warrants or premises of an argument.In the topos of justice, a policymaker argues that certain actions should or should not be taken because people should be treated equally (Reisigl & Wodak, 2001).In our study, this strategy allowed a policy insider to harness an idea that holds social currency, in addition to portraying him-or herself in a positive light (positive self-presentation).The argument of the director of the advocacy organization is an example of this strategy, in a roundabout way.She said that equality is undesirable when there is "a higher level of challenges" in some communities and schools, with the implication that equity should be promoted: "more funding" if they are in communities with fewer economic resources.The strategy of appealing for equity or equality was also present in a handful of responses from policy makers that mainly or only voiced deficit or teachers union discourse.We explain this phenomenon further below.
3) Listing disparities in school-based resources.Another strategy, which involved predication and was used by 10 policy insiders in association with structural inequity discourse, was listing disparities in school-based resources.For instance, an education consultant said, "Turnover rates among administrators and teachers is much higher in high-poverty, high-minority schools.Access to technology, access to learning materials is not as great in the kind of communities we're talking about."His list, then, included three concrete elements: turnover rates, technology, and learning materials.This format was similar to that of the others that used this strategy, including the civil rights lawyer quoted in the subsection on structural inequity discourse, who mentioned disparities in access to college and career curriculum, fully prepared teachers, and AP courses.A similar discursive move that some policy insiders made was to simply mention inequities in "resources," instead of or in addition to the list.(Those who mentioned "resources" without any list are not counted as examples of this strategy.)The list, even more than simply mentioning "resources," provided more weight to explicit or implicit arguments for the need addressing these disparities.

4) Contrasting the availability of resources between schools.
The final strategy, used by eight policy insiders with structural inequity discourse, entailed contrasting the availability of resources between schools.This strategy of predication involved characterizing different types of schools differently.The education consultant, discussed above, also used this strategy.After mentioning that "high-poverty, high-minority" schools have less access to technology and learning materials and face higher turnover rates, he showcased the opposite situation: "Schools in wealthier communities frequently are able to raise additional money through parent support, …[which] puts others at a disadvantage when they can't raise additional money for facilities and technology and field trips and a whole range of resources."By providing a contrasting case-"schools in wealthier communities"-he further underscored the gravity of the diminished resources at schools serving students of color and/or working-class students.
All four of these strategies allowed policy insiders to craft arguments supporting assertions that policy and/or funding should be changed to address inequity in education.Some policy makers stated this conclusion directly, such as the director of the advocacy organization, who argued in favor of increased funding for schools serving students of color and working-class students.Others left such a conclusion implicit.As CDA scholars would argue (Fairclough et al., 2011;Wodak et al., 2009), structural discourse may have transformative potential in challenging systemic racism and classism in education because of its policy implications.

Discursive Strategies Often Associated with Deficit Discourse
As discussed above, we found three strategies most often associated with deficit discourse: 1) asserting that those most negatively impacted by inequity cause inequity; 2) strengthening deficit discourse through interdiscursivity; and 3) naturalization (Bonilla-Silva, 2014), which involved three substrategies.The first strategy was associated with all 21 examples of deficit discourse, and is a defining aspect of it; however, the rest of the strategies did not appear in all the examples.Below we refer to the five categories of discursive strategies outlined by Wodak and her colleagues (Reisigl & Wodak, 2009;Wodak, 2009;Wodak et al., 2009).
1) Asserting that those most negatively impacted by inequity cause inequity.The argumentation strategy of asserting that those most negatively impacted by inequity cause inequity is the reasoning supporting deficit discourse in general (Valencia, 2010)-not just that of the policy insiders.For this reason, the strategy was associated with all 21 instances of deficit discourse.This strategy entailed not only maligning working-class students and students of color and their families but also implicitly or explicitly arguing that perceived deficits cause educational inequity.Through this claim, the strategy involved the use of several "fallacies," which are false or invalid argumentation schemes, according to DHA and argumentation theory (Reisigl & Wodak, 2001, 2009).
Since our interview protocol asked participants to explain why they thought not all schools provided the same opportunities, their answers were, by default, causal arguments.Of those who answered with deficit discourse, some made this causation explicit.This was the case with a business leader, whom we quote at length to illustrate this strategy and others: Business leader: So we have a lot of settings in the urban and more impoverished areas that do not expose kids to higher-order thinking skills and certainly not college readiness, college preparatory curriculum.And the suburbs in the more affluent communities certainly [have] college-going culture and access to…[college preparatory curriculum] and much more academically rich environment….And then there's this talk [of] excellency, great teachers, and figure out how to get it done or a certain school, some schools that we hopefully are helping to get it done.Melanie Bertrand: And so why do you think that there are differences?Business leader: Well, I mean, obviously the economics are affecting it, but way beyond that, the local neighborhood, the local family construct, the educational attainment of the parent or guardian or guardians or lack thereof, obviously affecting those questions, causing differences to occur.The environment and the neighborhood, the whole setting is affecting what happens outside the school day, that sort of norms establish for behavior, one being like maybe homework and studying, or the other being maybe not doing that and then alternative activities.The lack of positive adult mentors and relationships… But in a more affluent community…there's hopeful role models that, "If I do this, boy, I might be like her or him"-my dad, my uncle, my whatever.And in impoverished urban settings, I may see somebody ten years older than me that has a lot of toys, has a great car, and all sorts of things, and they got there through being in the gang and then selling drugs or some other sort of illicit activity, and that seems like, that's what I'm seeing, and I'm not seeing a lot of people that went to Harvard, and so that seems like logical, to sort of make that choice.This quote from the business leader illustrates the strategy of asserting that those most negatively impacted by inequity cause inequity.Overall, he started with structural discourse, describing disparities in school resources, and then appeared to argue that these disparities are actually caused by perceived deficits.One of the fallacies in his response was that of "hasty generalization" (Reisigl & Wodak, 2001), in which a generalization is based on insufficient evidence or bias.In this quote, the business leader characterizes "impoverished urban settings" in general as discouraging homework and lacking positive role models.This fallacy provided the foundation for another fallacy: the fallacy of confusing correlation with causation (called "post/cum hoc propter hoc").He argued that the "local neighborhood" and the "local family construct" cause the disparities in school-based resources, such as college preparatory curriculum.Because of his language throughout the quote, we know that the neighborhoods and families he is talking about are those who live in "impoverished urban settings."Through the prepositional phrase "but way beyond that," the business leader showed that he thought the supposed deficits of these neighborhoods and families were more of a causal factor than "economics" (of the schools or neighborhoods).The fallacy is the assumption that schools serving working-class students lack resources because of the perceived deficits of the neighborhoods and families of the students.
In addition, the strategy, as illustrated here, entailed circular reasoning, which may follow this pattern: A is true because B is true.B is true because A is true.Reisigl and Wodak (2001), explain this strategy as the use of an unproven claim as the starting point of an argument.The business leader stated that some of the reasons for inequity included the limited educational attainment of parents and the lack of role models in "impoverished" communities.A lack of positive role models, he suggested, would lead some students to make the "logical" choice to join a gang instead of attend Harvard.In other words, he argued that perceived deficits related to poverty (e.g.low formal educational attainment and a lack of role models) lead to students choosing gangs over education, which leads to low educational attainment.The end of the argument matches the beginning of the argument.
The business leader's quote exemplifies the strategy of asserting that those most negatively impacted by inequity cause inequity, a strategy found in all examples of deficit discourse either explicitly or implicitly.The strategy involved making negative generalizations (often stereotypes), which were assumed to have a causal relationship with other phenomena (such as graduation rates and test scores), in this way creating a circular argument.
2) Strengthening deficit discourse through interdiscursivity.In 12 of 21 instances of deficit discourse, policy insiders blended deficit discourse with structural inequity discourse and/or teachers union discourse.Of the 12 instances, seven involved interdiscursivity of deficit and structural inequity discourses, and seven involved interdiscursivity of deficit and teachers union discourses (with two cases involving all three discourses).There were two main functions of interdiscursivity, and both served to strengthen deficit discourse.One of these was positive selfpresentation and negative other-presentation (van Dijk, 1992(van Dijk, , 1997;;Wodak et al., 2009), which entailed positioning oneself vis-à-vis a discourse (perspectivization), characterizing oneself positively (predication), and diminishing the importance of structural causes of inequity (mitigation).By voicing structural inequity or teachers union discourse, some policy insiders aligned themselves with the ideals of fairness, what Bonilla-Silva (2014) calls abstract liberalism.Indeed, some even voiced the topos of justice, more commonly associated with structural discourse, explicitly or implicitly referring to equity and/or equality.By presenting themselves as espousing the ideals of abstract liberalism, policy insiders cast themselves in a positive light, allowing them the latitude to mitigate and dismiss structural arguments and advance deficit discourse.
For instance, the business leader, in the quote above, began by describing structural inequities between schools, implying that he was concerned about the disparities.He also mentioned the philanthropic work of the business organization of which he was a member, referring to "some schools that we hopefully are helping to get it done."These were schools that had received assistance from his organization to improve teaching and learning.In this way, he characterized himself as concerned about disparities in school resources and helping to improve schools.However, moments later, he appeared to dismiss the importance of these disparities (by saying "but way beyond that"), and instead asserted that parents and communities were "causing differences to occur." The other function of interdiscursivity, involving argumentation, in addition to other categories of strategies, heightened the supposed causal power of parent and community deficits by drawing upon teachers union discourse specifically.Recall that all but one of the eight instances of teachers union discourse co-occurred with deficit discourse, leading us to view it as a helper discourse.In blending the two discourses, policy insiders asserted that student, family, or community deficits caused disparities in access to teachers through complex interactions with teacher unions.For example, one policy insider suggested that teacher seniority was an "equity issue," allowing the best teachers to abide by "human nature" and move from schools serving students of color and working-class students to schools serving White and middle-class students.The president of a business organization (not the same person as the business leader referenced above) made a similar argument: Well, part of it [is] the function of the tenure system and seniority system.It gets schools in difficult areas, high-poverty, high-minority, high-crime areas; a teacher with seniority and tenure most likely doesn't to go there.So who winds up there?Less competent teachers.In essence, he argued that perceived human deficits found in some areas lead tenured teachers to find other schools, leaving the "less competent teachers" to fill the open positions.Also, he was explicit about what constituted an undesirable area-poverty, people of color, and crime-which he presented as interdependent characteristics.
In this way, interdiscursivity of deficit discourse with teachers union discourse allowed policy makers to seemingly locate the cause of educational inequity within a set of institutional and political dynamics, while subtly directing blame toward families.Hence, interdiscursivity bolstered the legitimation power of deficit discourse by reinforcing the argument that communities of color and working-class communities cause the very conditions that negatively affect them, such as inadequate access to quality teachers.
3) Naturalization.Another discursive strategy that some policy insiders used was "naturalization," which Bonilla-Silva (2014) describes as the practice of explaining away systemic racism as a natural occurrence.In analyzing our data, we came to expand the concept.For us, naturalization was a meta-strategy with substrategies that contributed to systemic inequity seeming normal and natural.The substrategies were: 1) obscuring the identity of those who are harmed by inequity, 2) listing several deficits together, and 3) highlighting deficits by providing contrasting cases.
The first substrategy of naturalization, observed with 17 policy makers who voiced deficit discourse, was obscuring the identity of those who inequity harms, specifically by using "metonymy," in which the name of one entity is replaced by the name of another entity "which is closely associated with it in either concrete or abstract terms" (Wodak et al., 2009, p. 43).By never directly naming individuals, this strategy7 made inequity appear to simply exist, naturally, without victims.Policy insiders used phrases such as "urban settings" to allude to the people who lived there.Interestingly, some policy insiders employed this strategy in conjunction with structural inequity discourse, (probably inadvertently) undermining its discursive potency.Often metonymy involved what one author has called place-name discursive deracialization (Bertrand, 2010), using geographic names to subtly refer to racialized groups.
These subtle references to race appeared in the deficit discourse of a teachers union leader.He responded to our question about whether all state schools provided the same opportunities with this explanation: "Structures with the demographics, you know, a school that's sitting in a White, middle-class neighborhood, you know, those kids are going to read White, middle-class language and White, middle-class textbooks when they're in school."Here he explicitly named the White racial group, but later alluded to students of color and working-class students in a much more oblique manner.After describing the "rich…language experiences" of the White, middle-class students, he painted a sharply contrasting picture: Now, let's just go across the causeway here in Sierra Linda and move into West Sierra Linda, and I can come into a neighborhood where every single family is living below the poverty line….
[T]hose kids don't have the books at home, they don't have the language accessibility, their interaction with the parents is significantly less, they have a lot more interaction with the television and with video games because those two vehicles become much more effective child monitors.Through this explanation, the union leader referenced race-specifically people of color-in subtle ways.He provided a specific geographic place name, West Sierra Linda, indicating that this area was "across the causeway" from the White, middle-class area he had first described.He may have assumed that his listener was familiar with the demographics of the area, specifically its racial composition.By obscuring the identity of the West Sierra Linda residents, he naturalized their supposed deficits and racial segregation, while building an argument based on connotative shortcuts.The place name of "West Sierra Linda" was standing in for not only its residents but also for working-class communities of color in general.
The substrategy of listing several deficits together, used by 14 policy insiders in association with deficit discourse, involved predication and implicit argumentation, in which the policy insider implied that there was an automatic and unquestionable link among perceived deficits.Also, this entailed the fallacy of a hasty generalization, based on stereotypes of communities of color and working-class communities.For instance, the business leader quoted above said and implied that "impoverished urban settings" discourage studying and lack "positive adult mentors and relationships."The education consultant for state Republican legislators, quoted in the section on deficit discourse, said that some parents do not take their child to school on time, emphasize the importance of education, or provide reading material.These sweeping generalizations were characterized as naturally and automatically inseparable.The implication was that if one deficit was present, so were stereotypical others.
The final substrategy of naturalization entailed highlighting deficits by providing contrasting cases.This strategy of predication, used by six policy insiders in connection with deficit discourse, positioned "affluent" and/or White schools, neighborhoods, or parents as a foil to people of color and working-class people, along with associated schools and neighborhoods.The business leader quoted above used this strategy.In an "affluent community," he argued, a student may want to be like "hopeful role models," such as a dad or an uncle.But in "impoverished urban settings," the role model may be "somebody ten years older" who has a lot of "toys" and "a great car" because of involvement in a gang and selling drugs.Similarly, the union leader contrasted the "rich…language experiences" of White, middle-class children with the homes of children of color or working-class children, which he claims lack books, "language accessibility," and interaction with parents.In both of these cases, and the other examples, the policy insiders used the foil of the White and/or affluent families to emphasize the supposed deficits of parents and families of color and working-class families.
Overall, when policy insiders used the strategies associated with deficit discourse-asserting that those most negatively impacted by inequity cause inequity, strengthening deficit discourse through interdiscursivity, and naturalization-they helped to bolster deficit discourse.In the process, they laid the foundation for arguments against public investment in addressing inequity and instead reinforced racism and classism.This analysis illustrates how the legitimation and social reproduction capacities of discourses discussed by CDA theorists can be enhanced through discursive strategies (Fairclough et al., 2011;Wodak et al., 2009).

Discussion and Conclusion
Overall, this article has demonstrated that the policy insiders in our study used relatively covert discursive strategies to advance discourses that either upheld or challenged the inequitable status quo in education.To help promote structural inequity discourse, policy insiders argued that structures cause inequity, made appeals for equity or equality, listed disparities in schools, and contrasted different types of schools.These strategies provided policy insiders the tools to focus attention on the social structure, align with ideals of equity and/or equality, which may hold some sway in policy arenas, and support their points through descriptions of differences in resources.To bolster deficit discourse, they asserted that those harmed by inequity are the causes of it, mixed the discourse with teachers union discourse or structural inequity discourse, or discursively normalized inequity by hiding identities, listing deficits, and contrasting groups of people.The causal argument and interdiscursivity strengthened deficit discourse by legitimizing the status quo and emphasizing the supposed causal power of the perceived deficits.Also, interdiscursivity-especially with structural inequity discourse-allowed policy insiders to characterize themselves positively while advancing racist and classist ideas.Meanwhile, naturalization helped normalize and even hide how deficit discourse functioned to maintain and justify racism and classism in education.
From the perspective of the theoretical aspects of CDA, the policy insiders' discourses and strategies appeared to have served a socially constitutive function in some cases.Structural inequity discourse and the associated discursive strategies challenged the educational status quo by shining a light on the differences in educational experiences between working-class students and/or students of color in contrast to middle-or upper-class students and/or White students.Seen through the lens of CDA, this discourse, aided by certain discursive strategies, may support transformation in education policy, depending on contextual factors (Fairclough, 2011).However, this discourse was sometimes infused into deficit discourse, which undermined the structural arguments.This suggests that the discourse and accompanying strategies were potentially transformative when used without deficit discourse.
On the other side of the coin, deficit discourse and the associated strategies appeared to buoy the inequitable status quo.In that our participants were policy insiders, they were influential in education policy, meaning that their discourses may have swayed not only their own actions but also the actions of others in the policy arena and beyond.For instance, participants used deficit discourse and discursive strategies to argue against increased education funding generally or for certain schools in particular.Such a stance on education funding supports the inequitable status quo and social reproduction, according to the theory.Another example can be found in the interdiscursivity of deficit and teachers union discourses.Participants claimed that deficits in families and communities caused the inequitable distribution of quality teachers, who choose to avoid "tough" schools.In this way, these policy insiders set the stage for specific policy solutions: dismantling of teacher seniority and teacher unions while diverting attention from structural issues and placing blame on communities of color or working-class communities.8In sum, the socially reproductive discourses and discursive strategies had specific implications in terms of concrete policy and social structures.
When examining these discourses and discursive strategies, it is important to consider who is voicing them.As described in the Methods section, we did not ask participants to identify themselves racially or ethnically in our interviews, but instead used Internet images and the presence or absence of a Spanish surname to attempt to categorize participants.Despite the flawed nature of this methodology, it is worth mentioning here what we found.Of the 21 policy insiders who voiced deficit discourse, 19 of them were White, one was Asian American, and one was racially unidentified.Also, three of them were women and 18 were men.Of the 16 who voiced structural inequity discourse only, seven were White, five were Latina/o, three were African American, and one was racially unidentified.Nine were women and seven were men.Due to the nature of the way we collected these numbers, we do not report them as reliable findings.Instead, we present these numbers to suggest that discourses were not used by everyone in the same way.Whites appear to have been more likely to use deficit discourse than policy insiders of other racial/ethnic groups.On a related note, when policy insiders voiced deficit discourse, they obscured the racial identity of those harmed by inequity (and deficit discourse).The dearth of discursive references to race, however, does not equate to a lack of racism, especially considering research highlighting taboos against being perceived as racist (for instance, Bonilla-Silva, 2014, & van Dijk, 1997).For these reasons, we assert that viewing the deficit discourse and the associated discursive strategies through a predominantly class lens would obscure the fact that they promote, or at least fail to challenge, systemic racism.
The main implication of our research is that policy insiders' discourses and discursive strategies either limit or expand possibilities for policy changes supportive of educational equity agendas.In light of this insight, it is important that efforts be taken in both political and academic arenas to galvanize the discourses and discursive strategies that promote equity and expose those that re-entrench racism and classism in education.One approach to this would be to conduct studies, such as the one presented in this article, that consider unscripted talk, which is more dynamic than written language, especially that found in policy documents.Also, our findings suggest that education policy conversations would benefit from an infusion of the voices of the families and communities maligned by socially reproductive discourses and discursive strategies.The promulgation of such discourses and strategies may be supported by the fact that policy insiders may not personally know the "other" whom they're characterizing.Families of color and working class families are, indeed, the experts on their lived experiences and education policy could be reimagined with their input.