TEACHERS AS DESERVING OR UNDESERVING: THE ROLE OF RACE, PLACE, AND POLITICAL LEAN IN THE SOCIAL CONSTRUCTION OF TEACHERS ON STRIKE By Sandra L. Waldron A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Education Policy – Doctor of Philosophy 2023 ABSTRACT In the United States, teachers have been both respected and villainized in popular culture and politics. These varying and conflicting cultural ideas and representations of teachers impact policy design, shaping the distribution of benefits and burdens imposed on teachers. The media play an important role in both crafting and reinforcing cultural images and did so during the wave of teacher strikes in 2018 and 2019. These strikes were part of the Red for Ed movement which framed teachers as undervalued and deserving of greater policy benefits, in direct contrast to recent accountability policies that described teachers as ineffective and needing more oversight. How did the media portray teachers on strike, and how were teachers able to shape how they were constructed? Drawing upon Schneider and Ingram’s (1993) theory of social constructions, I extend previous work to examine how policy targets (in this case teachers) actively participate in the social construction of their policy image. Social constructions are the “cultural characterizations or popular images,” embedded explicitly and implicitly in society broadly and in policy (Schneider & Ingram, 1993, p. 334). To do this, I use a comparative case study of seven strikes, and qualitatively analyze the text and images from 828 online news articles from eight national news sources with varying partisan political leanings. I consider how teachers are constructed by the media in each event, noting the critical role that race plays in shaping the deservingness of teachers. I also analyze how the political lean of the news source relates to the social constructions portrayed. I then consider how teachers framed themselves through handmade strike signs. I propose a new conceptual understanding regarding how targets can affiliate with other target groups to associate with their political power or deservingness. In this dissertation, I find that teachers as a whole are portrayed as an advantaged group, both deserving and having strong political power, during the Red for Ed strikes. I note that teachers were portrayed as deserving more frequently during strikes in less urban locations where unions have less political power and when images show a majority of white teachers. Left- leaning news sources are more likely to portray teachers as deserving, and one Right-leaning news source (Breitbart) portrayed teachers as undeserving more frequently than deserving. The messages from teachers’ strike signs coincided closely with the constructions conveyed via media images, highlighting one way teachers as a target group controlled the media narrative about themselves. Copyright by SANDRA L. WALDRON 2023 iv ACKNOWLEDGEMENTS I am grateful for all the support I have received throughout my academic journey. When I began my higher education studies, I remember listening to President Donna Shalala during freshman orientation at the University of Miami tell us that we would average seven careers in our lifetimes and that our undergraduate degree was intended to give us the skills to be lifelong learners capable of adapting to changing intellectual demands. At the time, I doubted the veracity of this statement, but at this stage in my life and careers, it feels both accurate and a fitting way to conceptualize my own growth and my future career contributions. Firstly, I would like to express my appreciation to Michigan State University and the College of Education for providing a rigorous doctoral education program, research training, and financial assistance. As a teacher at heart, I will be forever grateful to all my teachers for sharing their passion for their subject areas and the tools for learning. To my committee chair and advisor, Dr. Rebecca Jacobsen, I have been so fortunate to have found a mentor who encourages my growth as a scholar by actively engaging me in the research process, treating me as a colleague in our joint work, and pushing me deeper in my analytic thought processes. I can also never adequately express my gratitude to Dr. Jacobsen for also supporting me personally through my time at Michigan State. To my committee, I truly value and am thankful for the time and energy you have dedicated towards helping me develop this dissertation. To Dr. Katharine Strunk, thank you for pushing me to consider ways to enhance the rigor and policy relevance of my work and for providing me opportunities to develop my research skills. To Dr. Sarah Reckhow, thank you for helping me navigate the world of political science and public policy; your guidance and feedback v has helped me to define myself more clearly as a scholar. To Dr. Jane Lo, thank you for your willingness to put student learning at the forefront and to ensuring I have the resources I needed for this work. And to Dr. Melissa Arnold Lyon, I am beyond appreciative of your thoughtful feedback and professional mentoring. Your enthusiasm for your work is contagious. To all the faculty that have pushed me throughout my higher education career, thank you for always believing in my potential. To Dr. Chris Torres, I am grateful for your mentorship that helped build my solid foundation in qualitative methodologies and for including me as a thought partner in our collaborative work. To Dr. Abby Jones, I appreciate all your help training me on the Provalis software and helping me think through my dictionary building procedure. To Dr. Steve Permuth, thank you for seeing the potential in my work and guiding me through my first academic conference presentation and article. To Dr. Susan MacManus, I will be forever grateful for the opportunities you gave me first as a teaching assistant and then as a research associate, teaching me the culture of academia and how to think about the larger impact of my work. Without your influence and guidance, Dr. Permuth and Dr. MacManus, I likely would not have considered pursuing a Ph.D. To my undergraduate faculty, especially Dr. Jeanne Bergeron, Dr. Harry Forgan, Dr. John Fitzgerald, and Dr. Joshua Diem, thank you for inspiring me as an undergraduate teacher education major and for showing me how to think critically and creatively about my work. To Mrs. Ann Skipper and Mrs. Linda Janoff, I should have told you many years ago how fortunate I was to have learned to write and analyze in your classes. I use those skills every day as a researcher, in ways I would have never expected when sitting in high school English classes. To all my K-12 education teachers, I feel truly privileged to have learned from and with you. You have all shaped the teacher I have become. vi Throughout my time at Michigan State, I have been blessed with colleagues who have both challenged me intellectually and supported me emotionally. I was fortunate to be assigned Chris Thelen-Creps as my graduate student mentor, as she modeled how to pay-it-forward and shared the Education Policy program’s culture of community. She also helped me build a community of support that continues to share in each other’s joys and challenges– Tanner and Ellyn Delpier, Jesse Nagel, and Kelly Stec, you have all been invaluable colleagues and I appreciate all our deep policy discussions. I am grateful for my writing group, Andrea Chambers, Annie Hemphill, and Vanika Grover for motivating me to get my thoughts on paper. Jacqueline Gardner and Alexandra Stamm, thank you for the lunches, coffees, co-working sessions, and warm conversations. I’m also indebted to Alex and to Amy Cummings for the hours they committed to reliability tests of my image coding scheme; I appreciate your willingness to be thought partners and will forever buy the coffee when we meet up in the future. To Ariell Bertrand and Annie Gensterblum, I am so appreciative to have collaborators and co-authors who are excited by the research process and who have helped me become a scholarly mentor. And finally, to Danielle Sanderson Edwards, thank you for being my support, my confidante, and my much-needed push. I am so thankful to have a friend and colleague who champions me and my work. I truly value the community of graduate students I have been a part of and am thankful I will continue to have these friendships as I continue my career. Finally, I would like to thank my “village”, without whom this truly would not have been possible. To my extended family, I appreciate all the help given and offered over the years and the opportunity to be closer to you all. To my sister, Christi McGee, thank you for always being one of my biggest fans. And while not family by blood, thank you to my “village” that has helped whenever I have needed, especially Madison Brooks who made it possible for me to work vii by being the best co-parent we could have ever found. To Josh Roberson, who in addition to devoting hours to assisting with data collection, has also been a tireless supporter willing to talk politics and education anytime. To Megan Suau, my “oldest” friend who has been a sounding board, the best listening ear, and encouragement throughout this time. To my parents, Scott and Mary Frost, I cannot adequately express my gratitude. Thank you for always being willing to do whatever you could to support me throughout this degree program and for raising me to think critically and value my education. To my children, Spencer and Jasper Waldron, I so appreciate that I have been able to be both your mom and do this work. I hope that I can be an example for you of hard work, pursuing goals, and setting your priorities. I love you both exponentially. Thank you for your patience and understanding while I have tried to balance everything. And to my husband, Sam Waldron, I want you to know how much your support of this work has meant to me. You have made every effort to make sure I have the time to think, analyze, and write. I truly appreciate all the sacrifices you have made to make this dissertation a reality, and I cannot imagine doing this without you by my side as my partner. Thank you all for joining me on this journey. I am excited to share the work that you have helped make possible. viii TABLE OF CONTENTS LIST OF ABBREVIATIONS………………………………………………………………. x CHAPTER 1: INTRODUCTION…………………………………………………………... 1 CHAPTER 2: LITERATURE REVIEW…………………………………………………… 7 CHAPTER 3: DATA AND METHODS…………………………………………………… 28 CHAPTER 4: FINDINGS REGARDING OVERALL SOCIAL CONSTRUCTIONS OF TEACHERS ON STRIKE DURING THE RED FOR ED MOVEMENT…………………. 56 CHAPTER 5: FINDINGS COMPARING SOCIAL CONSTRUCTIONS ACROSS STRIKE EVENT CASES…………………………………………………………………... 76 CHAPTER 6: FINDINGS COMPARING SOCIAL CONSTRUCTIONS BY NEWS SOURCE.…………………………………………………………………………………… 90 CHAPTER 7: FINDINGS REGARDING TEACHER AGENCY IN INFLUENCING SOCIAL CONSTRUCTIONS VIA STRIKE SIGNS…………………………………....... 103 CHAPTER 8: DISCUSSION……………………………………………………………….. 115 REFERENCES…………………………………………………………………………….... 121 APPENDIX A: 2018 AND 2019 RED FOR ED TEACHER STRIKE AND PROTEST TIMELINE………………………………………………………………………………….. 137 APPENDIX B: IMAGE CODING SCHEME……………………………………………… 139 APPENDIX C: STAGES OF THE DICTIONARY BUILDING PROCESS………………. 152 APPENDIX D: WORD STAT DICTIONARIES WORD FREQUENCY ANALYSIS…… 158 ix LIST OF ABBREVIATIONS CHI 12 Chicago Teacher Strike 2012 WV 18 West Virginia Teacher Walkout 2018 KY 18 Kentucky Teacher Walkout 2018 OK 18 Oklahoma Teacher Walkout 2018 LA 19 Los Angeles Teacher Strike 2019 DEN 19 Denver Teacher Strike 2019 CHI 19 Chicago Teacher Strike 2019 WSJ The Wall Street Journal USA USA Today WP The Washington Post NYT The New York Times x CHAPTER 1: INTRODUCTION Beginning in 2018, a wave of teacher strikes, walkouts, and sickouts spread across the United States, impacting whole states’ school systems. This labor movement, commonly referred to as the “Red for Ed” teacher strikes,1 continued through 2019, and teachers’ demands went beyond traditional labor issues of teacher compensation and working conditions to include funding for increased support staff, decreased class sizes, and up-to-date supplies. Some of these strikes were successful, resulting in increased school funding, teacher pay raises, and improved working conditions (Karp & Sanchez, 2018; Loewus, 2019). Because these strikes were widespread, news media covered the strikes extensively shaping public opinion of both the strikes themselves and teachers through the information shared. National news media carried stories analyzing the way the Red for Ed strikes spread like “wildfire” (Goldstein, 2018). In addition to traditional newspaper sources (e.g., The New York Times, USA Today, etc.), online news articles proliferated from formerly television-focused news sources like Fox News and CNN, as well as newer, online format news sources such as Breitbart and Slate. Some of these non-newspaper sources are viewed as highly partisan in their coverage (AllSides, 2023). The media can be a powerful force in shaping and reflecting public opinion and setting the policy agenda (McCombs & Valenzuela, 2021). What the media decided to share about those strikes, and how they chose to portray teachers on strike impacts both public and policymaker opinions of the deservingness of the demands of teachers. Indeed, Hess & Martin (2018) note that teacher strikes were largely framed in supportive terms by national news sources. The 1 Depending on the political context of locations, including if it is legal for teachers to strike in a particular location, some of these teacher protest events are referred to as “walkouts” or “sickouts” instead of strikes. However, for clarity, I will refer to the broad collection of teacher protest events as “strikes” throughout this paper. 1 2019 EdNext Poll found public support for increased teacher pay was the highest it has been since 2008 and there was increased support for more school spending (Henderson et al., 2019). Teachers have not always been viewed so favorably. Historically, teachers have often been seen as “to blame” for the failures of the public school system, even when they are widely seen as being essential for quality education (D’Amico Pawlewicz, 2020). Recent policies towards teachers have tended to take a dim view of the profession. The New Teacher Project’s report, “The Widget Effect,” (Weisburg et al., 2009) claims that teacher evaluations are generally too similar and do not appropriately differentiate teachers by quality, enabling poor teachers to remain in the profession and ultimately harm student education opportunities. This report gained significant traction and influenced new teacher evaluation systems that increased accountability pressures (e.g., Amrein-Beardsley, 2017; DiCarlo, 2014). Federal Race to the Top competitive grants supported these changes, referencing “The Widget Effect” report specifically (US Department of Education, 2009), and the 2015 Department of Education evaluation of Race to the Top indicates participating states altered evaluation practices in response (US Department of Education, 2015). Similarly, a Gallup poll in 2014 saw widespread public support for linking teacher evaluations to student test scores (Gallup, 2020), providing evidence that there was general support of the perception that teachers were not performing as expected and needed to be held accountable for their performance to spur improvements. The positive media and public response to the Red for Ed teacher strikes seems to represent a notable shift in how teachers are being portrayed in cultural narratives in the U.S. It is important to understand how such shifts in perception can impact policy, particularly as there is some evidence that teachers’ demands were met in many of the strike resolutions (see Appendix A for a summary of major strike events from 2018-2019, including immediate policy outcomes). 2 Social construction theory is a useful framework for considering how perceptions of teachers relates to policy design. Schneider and Ingram (1993, 1997; Ingram & Schneider 2005) posit that the social constructions, or the sociological meanings associated with target groups in policy, can impact policy design and how a particular target group interacts with government. Target groups can be constructed along continua from deserving to undeserving, politically powerful to weak. Groups that are more positively and more powerfully constructed are more likely to receive policy benefits and avoid policy burdens. Constructions are not static, they can shift overtime (Ingram, Schneider, & deLeon, 2007). The images of target groups captured in media coverage can help us understand how these groups are perceived, and those perceptions in turn impact policy (Schneider & Ingram, 1993). If more positive, or deserving, constructions of teachers are prevalent during particular political events, such as strikes, this in turn may be associated with a shift in educational policies to benefit teachers. Because striking teachers have broadened the scope of issues they are advocating for to include increasing education funding and availability of classroom resources, this can also have wider implications for benefiting students. Race, gender, class and other identity characteristics and associated stereotypes are known to influence public perceptions of deservingness (Ingram & Schneider, 2005). The scope of the Red for Ed movement provides a unique opportunity to explore how race in particular influences the social construction of a particular target population in a particular period of time. The teaching profession is largely feminized, and teachers predominantly identify as white (U.S. Census Bureau, 2022), however, Black and Latinx teachers are over represented in urban schools, relative to the proportion of the total teaching workforce (Frankenberg, 2009). Considering these demographic differences, the Red for Ed strikes offer a unique opportunity to 3 explore the variation in racial composition of the teaching workforce to understand how within a target group sub-groups may be portrayed differently. Teachers from rural and suburban districts with high proportions of white teachers and teachers from urban city centers with lower proportions of white teachers went on strike at a similar time for similar reasons. Differences in how teachers are portrayed by news media could then be attributed to differences in contexts of strike locations, including teacher racial demographics, instead of the issues and rationales underlying the strikes. In addition to differences by strike context, it is also possible that teachers are portrayed differently by different news media sources. Understanding how the media presents teachers can help us better understand how teachers can be reframed as worthy of policy benefits. Media and public opinion can be mutually reinforcing, reflecting, and shaping each other. Increasingly, news sources vary in how they cover topics in ways that directly reflect the political leanings of the news source (AllSides, 2023). As news media continues to evolve in the online sphere, it is important to consider how different audiences may be exposed to different constructions leading to more polarized views of teachers based on their political ideology. If different news sources advance different images and stories about teachers, this could be related to differences in public perceptions and policymakers’ views of teachers and the policy issues they champion. This study examines these two potential explanations for differences in social constructions of teachers as a target group – strike event context and news source – are related to portrayals of teachers on strike in news media. I examine online news articles covering teacher strikes from eight national news sources across seven events, including one strike pre-dating the Red for Ed movement to understand how the movement does or does not represent a shift in how teachers are portrayed. I build on the field’s understanding of how teachers are constructed, how 4 those constructions vary by event context, and how constructions vary by news media source. In examining media coverage, it became apparent that while the media has editorial discretion over what is published, teachers were also making choices that influenced what was published and how they were portrayed, particularly in the creation and display of strike signs. As a result, I also consider how a target group like teachers can shape constructions through their actions. If teachers can be agentic in this process, this can increase our understanding of how teachers can actively participate in and influence the policy process as technical experts in the field of education. This study examines the following questions: 1) How are teachers socially constructed by news media during the 2018-2019 Red for Ed strikes? a. How do social constructions of teachers vary by strike event context? i. How do social constructions vary over time? ii. How do social constructions vary by racial demographics? b. How do social constructions of teachers vary by news source? 2) How do teachers actively influence social constructions of themselves as a target population? A Brief Overview of Teacher Strikes in the United States In the United States, teacher strikes have been relatively rare. Historically, teacher strikes are often a last-resort bargaining tactic used by teachers’ unions in the collective bargaining process – localized events rather than larger political movements (Lyon & Kraft, 2021; Lyon, 2022). Across a 44-year time period, Mother Jones (Lee & Liebelson, 2012) found a total of 839 teacher strikes, but all but 99 of those strikes occurred in one state (Pennsylvania). From 2007- 2019, Lyon (2022) finds record of 696 teacher strikes, with the vast majority of these being 5 coordinated strike events across multiple districts (78.45%), including the Red for Ed strikes. As Lyon counts strike events by district, the 2018 West Virginia teacher strike, a coordinated statewide event, would be counted as 55 strikes since 55 districts had work stoppages. Prior to the 2017-2018 school year, Lyon (2022, p. 2-3) shows that strikes were typically individual districts striking with relative infrequency, averaging less than 20 district strikes per year in a nation with over 19,000 school districts. There is evidence historically that teacher strikes are viewed negatively, with Goldstein (2014, p. 133) titling a chapter covering teacher protests and strikes during the community control era as “We both got militant” in reference to a quote from New York City teacher Deborah Meier talking about teachers and parents in conflict with each other. The Red for Ed strikes are a departure from typical strikes. First, these strikes began in states with Republican-controlled governments in contexts that are unfriendly to unions, collective bargaining, and striking. In fact, in seven states that had strikes in 2018-2019 it was illegal for teachers to strike (Sanes & Schmitt, 2014). As a result, a number of these events are called walkouts and a few, such as Kentucky, held sickouts where teachers called in sick en masse rather than formally declaring a strike. Lyon (2022) finds coordinated strike events like the Red for Ed strikes were more common in the South and West regions of the U.S., while individual district strikes like those that were most common prior to 2017 were concentrated in the Northeast, where historically unions have had a greater presence and political power. The strikes were widespread and prolific, and as previously noted there was a unified agenda that targeted both teacher compensation and broader education issues, including advocating for more education funding broadly, demanding hiring of additional support staff, and fighting for smaller class sizes. 6 CHAPTER 2: LITERATURE REVIEW To lay the groundwork to explore these research questions, it is first necessary to understand the theory of social constructions of target populations and how I intend to use this theory as a framework for analysis. I provide a brief overview of key tenants of the theory, a description of what is known about changing social constructions, and the role of race in social constructions. I then consider what the literature says about how teachers’ unions and teachers have been constructed together and separately. I justify my use of news media as a data source for studying social constructions and conclude with a summary of key points. Theoretical Framework: Theory of Social Constructions of Target Populations Schneider and Ingram (1993, 1997; Ingram & Schneider, 2005) introduced the idea of social constructions of target populations to better explain the agenda setting process and how policymakers craft policy to distribute benefits and burdens amongst target groups. Target populations are a group of people, typically a subset of citizens, who receive the benefits or burdens of a policy or whose behavior is targeted for change by a particular policy (Ingram & Schneider, 1991; see also Schneider & Ingram, 1993; Smith & Larimer, 2017). Social constructions2 are “the variable meanings that individuals or groups assign to various objects or processes associated with public policy” (Shanahan et al., 2017, pp. 178-179). These meanings and associated values result from experiences and exposure to stimuli, including discourse about the constructed group or construct (Schneider & Ingram, 1997). Schneider and Ingram (1993, 1997) argue that by looking at target populations, researchers can identify constructions that are 2 Schneider and Ingram (1993) are clear that, while social construction of groups and ideas is a concept from sociology that has a wider application to meaning making in general, they specifically focus on social constructions of target populations for their application to the fields of political science and policy studies. As such, I will use the term “social constructions” to apply specifically to the constructions of target populations throughout this dissertation. 7 pervasive in societal discourse and apply their understanding of these constructions to explain why some groups received more favorable policy benefits while others face greater policy burdens. There are political benefits to policymakers implementing policies with rationales and policy levers aligned with understood social constructions (Schneider and Ingram, 1993, 1997). Elected officials are motivated both by a desire to be re-elected and the intent to do public good when they design policy. These sometimes-competing motivations can result in strategic choices of who benefits and how, particularly through the use of different policy levers. Policies that benefit target groups constructed as deserving and burden those constructed as undeserving are more likely to be viewed as desirable or rational by constituents. In turn, constituents are more likely to vote for policymakers whose policies seem aligned with a broad societal understanding of which groups need and deserve intervention from the government. Thus, policymakers’ understandings of social constructions can impact who they view as needing policy or whose problems should be a part of the policy agenda. Once an issue associated with a target group is part of the agenda, the constructions influence how the policymakers attempt to change groups behaviors. For example, target groups that are believed to be more deserving are more likely to receive inducements to encourage a behavior while those targets considered undeserving are more likely to have behaviors discouraged through sanctions. Less politically powerful groups are more likely to have authority tools that expect compliance. After a policy is enacted, targets’ interactions with government based on shape how that group views government and government participation generally, impacting political participation beyond the particulars of specific policy. 8 Typology of Target Populations Schneider and Ingram (1993, 1997, 2019; Ingram & Schneider, 2005) conceptualize social constructions of target populations as a four-part typology theorizing that benefits and burdens are allocated based on the interaction of two characteristics of target populations: how deserving of benefits the target group is and the political power of the group. Target populations that are constructed as more deserving or having more political power are more likely to be distributed policy benefits, as it is politically advantageous for policymakers to reward such groups. Target populations that are perceived as either less deserving or less powerful are more likely to carry policy burdens. A summary of Schneider and Ingram’s typology and the distribution of benefits and burdens can be found in Figure 1. Figure 1. Schneider & Ingram’s (2019) Typology of Target Populations Social Constructions Deserving (+) Undeserving (-) Advantaged Contenders Business, Taxpayers, Big Unions, Big Banks, Strong Middle Class, Property Owners Wall Street, The Rich Political Power Mostly policy benefits, Benefits sub rosa, limited burdens burdens more symbolic Dependents Deviants Children, Women, Criminals, Illegal immigrants, Weak Minorities, the Disabled terrorists, sex offenders Benefits mostly rhetorical & symbolic, Benefits only when necessary, some burdens mostly burdens Note. Adapted from Schneider & Ingram (2019), Figure 1, p. 208. Schneider and Ingram indicate that target populations fall into one of four quadrants: the advantaged, contenders, dependents, and deviants (see Figure 1). An advantaged group (upper left), such as business, tend to have overt policy benefits and few burdens. Because business 9 owners are generally perceived as contributing to economic growth (deserving) and have influence via organized interest groups (politically powerful), policies with benefits such as tax exemptions can be justified as not only benefiting the individual businesses but encouraging economic growth that will benefit the public broadly. While contenders (upper right) also have access to policymakers, often through lobbying groups, society often frames these groups as more self-serving and perceive policies benefiting these targets as less helpful for the common good (undeserving). Policy benefits conferred to contenders, like unions, thus are often hidden and symbolic burdens that do not inconvenience the targets greatly may be emphasized by policymakers. Groups that are politically weak do not have the same influence and access to policymakers. Dependents (bottom left), like children, may be generally seen as worthy of benefit, so it is politically advantageous for policymakers to design policies to meet these targets’ needs. In the case of children, for example, their lack of voting power means that policy benefits may be more symbolic; the benefit is a positive talking point for a policymaker but often stops short of substantially addressing the policy problem or the access to the benefit may be limited to a narrow group or complicated through bureaucratic paperwork. When a group lacks political power and is perceived as undeserving, policymakers are most likely to craft policy that punishes deviant (bottom right) target populations. Tough punishments for criminals are often popular policies as society generally feels the burdens are warranted for those who do not follow the law. Schneider and Ingram (1993) argue that it is easiest for policymakers to address issues pertaining to the extremes– the advantaged and the deviants– as it is more clear what types of policies will be perceived as most rational. For policy to be congruous and make sense to both policymakers and constituents, the rationale for the policy should align with the social 10 construction of a policy target. Thus, when a negatively perceived group receives benefits, the rationale provided will likely de-emphasize these benefits, and when positively perceived groups receive burdens, these may be framed as unavoidable and minimized. Distribution of Benefits and Burdens. The level of political power of a group can indicate the level of influence and control they have on the policy design process. Schneider and Ingram (1993) note that advantaged groups are likely to be oversubscribed benefits, sometimes benefitting from policy addressing an issue that may be more directly addressed through targeting another population, as deviant populations are most likely to be oversubscribed to burdens. Advantaged groups tend to have high levels of control over the policy process, typically benefiting often and receiving few burdens. As such, assigning benefits to advantaged groups can be seen as a political opportunity for policymakers (Schneider & Ingram, 1997). Typical policy tools for delivering benefits to advantaged populations include subsidies, entitlements, free information, and outreach programs, while burdens will include inducements, charges, and occasionally sanctions (Schneider & Ingram, 1997). Contenders have higher control over burdens than they do for benefits, so they may avoid negative impacts of policy but benefit only in less visible or hidden ways. The policy tools used for distributing benefits to contenders can be unpredictable, but burdens often single out a high-profile group for heavy sanctions to make a visible example (Schneider & Ingram, 1997). Neither dependents nor deviants have much control over the policy process, and thus tend to both be oversubscribed to burdens and undersubscribed to benefits. Dependents tend to be subject to more paternalistic policies but are still conferred some benefits through income-tested or other regulated subsidies where the client proactively reaches out to government institutions to receive assistance. In this case, many of these benefits are a burden to receive or may be 11 stigmatized, even while reflecting cultural norms that these groups deserve assistance. Deviants typically only receive benefits with the threat of punishment attached, and burdens typically have more severe consequences, including death when considering the criminal justice system. Because cultural perceptions of undeservingness and policy rationales for punishment of deviants often align, policy burdens or punishments for deviant populations may also be politically advantageous and so can be more overt (Schneider & Ingram, 1997). Degenerative policy design (Schneider & Ingram, 1997) describes circumstances where the political motivations for policy design outweigh policy fairness, particularly for some policy issues. Schneider and Ingram argue that this can reinforce deception and manipulation in policy practices so that advantaged groups learn they benefit through self-promotion and mobilization, contender groups learn to be deceptive to strategically make policy gains, and dependents and deviants learn that no matter their level of engagement in the political system or mobilization that they lack the power to make changes. For example, in their study of 2016 Arizona state legislation, Schneider and Ingram (2019) find that most benefits are conferred to deserving populations, but that legislators employ deceptive feedback to hide these benefits from the general public, thus protecting themselves politically while also rewarding advantaged groups. While this can seem to be a self-perpetuating cycle, Schneider & Ingram (1997) note that degenerative policy design can be issue specific and as new information emerges, new leadership rises, or social constructions of groups change, fairness can be re-established within the system. Key to understanding policy changes, Schneider and Ingram (1993, 1997; Ingram & Schneider, 2005) indicate that it is possible for a target population to move positions within the matrix by changing how deserving or undeserving their group is viewed or by gaining political power. Schneider and Ingram argue that social constructions vary on several dimensions for a 12 given group of people, from pervasive to nonexistent, from positive to negative, from historically stable to constantly evolving. Those groups with few constructions attached to them seem to be less likely to be a target groups for policy; Schneider and Ingram (1997) provide the example of left-handed people to illustrate this. Similarly, prior to the advent of labor unions, workers were less likely to be constructed as a politically powerful group; with the organization of unions during the Industrial Revolution, the working class was able to minimize burdens via workplace safety reforms, among other related policies. Connections to other policy design theories Social construction theory builds on punctuated equilibrium theory, policy feedback theory, and policy learning theories, and is also a precursor to some of the ideas in Narrative Policy Frameworks. Baumgartner & Jones’s (2009) punctuated equilibrium theory holds that there is stability in policy over long periods of time that are “punctuated” or interrupted by periods of change when ideas and policy images shift. Schneider and Ingram’s social constructions provide one way of examining policy images and how they change. Policy images are how a policy is understood by different groups of stakeholders (Stone, 2012). Because different groups can have different experiences with the same policy, there may be multiple images of a particular policy (Baumgartner & Jones, 2009). Social constructions framework considers how target populations react to policies, which can include the policy images that may be influenced by how the group is framed as deserving or underserving within a policy. Punctuated equilibrium theory also notes that policy change occurs when images shift; social constructions framework acknowledges that constructions can change overtime and that this will impact policy designs. Policy images can change, according to Baumgartner & Jones (2009), when issue attention changes, a crisis occurs, or the political venue for the issue changes (i.e., a 13 new court ruling when an issue has previously been a legislative issue, such as in numerous Civil Rights policy changes). Ingram and Schneider (2005) posit that change in constructions may occur due to external events, which can encompass the triggers for change highlighted by Baumgartner and Jones (2009). Strikes are an external event that provide an opportunity for issue attention to change, providing an opportunity to examine shifting social constructions. Policy feedback theory, or path dependency theory, argues that the more policy choices that are made in a particular direction, the more difficult it is to change the direction of the policy going forward (Mettler & Sorelle, 2017); the policy feedbacks into itself and reinforces the images it created. Change occurs when there is feedback in conflict with this reinforcing loop from political actors and events, including mobilization of interest groups to promote benefits or protest burdens. Schneider and Ingram expand on this argument through their typology of target populations. This feedback can result from interpretive effects, or the messages conveyed by policy that impact how people perceive groups, government, and their relationship with government. Social constructions conveyed through policy can be interpretive effects, and Schneider and Ingram (1993, 1997, Ingram & Schneider, 2005) are keenly interested in their impacts on democratic behaviors. Schneider and Ingram (2005b) also argue explicitly that their typology helps explain how conferring benefits and burdens on particular target populations can be self-reinforcing (path dependent) until a tipping point (punctuation). For example, conferring benefits on advantaged groups has political benefits for policy makers, and will be reinforced by advantaged groups mobilizing to protect benefits (positive feedback), until such a point that the benefits become so great that the public begins to construct an advantaged population more negatively as “greedy” and thus less deserving. Changing public opinion can change the policy rationales and policy 14 images conveyed, and ultimately result in policymakers changing policy design – particularly the distribution of benefits and burdens– in response to negative feedback. Schneider and Ingram (2005b) state that their framework contributes to policy learning theories, including advocacy coalition framework (Sabatier, 1999) and institutional analysis and development framework (Ostrom, 1990), by highlighting political conditions that both hamper and encourage learning. In identifying categories of policy learning, May (1992) classifies one of his three types as social learning. He specifically discusses how changes to social constructions of policies or problems, including target populations, can change policy goals and expectations, highlighting the crossover role social constructions framework plays amongst policy studies theories and frameworks. In part as a response to criticisms that social constructions framework does not adequately consider institutions (Lieberman, 1995), Schneider and Ingram (1997, 2005b) consider the role of institutions in decision-making with their separation of degenerative and scientific and professional political arenas. Degenerative politics are self-reinforcing and discourage learning because there is little political benefit to change policy until there is political pushback from oversubscribing benefits or burdens to a particular target population. Alternately, scientific and professional arenas are more likely to value changes in knowledge that could alter the rationales and policy logic that support the social constructions of target populations embedded in policies, and thus new policy designs. Schneider and Ingram acknowledge that policy arenas and institutions differ in these decision-making characteristics, and thus differ on their likelihood of engaging in policy learning. Social constructions framework also complements Stone’s (2012) discussion of the role of policy images. Schneider and Ingram (1993, 1997, 2005b) acknowledge the importance of language in conveying social constructions. Stone (2012) provides examples of ways language 15 conveys policy images, which can include social constructions, including symbols embedded in narratives (stories, synecdoche, and metaphors), numbers, and identification of causes, among other devices. Within the study of social constructions of women on welfare, for example, elements of Stone’s description of synecdoche in policy can be found in the ways this target population is defined and segmented using labels such as “jezebels,” “matriarchs,” and “welfare queens” (Bensonsmith, 2005). One original criticism of social constructions framework is that the theory did not generate testable hypotheses (Ingram et al., 2007). Narrative policy framework (NPF) address some of these criticisms as it provides a systematic and empirical way to test hypotheses regarding how constructions impact policy and how policy in turn impacts democratic participation (Shanahan et al., 2013; Smith & Larimer, 2017). Studies applying NPF can use quantitative methods to determine the effect of a narrative on individual public opinions, larger beliefs and opinions embedded in policy, and less-often studied impact of institutional or cultural ideas. Studies employing this method often utilize surveys with experimental items to link narratives and public opinion or content analyses to quantifiably relate constructions of narratives to policy outcomes. Thus, NPF is a framework for quantifying static or changing constructions and apply the underlying ideas of the social constructions framework when the narratives studied focus on how target populations are constructed. Changes in Social Constructions As social constructions are theoretically linked to the distribution of policy benefits and burdens, it stands that changes in how a target group is constructed or in their political power would change policy. Schneider and Ingram (2005b) indicate that constructions can be fluid. Understanding the dynamic nature of constructions in specific context can help illuminate how 16 policies can change in how they treat particular target groups. Given the apparent shift towards positive portrayals of teachers presented in the media coverage of the Red for Ed strikes, it’s important to consider how and under what conditions perceptions of deservingness can change. In his review of Schneider and Ingram’s Deserving and Entitled (2005a), deLeon (2005) suggests their theory of social constructions needs further development regarding how social constructions change. Echoing this emphasis on changing constructions, Ingram et al. (2007) assert “a key challenge for the theory of social construction of target groups is in understanding the mechanisms underlying transitions from one cell to another” (p. 119). In their review of peer- reviewed studies applying social construction theory, Pierce et al. (2014) find that 43% of the applications of the theory they reviewed included some reference to social constructions changing. This is evidence that there is significant scholarly interest in understanding how constructions can change along the dynamics of power and deservingness, and, as Schneider & Ingram (2005b) point out in their response to deLeon’s critique, the role of policy design in the changes to construction and policy. Teachers’ Strikes as Political Activity To build on how constructions of teachers specifically have changed, it is important to first understand what events or conditions may cause a change in construction. Pierce et al. (2014) summarize Schneider and Ingram’s (2005b) comments, noting, a shift in categorization of a target population should be the result of (i) changes in perception of a target population from being deserving to undeserving or vice versa, (ii) external dramatic events, (iii) opportunities, and/or (iv) skillful manipulation by entrepreneurs (p. 19). Strikes and protest events are examples of potentially construction-changing external events or 17 social movements. Lipsky (1968) classically theorizes that protests are a political resource for the less powerful. While strikes might not typically be seen as protests, the wide scope of the Red for Ed events and the focus on issues beyond traditional collective bargaining have noticeable similarities to social movements (Lyon & Kraft, 2021). Wasow (2020) argues that strikes and protests are an opportunity for subordinate groups (or less politically powerful, in Schneider & Ingram’s terminology) to engage in agenda seeding – attempting to influence the policy agenda by engaging in events that attract media attention. This media attention is an opportunity to bring public and elite attention to particular political issues and to frame the conversation around those issues through actions and messages conveyed through the media coverage of such events. Evidence from the Civil Rights Movement showed that peaceful protests can result in more sympathetic feelings towards activists compared to violent protest (Wasow, 2020), indicating that political involvement by activists can have differential impacts on public opinion depending on the actions of the event participants. Teacher strikes can be perceived as both peaceful and as disruptive; while the Red for Ed strikes were not violent, the strikes did close many schools and interrupt classroom learning, potentially harming children and inconveniencing parents and their employers. As such, it may not be clear how the media would frame coverage- would strikes generate positive narratives highlighting education policy issues or would strikes instead promote negative portrayals of teachers abandoning their duties? Using survey data, Hertel-Fernandez et al. (2021) find that the long-term political impact of teacher strikes in 2018 includes increased support for teachers and the labor movement broadly. Similarly, the annual EdNext polls (Education Next, 2020; Henderson, et al., 2019) found a spike in support for increasing teacher pay in 2019 after the Red for Ed strikes. While these studies focus on changes to public opinion, Lyon and Kraft (2021) show that educational issues also 18 became more salient in congressional election campaigns following the Red for Ed strikes. Overall, we see evidence that strikes represent key events in how participant groups are portrayed in society, and that in particular the media coverage of these events can be a tool for participants to shape political conversations, thus the Red for Ed strikes are an ideal case for examining the ways media constructs teachers. Target Populations’ Agency in Changing Constructions Target populations themselves have some agency in how they are constructed. DiAlto (2005) explicitly notes the role of Japanese Americans in mobilizing to change how they as a group were treated in policy. Hudson and Gonyea (2012) also highlight that one of the reasons the elderly are historically constructed as advantaged is due to this demographic group’s mobilization to vote and to form advocacy organizations, contributing to the shift in political power. If target groups can be agentic in changing how they are constructed broadly, then these changes can impact policies beyond those the target group may actively advocate for or against. Teachers on strike present an interesting case for examining agency in constructions. Through handcrafted signs and messages, teachers may be able to frame the political conversation, media coverage, and public opinion by presenting themselves as sympathetic and deserving of benefits, despite the fact that teacher strikes often inconvenience children and parents (among others). Race as Important to Understanding Social Constructions It is impossible to consider how social constructions of target groups change without acknowledging that constructs such as race, gender, and class that are deeply entrenched in our social systems impact social constructions, particularly in the context of the United States. Ingram and Schneider (2005) acknowledge that social constructions reflect a human tendency to identify with and elevate their own group by constructing an “other” group that is in some way 19 less-than themselves. Social roles and perceptions of differences between groups are exaggerated compared to actual, observable differences. Ingram and Schneider (2005) emphasize that while there may be some group characteristics that are real or observable, the positive and negative perceptions of these differences and the societal power structures that they create are “the product of social and political processes” (p. 3). Intersectionality, or the complex interplay of an individual’s identities and experiences with dominant and nondominant culture, plays an important role in understanding group identities, how power and privilege are distributed unequally, and how culture and society perpetuate these inequalities (Sensoy & DiAngelo, 2017). Many target populations can be thought of as either dominant or non-dominant groups. For example, Pierce et al. (2014, see pp. 14-15) identify 141 target populations and find 9 invoke non-dominant gender constructions, 27 invoke non-dominant racial and ethnic constructions, 31 invoke non-dominant class constructions, 13 invoke non-dominant age constructions, and 14 invoke non-dominant health and ability constructions.3 Even within simple identifications of target populations, identity markers and the intersection of these identities plays a crucial role in defining these groups. Belonging to a non-dominant group can increase the likelihood of experiencing inequities relative to dominant groups and can shape policies that target these groups. The wider social construction literature, especially studies concerning the portrayal of women accepting welfare in policy, finds that the intersections of race, gender, and class all impact the ways target groups are discussed in policy (e.g., Amundson & Zajicek, 2018; Bensonsmith, 2005; Boris, 2007; Cammett, 2014; Cassese & Barnes, 2019; Cocca, 2002; 3 Some target groups did not fit into the listed designations, such as environmental groups, so they were not included in this illustrative count. Similarly, some target populations, such as middle-class women, were counted in two identity categories when appropriate. 20 Constance-Huggins, 2011; Foster, 2008; James & Rashid, 2013; Limbert & Bullock, 2009; Schram, 2005 among others). For example, some of these studies find that Black women accepting welfare assistance are particularly negatively constructed in comparison to other groups, including the now infamous portrayal of Black women as “welfare queens” even though white women represent a larger proportion of women accepting welfare assistance (e.g., Bensonsmith, 2005; Constance-Huggins, 2011; Hancock, 2003). While the research concerning social constructions of teachers does not often directly consider the impact of race and gender on these constructions, teaching in the U.S. context is a largely feminized profession (Goldstein, 2014), with greater concentrations of Black and Hispanic teachers in urban schools, relative to the demographics of the overall teacher force (Frankenberg, 2009). As such, it is critical to consider how race4 can impact how teachers are constructed as well. Social Constructions of Teachers Vary in Levels of Deservingness Some target groups, such as women accepting welfare assistance, have been studied extensively. Likewise, some education policy stakeholders, including children and teachers’ unions, have clearly defined, long-standing social constructions. However, teachers as a target population have less clearly defined social constructions. Teachers are often considered essential to the success of education policies (e.g., D’Amico Pawlewicz, 2020), thus they are a critical target population within education policies. Changing teacher behavior is perceived to have large impacts on educational outcomes. Thus, it is important to expand our understanding of how teachers are constructed. Social Constructions of Teachers’ Unions While teachers and teachers’ unions are clearly linked target groups, particularly as the 4 As the teaching profession is a female-dominated profession and as social class is closely tied to profession, I focus on race alone in this study. 21 latter represents the former in collective bargaining processes and are generally responsible for organizing, calling for, and ending a strike, the social constructions of these target groups vary on both dimensions. Within the politics of education, teachers unions are often considered critical players in shaping policy, to the point of being so powerful in local settings to have considerable control over school board elections (e.g., Moe, 2001; Finn & Petrilli, 2013).However, teachers themselves are rarely discussed as having such power; research regarding the accountability notes the lack of autonomy and power granted to teachers (e.g., Cochran-Smith & Lytle, 2006; Crespo, 2017). Historical accounts of the profession also point to unions as a source of political power, and teachers without union backing as victims of labor policies that depressed wages, benefits, and job protections (e.g., D’Amico Pawlewicz, 2020; Goldstein, 2014). While teachers’ unions are inherently made up of teachers, the union as a target group has notably more political power than teachers independently. Empirical research finds teachers’ unions typically constructed as contenders, reflecting Schneider and Ingram’s (2019) placement of unions within their typology. In particular, research emphasizes negative portrayals of unions (Bulkley & Gottlieb, 2017; Crespo, 2017; Goldstein, 2011). Osborne-Lampkin et al. (2017) indicate teachers’ unions are traditionally associated with preventing school improvement reforms (despite the mixed empirical evidence regarding differences in student outcomes by differences in union characteristics: see Cowen & Strunk, 2015). As such, while teachers are often conceptualized as crucial to improving student educational outcomes, teachers’ unions are often constructed negatively as obstructionists. This contrasts with the literature that indicates teachers’ unions enhance teacher voice in the political process for the betterment of the education system (e.g., Choi & Chung, 2016, Han & Maloney, 22 2019, Hirschman, 1970, Ravitch, 2006). Since Janus v. AFCSME (2018) ruled that public sector unions, including teacher unions, could only collect fees from members, teachers’ unions have been limited in how they can fund union activities. Lyon (2021, p.2) illustrates how states have increased restrictive labor laws that limit teachers’ union political power. If teachers’ unions have less power than they have historically had in education policy and continue to be constructed as less deserving, then it may be less advantageous for teachers as a target group to affiliate with unions. Instead, teachers may dissociate to reframe themselves as educational experts and advocates who have the experience to improve the system for children. Social Constructions of Teachers Generally Research that considers social constructions of teachers often finds both positive and negative portrayals of teachers (Bulkley & Gottlieb, 2017; Cochran-Smith & Lytle, 2006; Crespo, 2017; Goldstein, 2011). Recent accountability policies have represented teachers in more negative ways (Crespo, 2017) and granted teachers little agency (Cochran-Smith & Lytle, 2006), suggesting teachers are both less deserving and lacking political power. Considering how teachers are described in public discourse, Goldstein (2011) analyzes mass media portrayals of teachers and notes more than half of the articles portrayed teachers negatively. These negative portrayals of teachers often associate teachers with teachers’ unions (Crespo, 2017; Goldstein, 2011). There is also evidence of more deserving constructions of teachers. Bulkley & Gottlieb (2017) find that elite policy actors view teachers along a spectrum of three archetypal images: teachers as professionals, individual great teachers, and teachers trumped by systemic dysfunction. These archetypes captured nuanced views of teachers, with some groups of teachers 23 being constructed more positively than others. For example, the image of the great individual teacher positively highlights the important and potentially transformative role of teachers in student experiences, but also emphasizes that these “great” teachers are a departure from much of the teaching force. Many policy elites conceptualized some teachers as worthy of policy benefits, but not all. This is also evidence that constructions of teachers may differ based on sub- groups of teachers. While few studies clearly place teachers within Schneider and Ingram’s typology, Kreitzer and Smith (2018) use crowdsourcing via Amazon’s Mechanical Turk to place teachers in the dependent category; however, there is notable variation in patterns of consensus when viewing standard deviations in response. Kreitzer and Smith (2018) note that children have significant consensus in how they are portrayed as dependents, while other groups, like labor unions have more variation while fitting the definition of being a contender. In the context of a labor activity like a teacher strike, it may be argued that teachers are acting as union members, not as teachers alone. However, the organization of the early Red for Ed strikes (West Virginia, Oklahoma, Arizona, and Kentucky) was primarily through Facebook groups run by teachers, and unions were seen as primarily serving a support role (Will, 2018). Unlike past teacher strikes, these strikes occurred in areas with relatively weak teachers’ unions, particularly in Republican-led right-to-work states. Thus, one way I hope to disentangle when teachers are being constructed as teachers versus as teacher unions is to look specifically at these early Red for Ed strikes to see if teachers were portrayed differently. Also, as teachers are seen as slightly less deserving than children and less powerful than unions (Kreitzer & Smith, 2018), I hypothesize that by associating with either children or unions, teachers as a group may borrow perceptions of deservingness or power, respectively. 24 The News Media as a Source of Social Constructions Ingram and Schneider (2005) indicate that the news media is one important source of social constructions. Media can be particularly important to study as the news media can serve as a political agenda-setter, because the attention they provide an issue and how they present that issue can impact the salience of the issue to the public and to policymakers (McCombs & Valenzuela, 2021). While the media does not always change public opinion, it can inform the public regarding issues and help determine what information is important for the public to know (McCombs & Valenzuela, 2021). The news media is commonly used as a source for identifying variation in constructions. For example, DiAlto (2005) traces how constructions of Japanese Americans shifted from undeserving to deserving since the beginning of the twentieth century, focusing specifically on how media, public policy, and court proceedings reflect constructions that progress from “problem minority” to “enemy aliens” with the advent of World War II, and then to the more positively constructed “model minority” construction post-war. Several content analyses of media also trace changes in construction. Bleich et al. (2018b) and Bleich & van der Veen (2018) use sentiment analysis to trace how the tone of media stories (positive, negative, neutral, or mixed) for particular target groups changed over time and McGinty et al. (2016) examine news stories to trace the development of new policy narratives. When using the news media to examine constructions, researchers have focused on article language (e.g., Bain et al., 2017; McGinty et al., 2016; Rose & Baumgartner 2013), tone (Bleich et al., 2018a; Bleich et al., 2018b; Bleich & van der Veen, 2018), and images (e.g., Farris & Silber Mohamed, 2018; Silber Mohammed & Farris, 2020). These studies link the mass media’s portrayal of groups, particularly how groups are negatively constructed, to how target 25 groups are constructed in public opinion and public policy. Political Polarization in News Media Increasingly, news media exists in a politically polarized landscape within the United States (Wilson et al., 2020). News media contributes to this trend as some sources have developed more ideological leanings and identities (Groeling, 2013), and increasing polarization in media can be traced back as early as 2014 (Jurkowitz, et al., 2020). In addition to some news sources becoming more politically polarized in their content, their audiences are also becoming more polarized in their media source selections; members of the public can self-select to consume media that tends to align with their ideological views. For example, Jurkowitz, et al. (2020) found that 60% of those who identify as Republicans or Lean Republican rely on Fox News for their political news, while Democrats or those who Lean Democrat trusted CNN. This is potentially concerning, because consuming news sources from across the ideological spectrum can moderate consumers’ viewpoints (Broockman & Kalla, 2022). There is also evidence of increasingly partisan views of educational issues, previously thought to be “exceptional” from other policy issues in U.S. governance in part because of the lack of partisan division (Henig, 2013). However, even as early as 2019 more people who identify as Democrats or as Leaning Democratic thought education should be a top priority (76%) compared to those who identified as Republicans or Leaning Republican (58%) (Mitchell et al., 2019). Van Green (2021) also finds that Republicans are increasingly critical of public schools. Given that media is an important source of agenda setting, has become more polarized, attitudes around issues in education have become more partisan, and consumption patterns can relate to public opinions, it is important to consider how different media sources may portray teachers differently. 26 Social Constructions of Teachers, the Red for Ed movement, and News Media Social constructions impact policy design and target populations experiences with government and the policy process. While some target populations, like children and unions, are clearly defined, social constructions of teachers have been nuanced, often varying in levels of deservingness and political power that results in a nebulous conception of where teachers as a target population fall on Schneider and Ingram’s (1993, 2019) typology. As the racial composition of the teacher workforce varies by geographic location and because racialized policy images and stereotypes can influence social constructions, it is important to also consider how teachers may be constructed differently based on race. Teacher strikes, as a form of political protest and opportunity to exert political power, represent an opportunity to examine policy images and social constructions. By initiating in strikes and carefully crafting their messaging around strike issues, teachers may also be able to display agency in shaping constructions. In particular, teacher strikes can represent an opportunity for agenda seeding (Wasow, 2020), where teachers as a potentially less-politically powerful group can influence the policy agenda by attracting media attention. Playing the role of political agenda-setter, the news media is a rich source of social constructions data. The increasing political polarization of news media and increasing presence of partisan differences in opinion about education policy issues suggest that it is also important to consider how news media’s political biases may shape coverage. As a result, I designed a study that seeks to compare social constructions of teachers across location contexts and news media sources to identify how teachers were portrayed by the news media during the Red for Ed movement and if those portrayals varied across cases. I also consider how teachers were agentic in shaping media’s construction through the messages pictured on handmade strike signs. 27 CHAPTER 3: DATA AND METHODS To examine social constructions of teachers on strike across several dimensions of variation, I employed a multiple case study design. By selecting multiple cases, I can compare amongst similar and different contexts to enhance the external validity and generalizability of my findings (Merriam & Tisdell, 2016). I purposively selected seven teacher strikes or walkouts as cases (see Table 1) to examine media coverage in detail for evidence of social constructions. While these events all can be characterized as large-scale teacher labor stoppages with many similar or related reasons for striking, they differ in dimensions of time, location context, and outcomes. I utilized this variation to allow me to compare which constructions of teachers are prevalent in differing contexts. Because media plays a key role in agenda setting for the policy process, I analyzed the images and text of online news articles. I created a dataset of 828 news articles from eight national news sources. Sources vary in political lean and in their original format (newspaper, television network, or online news outlet). The resulting dataset allowed for comparison among strike events and news media sources. Multiple Case Studies Design In 2018 and 2019, there were over eighteen major5 teacher strikes or walkouts across the U.S. A summary of these major strike events can be found in Appendix A. Across these strikes, several key issues appear again and again: (1) teacher compensation, including salary and pensions, (2) school funding, and (3) improved working conditions, such as increasing support staff and decreasing class size. While the reasons for the strikes were consistent, there were differences among the strikes based on time and location. And while generally the strikes are viewed as collectively successful in securing policy benefits for teachers, some strikes were 5 I define major strikes, walkouts, or sickouts as those that received national news coverage. 28 decidedly more successful —such as the 2018 West Virginia teacher strike that resulted in 5% pay raises — than others, like the 2018 Oklahoma strike that resulted in no tangible policy gains beyond those already conceded by the state prior to the strike. I selected the three strikes that were considered the beginning of the Red for Ed movement in the spring of 2018 – West Virginia, Kentucky, and Oklahoma – to capture how media initially portrayed the movement. I focus on the first three statewide strikes of 2018 because as the number of strikes increased news coverage became increasingly overlapping and focused on the full collection of strikes throughout the year, rather than on particular strike events. There also seemed to be decreasing amounts of coverage due to saturation of the topic6. These 3 events were all statewide efforts, typically portrayed as more grassroots rather than union-led, and occurred in states that are traditionally less friendly towards labor unions and strikes and some that even specifically outlaw them. Kentucky, in particular, threatened discipline action against teachers that called in sick to participate in later protest events in 2019 (Schreiner, 2019). All three states also had Republican governors and Republican-controlled legislatures at the time of the strike (Ballotpedia, 2022a-i). To understand how the Red for Ed movement evolved over time, I selected three major strike events from spring and fall 2019 – Los Angeles, Denver, and Chicago. I selected three strikes occurring in three different states (California, Colorado, and Illinois); I selected the Los Angeles strike instead of the Oakland strike because of the size of the district, as one of the largest in the country, and because the Los Angeles strike preceded the Oakland strike. These strike events differ from the three initial cases in that they were districtwide strikes within major metropolitan cities, strikes were more clearly union-affiliated, and the cities’ mayors associated 6 For example, The New York Times published 1 online article covering the North Carolina teacher walkout in late spring of 2018, compared to 22 articles covering the West Virginia strike. 29 with the Democratic party, who may be perceived as more willing to work with labor groups and unions. Thus, I considered how the overlapping contexts of urbanicity and political lean interact. As strikes within charter networks occur in a different political context and governance structure from those focusing on traditional public schools, I exclude charter network strikes for the sake of comparability. The strike locations also differ by teacher demographics. Two statewide strikes –West Virginia and Kentucky– have a higher percentage of white teachers compared to the national average, while two of the citywide strike locations – Los Angeles and Chicago7 – have a noticeably lower percentage of white teachers than the nation as a whole (U.S. Census Bureau, 2022). Denver and Oklahoma have a similar percentage of white teachers compared to the nation (U.S. Census Bureau, 2022). The 2018 and 2019 Red for Ed teacher strikes were often discussed as a new era of teacher activism, reminiscent of the wave of teacher strikes in the 1960s-1980s (Reilly, 2019). To understand if teachers are being portrayed differently during the Red for Ed strikes than they have been when on strike in recent times when strikes were less common, I included news coverage from the 2012 Chicago teacher strike as a comparison case to the Red for Ed strikes. Unlike the 2018 & 2019 strikes, the Chicago strike was a localized, stand-alone event. This strike was widely covered in media and was a large-scale strike event in relatively close time proximity to the more recent Red for Ed strikes. I compared the 2012 Chicago teacher strike’s news coverage to that of the 2019 Chicago teacher strike that was near the end of the Red for Ed strikes to see how coverage differs within the same location in a relatively short period of time. A summary of the selected strike cases can be found in Tables 1 and 2. 7 The Chicago 2012 strike occurred between ACS data collections, but there was relatively little change in teacher profession data. 30 Table 1. Summary of Case Study Strike Events and Characteristics Strike Strike Event Characteristics Event Strike Strike Data % teacher Government Strike Outcomes3 Dates Length Collection population Partisan & Scope Range white1 Lean2 Chicago 9/10/12- 9 days, 8/27/12- 53.22% Democrat Wins- 17.6% pay raise over 4 years with other 2012 9/18/12 citywide 10/2/12 protections, test scores count less in teacher evaluations Losses- extended school day, no priority rehiring of laid-off teachers West 2/22/18- 14 days, 2/8/18- 95.54% Republican Win- 5% pay raise for teachers Virginia 3/7/18 statewide 3/21/18 2018 Kentucky 3/30/18, 3 days, 3/16/18- 92.41% Republican Win- increase in per pupil funding 2018 4/2/18, statewide 4/27/18 Loss- did not stop passage of pension reform bill 4/13/18 teachers opposed Oklahoma 4/2/18- 11 days, 3/19/18- 79.94% Republican Win- increase in school funding and teacher pay via a 2018 4/12/18 statewide 4/26/18 tax increase passed prior to the strike Loss- no changes after strike Los 1/14/19- 8 days, 12/31/18- 25.81% Democrat Wins- 6% pay raise, reduced class sizes, additional Angeles 1/21/19 citywide 2/4/19 school support staff 2019 Denver 2/11/19- 4 days, 1/28/19- 77.97% Democrat Win- universal teacher pay raise in place of pay-for- 2019 2/14/19 citywide 2/28/19 performance Chicago 10/16/19- 17 days, 10/3/19- 55.35% Democrat Wins- 16% salary increase over 5 years, reduced class 2019 11/1/19 citywide 11/14/19 sizes, additional support staff 1 2 3 Sources. U.S. Census, 2022; Ballotpedia, 2022a-i; Loewus, 2019; Pearson, 2012; Smith et al., 2009; Turner et al., 2018 31 Table 2. Summary of Union Power by Case Study Events Strike Event Union Power Characteristics State-level Strike Collective Legal for % teachers union power characterized bargaining teachers to in union or ranking as union-led allowed1 strike1 similar org.2 (N=50)3 (unavailable Chicago 2012 union-led required Yes 8- strongest ) West Virginia grassroots permissible No 73.9% 13- strong 2018 Kentucky 2018 grassroots permissible No 56.7% 28- average Oklahoma 2018 grassroots required No 59.4% 43- weakest Los Angeles 2019 union-led required Yes 90.7% 6- strongest Denver 2019 union-led permissible Yes 57.2% 35- weak Chicago 2019 union-led required Yes 96.3% 8- strongest 1 2 3 Sources. Nittler, 2019; National Center for Education Statistic, 2018; Winkler, et al., 2012 News Media Sources Selection As news sources have editorial discretion, it is likely that different news sources will cover events in varying ways. To consider how this coverage and the constructions embedded in it vary, I purposively selected news sources that varied in audience and political lean. I collected articles about all seven strike cases from eight general national news sources: Breitbart News, CNN, Fox News, New York Times, Slate, USA Today, Wall Street Journal, and Washington Post. Four newspaper sources – New York Times, USA Today, Wall Street Journal, and Washington Post – were among the top 5 newspapers in circulation in 2019 (Mercier, 2019). The New York Post was the fourth-ranked newspaper and was excluded from data collection as the modern version of the paper is publish in a tabloid format, rather than a traditional daily newspaper (Muck Rack, 2023). To capture how online news outlets may differ from traditional newspapers, I also collected data from 4 online news sources – Breitbart News, CNN, Fox News, and Slate. 32 Two of these sources – CNN and Fox News – were originally television news networks that also produce online news articles. The remaining two news sources – Breitbart and Slate – developed as online news sources. Table 3. Summary of Selected News Sources National News Source News Source Characteristics Newspaper AllSides News Partisan Lean Ranking by Media Bias Source Analysis Group National Ratings1 Type Circulation2 Online Breitbart News Right-leaning Right N/A News Television Fox News (Online News) Right-leaning Lean Right N/A Network Wall Street Journal Center Center 5 Newspaper (News) USA Today Center Center 3 Newspaper Washington Post Left-leaning Lean Left 1 Newspaper New York Times (News) Left-leaning Lean Left 2 Newspaper Television CNN (Online News) Left-leaning Lean Left N/A Network Online Slate Left-leaning Left N/A News Note. Wall Street Journal’s Opinion section was ranked as Lean Right, and New York Times’s Opinion section was ranked as Left in 2019. As Opinion articles are included in the dataset, this has placed each of these at further extremes of the spectrum than similarly ranked news sources. Sources. 1AllSides, 2023; 2Mercier, 2019 Recently, media sources have been labeled as having particular political biases that influences the tone and topics of coverage. Pew Research Center (Mitchell et al., 2021) found that about a quarter of people who identify with either the Republican or Democratic party will only consider sources that cater to an audience that matches their political ideology. While this trend in polarization has increased notably over the past 2 years, the trend in polarized trust in 33 news media based on partisan ideology existed as early as 2014 (Jurkowitz et al., 2020). For example, Republicans are more like to trust media coverage from Fox News while Democrats are more likely to turn to CNN (Jurkowitz et al., 2020). To capture this variation in news coverage, the eight national news sources that I consider vary regarding media bias, using AllSides 2019 Media Bias Ratings (2021). Two sources are considered either Right or Right Leaning – Breitbart and Fox News. Two sources are considered Center in their coverage, leaning neither Right nor Left consistently – Wall Street Journal and USA Today8. Four sources are considered Left or Left leaning – Washington Post, New York Times, CNN, and Slate. By capturing the full political spectrum of media coverage, I was able to compare if differences in teacher social construction are an artifact of the media source’s biases or if constructions are in fact more consistent across sources. A summary of these national news sources’ political biases can be found in Table 3. Sampling and Data Collection For each of the strike cases, I gathered the full collection of online news articles that include images from each of the selected news sources available via each news organization’s search function for the time period of two weeks (14 days) before the strike began until two weeks (14 days) after the strike ended. In the case of Kentucky, which had three one-day sickout events, I searched for coverage from the entire time period from the day before the first protest event (March 29, 2018) until the day after the last protest event (April 14, 2018) to capture any potential articles capture the collection of the events at the time they occurred. I used online versions of articles as these are commonly accessible to both subscribers and the general public, 8 In 2019, USA Today was rated with a Center media bias from AllSides. In 2021, AllSides updated their rating to Leans Left (AllSides, 2023). However, because data was collected from 2018-2019, for the purposes of this study, I use the 2019 ratings. 34 compared to the more limited reach of paper newspapers. This also results in a greater number of news stories, as some coverage does not make the space-limited paper forms of these newspapers. Likewise, while a paper news article contains a small number of photos, often only one, due to space and cost constraints, online versions can easily embed hundreds of photos in photo galleries and within articles. Articles were collected directly from the media source’s website using the search terms “(strike location) teacher strike.” In the case of USA Today, the search function did not produce a robust data set, so I reviewed article titles from each media source’s Site Map for each day within the data collection range. I included any article whose title was related to the event and included the words “strike,” “teacher,” “walkout,” “protest,” “school,” and the location name. I collected all articles with images that discuss each of the seven teacher strike events for two weeks before the strike event through two weeks after the strike event. A summary of articles collected organized by strike event and news source can be found in Table 4. Table 4. Summary of Number of Online News Articles Collected by News Source and Strike Event News Strike Event Total Source Chicago West Kentucky Oklahoma Los Denver Chicago 2012 Virginia 2018 2018 Angeles 2019 2019 2018 2019 Breitbart 6 1 0 1 16 4 8 36 Fox News 27 7 10 9 4 6 7 70 WSJ 15 6 6 7 16 3 11 64 USA 20 21 28 47 34 23 34 207 WP 31 19 21 32 22 8 10 143 NYT 23 22 21 17 29 4 14 130 CNN 40 26 19 30 13 13 17 158 Slate 6 6 2 3 1 0 2 20 Total 168 108 107 146 135 61 103 828 35 Table 5. Summary of Number of Total Images in Articles by News Source and Strike Event News Strike Event Total Source Chicago West Kentucky Oklahoma Los Denver Chicago 2012 Virginia 2018 2018 Angeles 2019 2019 2018 2019 Breitbart 6 1 0 1 29 13 9 59 Fox News 27 12 18 18 5 8 12 100 WSJ 43 8 18 20 27 3 14 133 USA 70 46 100 138 121 83 120 678 WP 85 36 37 49 28 11 12 258 NYT 55 33 46 33 41 4 30 242 CNN 91 67 76 117 49 45 67 512 Slate 6 6 2 3 1 0 2 20 Total 383 209 297 379 301 167 266 2002 Only articles with images were included in the dataset as my analysis focuses on how teachers were visibly portrayed. In an online news environment where readers can view large amounts of news data, the focus on the images captures what subtle messages are being conveyed about teachers on strike even to the casual reader skimming articles and headlines. Images include any still photographs, infographics, and screenshots of the beginning of video clips to simulate what a reader would see in a quick skim of a piece. When a news article includes a link to a photo gallery, I include any embedded images within the original article, as most use a photo as a link to the larger gallery and a reader would see this image without following the additional link. I then include the linked photo gallery as a separate news article, with all photos and captions. If an image appears multiple times, I include it each time it appears, as the relative salience of the image and the constructions it conveys is higher due to the potential for greater exposure to the public. Some articles, particularly in more recent times, also include embedded tweets; however, as the original author of the tweet can remove these 36 statements and because these were not fully controlled by the media source, they were excluded from this analysis. I do not analyze unrelated images, defined as an image that is not related to the coverage of the strike event or educational issues. For example, many news sources compile brief descriptions of the top news stories of the day. While the strike-related coverage within these summary briefs may be particularly salient to readers, as this digest version may be their only limited exposure to the issue, the coverage of other unrelated events is not closely related to the research questions. In total, the dataset includes 2,002 images for analysis (see Table 5). Differences in Sample Collection Procedures for Data Sources I applied the search criteria described above to collect my sample for all sources except when the news site embedded search features did not return appropriate results. When this occurred, I supplemented the initial data collection by expanding search terms and using features unique to that source’s website. In all cases, I attempted to collect all related articles within the time frame from the source and ended the search when I found no new articles to include. Breitbart. Breitbart’s search feature would only bring up the most recent articles. As the searches were conducted after the completion of the strike events, this sometimes meant that the search terms “(location) teacher strike” would not bring up articles within the intended time frame. To supplement this search process, I added the year the strike occurred to the search terms “(location) teacher strike (year).” Breitbart also uses a series of tags to categorize articles, so I reviewed all articles with the following tags to see if they fit within the inclusion parameters: Red for Ed, #RedforEd, Red4Ed, teacher’s strike, and teachers’ union. USA Today. USA Today’s search functions only return five pages of results, and the relevance of those results varies. To circumvent this, I used the Sitemap on the website to view the full listing of news articles published each day during the search periods. I reviewed titles of 37 all articles to identify related stories. To ensure I captured as many related articles as possible, I supplemented my review by using the web browser’s Find function to identify titles with the following key words: “teacher,” “school,” “strike,” “protest,” “walkout,” and the particular location (i.e., Kentucky or KY). Because the Sitemap is limited to articles dated October 2012 or more recently, I used a snowball sampling method to compile the Chicago 2012 USA Today articles. I first used the ProQuest Newspaper search to find all articles from USA Today in the time frame for the terms “Chicago teacher strike.” Then I used Google to search for each article title from ProQuest to find the online news article version. I then clicked embedded links to additional articles within my Google results to capture additional coverage. Qualitative Content Analysis I conducted a qualitative content analysis to better understand how media conveys social constructions of teachers, with attention to nuanced differences based on case context. A qualitative approach fits this study as it is intended to be contextual, explanatory, and generative (Ritchie & Ormstrom, 2014). Qualitative research describes more fully the context of social constructions of teachers, describing how teachers are portrayed across a variety of media sources and strike events to gain a better understanding of how teachers are viewed as worthy or unworthy of policy benefits. The multiple case study approach also allows for this study to have explanatory power (Bryne, 2009), linking strike event contexts and news source characteristics to the presence, absence, or prevalence of particular constructions. I use a content analysis to structure my work. I consider both the content — the images and text included in the news articles — and the context of the strikes in my analysis (Spencer, et al., 2014), linking aspects of language, symbolism, and the messages conveyed by the media to the point in history, characteristics of the strike event and location (including union involvement 38 and urbanicity), and characteristics of teachers depicted on strike, specifically the race. This approach allows me to focus on the interpretation of how particular social constructions are conveyed in subtle ways. Content Analysis in Studies of Social Constructions Other research on social constructions has employed content analysis to gain deeper understanding of the applicability of the theory to the policy process. Regarding research on the target group of women accepting welfare assistance, content analyses allowed researchers to describe the tone of constructions, how it has changed over time, and link those changes to variations in policy design. Systematic code applications assist in quantifying occurrences of particular constructions in order to track variation. Some content analyses use coding schemes to identify themes related to the portrayal of specific target groups in news media and track the evolution of those themes (McGinty et al., 2016), while others utilize sentiment analysis software to identify the tone of media coverage and how it changes over time (Bleich et al., 2018b; Bleich & van der Veen, 2018). Complementary Analyses of Image and Text Data My study combines these two methods to be able to analyze both the images and the text of the articles in my dataset. I analyzed the content of my data in two different ways. First, I analyzed the images included in the news articles using a coding scheme developed both deductively based on the theory of social constructions of target populations and the literature about policy images of teachers and inductively through an iterative review of the image data. Second, I analyzed the text of the articles using dictionaries developed deductively and then inductively using Ingram et al.’s (2007) definitions of political power and deservingness, coupled with sentiment analysis of the text. I employed analytic software to aid in the review of large 39 amounts of data that allow my findings to be more generalizable while also being sensitive to the nuance of visual and language data. For both analyses, I considered how social constructions did and did not vary based on the strike event and characteristics of the locations, over time, and by the news source. Table 6. Summary of Number of News Articles in Text Analysis by News Source by Strike Event News Strike Event Total Source Chicago West Kentucky Oklahoma Los Denver Chicago 2012 Virginia 2018 2018 Angeles 2019 2019 2018 2019 Breitbart 4 1 1 0 16 4 8 34 Fox News 15 7 9 10 3 6 4 54 WSJ 15 6 7 6 16 3 11 64 USA 17 13 23 15 18 12 10 108 WP 27 16 28 21 17 8 8 125 NYT 19 17 11 13 18 3 7 88 CNN 28 20 24 16 11 10 16 125 Slate 6 6 3 2 1 0 1 19 Total 131 86 106 83 100 46 65 617 To address my research questions, I consider my dataset in 3 subsets. First, there is the full dataset, which includes all 828 articles and their 2,002 images across all 7 cases. The content analysis of images is conducted across this full dataset. As the full dataset includes data from the Chicago 2012 strike that pre-dates the Red for Ed movement, the portion of the analysis that considers how the movement was portrayed focuses on a subset of the 6 cases from 2018 and 2019 (West Virginia 2018, Kentucky 2018, Oklahoma 2018, Los Angeles 2019, Denver 2019, and Chicago 2019), which includes 660 articles with 1,619 images. Linguistic analysis of text is more reliable when the document has more words, and LIWC software cautions against using text with less than 50 words (Pennebaker Conglomerates, 2023). As some of the articles included 40 in the full dataset are news briefs that have limited text about the strikes or very short descriptions coupled with video news coverage, I limit my text analysis dataset to those articles with at least a paragraph of text, which results in a subset of 617 articles (see Table 6). Image Content Analysis Coding Scheme. The qualitative coding scheme included both deductive and inductive codes focusing on the social constructions conveyed by images and image context, including the content of protest signs, the racial and gender composition of teachers pictured, the role of the individuals pictured (teacher, student, political leader, etc.), and the strike event pictured. Early versions of this coding scheme focused on the photo context. These codes were tested on a small subset of images (N=12) and were then refined through an iterative process of conversations with a research thought partner and analytic memos, in keeping with the recommendations of Saldaña (2016). This process was repeated with a larger section of the dataset – data from The New York Times from the time period of one day before the strike through one day after the conclusion across all seven strike events (N=99) – and ultimately resulted in the codebook in Appendix B. To understand how teachers were socially constructed by images, I created deductive codes that mirror Ingram et al.’s (2007) typology of social constructions. I used Schneider and Ingram’s definitions of deservingness and political power9 as a starting point for this code. I then augmented these definitions using literature about how teachers are portrayed. For example, Bulkley and Gottlieb (2017) identified three archetypes10 of teachers and teaching as a profession 9 “The political power of the target group is one dimension indicating the extent of its political resources, such as whether it is large, united, easy to mobilize, wealthy, skilled, well-positioned, focused on issues of concern to it, and accustomed to voting, contacting public officials, and so on. […] The second dimension […] refers to the valence or the positive and negative social construction of the group as more-or-less worthy and deserving and as contributing more-or-less to the general welfare.” (Ingram et al., 2017, p. 101) 10 Bulkley & Gottlieb’s (2017) three archetypes are: “Profession of Teaching Struggling Against Difficult Circumstances,” “Dysfunctional Structures of Teaching Trump Teacher Quality,” and “Individual Great Teachers Can Overcome All Obstacles.” 41 that I associated with Schneider and Ingram’s typology. I also used Goldstein’s (2014) history of the teaching profession to identify other common portrayals of teachers and teachers’ unions. For example, Goldstein (2014) discussed how early expansion of public education led to portrayals of young female teachers moving west as missionaries who were spreading democratic ideals and American ethics (p.13). She also discussed how teachers’ unions in the Civil Rights Era were often depicted as militant in their fight for policy changes. I considered how the emotions of teachers reflect deservingness. I associated teachers with facial expressions depicting happiness as deserving and images showing teachers as angry as undeserving. To accurately capture these emotions, I utilized images of emotions used to help children recognize facial expressions (e.g., Morin, 2020) and identified key facial cues. Happiness was characterized by smiles and relaxed facial expressions, while anger was characterized by narrowed eyebrows, furrowed forehead, and a mouth with downturned edges or potentially open. Finally, I included descriptions of images and themes that I noticed through the iterative review process that seem to be associated with deservingness or political power. For instance, images that include strike signs portraying teachers as strong women generally conveyed messages of deservingness, and images with large numbers of teachers on strike emphasized the high political power of teachers through the size of their group and organization. Inductively, I grouped ways that teachers could be perceived as deserving into three themes: deserving because of association with children, deserving because education is a public good, and deserving because teachers are worthy of respect or economic benefits. Similarly, I also categorized ways teachers can be perceived as high power as either being associated with teachers’ unions or being associated with the #RedforEd movement. To illustrate how these constructions can be portrayed via images, I offer some descriptions of exemplar images in Table 7. 42 Table 7. Examples of Applications of Image Coding Scheme Social Constructions Codes Code Description of Exemplar Image Explanation of how code fits Deserving Photograph of three Los Angeles Strike signs reference important teachers facing the camera and smiling, skills teachers develop in students holding handmade strike signs that for the betterment of society, include the following text in bright specifically how to self-advocate colors: “I teach MY STUDENTS to and how to read. The teachers are STAND UP for THEMSELVES & also smiling and posing with their here’s MY […]” and “CAN YOU READ signs, a non-aggressive stance that THIS? YOU’RE WELCOME” (Davis, showcases their message. 2019). Undeserving Photograph of parents protesting during Parents are shown in opposition to the Chicago 2012 strike. The parents are the teachers on strike, and not wearing red and are not with specifically carry messages teachers. This is a counter-protest, with alleging the strike causes mass-produced signs that read “350,000 significant harm to children by CPS HOSTAGES! Let Our Children equating them to “hostages.” Learn” in yellow type on black signs (Nicas, 2012). High Power Photograph of hundreds of teachers Teachers are shown as having high marching in West Virginia. Most are levels of political power through wearing red clothing and holding hands their display of solidarity with with their arms overhead. Shirts from the unified clothing and handholding, front line include messages such as as well as with the large number of “United” over an outline of the state. No teachers present. shirts reference a teachers’ union (Dorst, 2018). Low Power Photograph of state lawmakers in Rather than showing teachers Kentucky debating the teachers’ pension protesting to influence the state plan. There are no clear images of legislature, the image shows teachers. The image appears under a lawmakers discussing the pension headline that reads, “Kentucky reform bill without teachers lawmakers pass pension overhaul despite present. The teachers’ political loss teacher protests” (Associated Press, is also associated with the image, 2018). highlighting a lack of political power to maintain policy benefits. 43 During the iterative development of the coding scheme, it was apparent that the strike signs pictured were crucial to how teachers were being portrayed. The signs are also important to understand the role of two key groups of actors in contributing to societal impressions of social constructions: news media and the target population-teachers. The news media maintains control over which images are collected and published, serving as a curator for images the mass public is exposed to. Teachers themselves show agency in the construction process by crafting the messages on their strike signs; the media cannot publish photos of strike signs that the teachers do not create and display. Given the important role of strike signs in conveying symbols and stories that contribute to the social constructions of teachers in these news images, I created descriptive codes to identify the types of messages displayed by teachers and to identify when strike signs were handmade (crafted by teachers) or not (typically supplied by unions en masse). Examples of strike signs exemplars can be found in Appendix B. Because social constructions are shown to be related to stereotypes based on identity characteristics such as race and gender (Ingram & Schneider, 2005), I also considered the racial and gender composition of the groups of teachers in photos. I coded the apparent race and gender of the teachers in the photos by coding if the majority of teachers pictured seemed to present as either people of color (POC) or white and male or female. In cases where teachers belong approximately equally to these dichotomous groupings, I coded the photo as representing a diverse racial or gender composition. In cases where the race or gender of teachers was not easily determined, such as aerial shots of large groups or the backs of teachers’ heads, the photo’s racial or gender composition was coded as “can’t determine.” In cases where photos either contained a singular teacher or foregrounded a singular teacher, I coded if the teacher was a “single POC,” “single white,” “single male,” or “single female.” To determine race and gender, I use 44 phenotypical markers, such as skin color or hairstyle, and photo captions including titles and names of teachers, using the example of Hernández (2016). There are inherent shortcomings of using dichotomous variables and phenotypical markers as these may or may not reflect the individual’s identification with these groups and neglects the complexity of socially constructed categories of race and gender. Within the U.S. social context, there are some racial and gender visual cues that associate individuals with categories of race and gender and related stereotype, however inaccurate these cues may in representing individuals’ actual lived experiences and identities as a person. Hernández (2016) noted that visual cues, such as phenotypical markers of race and gender, are important for understanding when race and other identity groups are implicitly invoked, particularly in a society that emphasizes color-blindness. Thus, these codes assume that the average reader may associate teachers on strike with the identity groups represented by the images, and these codes were analyzed to see how characteristics of race and gender do and do not relate to social constructions. Final components of my coding scheme included three sets of descriptive codes: the role of people pictured, the strike pictured, and the strike discussed in the text. As whom is in each photo could impact how teachers are portrayed, I coded if the individuals pictured are teachers generally, union teachers, students, parents, political leaders, union leaders, non-teacher union members, or if the image does not include people at all. I also coded which strike each image shows, as some articles include images from a range of strikes across several locations. Similarly, I code the title of each article to indicate which strike(s) are discussed in the text. While an article may have been gathered while searching for the 2018 Oklahoma strike, for example, it may actually focus on the 2018 West Virginia or Kentucky strikes, the 2018 Arizona strike that followed the case study strikes, or a combination of all these strikes. In this way, I am 45 able to sort my data by the strike pictured or by the strike discussed to have a more accurate analysis of how each event was portrayed. Use of Qualitative Data Analysis Software. To facilitate coding and analysis with the large volume of data, I used Dedoose analytic software to store, manage, sort, and conduct preliminary analyses of image data codes (Bazeley, 2013). The use of media file descriptors for contextual factors along with the coding scheme allows for easier comparison among different comparison groups or slices of data. To ensure that my research remains embedded in the qualitative data, I engaged in a constant iterative process that included short analytic memos using features linking memos to codes in Dedoose to track early observations and rationale for coding decisions. I also tracked decisions related to data analysis in research notes and discussed choices regarding code applications with a research thought partner to be sure I could clearly articulate how and why I applied codes as I did. Reliability and Validity. I applied my coding scheme in ways that enhance both reliability and validity. I used the interrater reliability testing feature within Dedoose to calculate Cohen’s Kappa for a subset of the image data, both for intra- and inter-rater reliability. Additionally, the size of the dataset helps increase the confidence that findings are more likely to hold beyond the sample of articles selected, reducing potential sampling bias (Maxwell, 2013). Dedoose allows for images to be coded but does not currently allow for images to be used in their training functionality to calculate inter-rater reliability. To circumvent this challenge, I created a Microsoft Word document with a subset of images from the full dataset, representing a variety of sources and strikes. I included text labels for each image within this document using the image captions. I then uploaded the document to Dedoose and coded only the image labels with the codes applied during analysis. Both I and two outside thought partners coded the test 46 document manually, writing our selected codes on the document. I then compared the manual codes to the image labels and input those responses into the Dedoose training test to calculate Cohen’s Kappa for the Social Constructions, Strike Signs, Race, and Gender codes. The results indicate strong intra-rater reliability with a kappa of 0.90 and strong inter-rater reliability, with an overall kappa of 0.75 and 0.95 with each thought-partner. Text Computer-Assisted Content Analysis and Sentiment Analysis. To complement the qualitative content analysis of images and to enhance the trustworthiness of my analysis, I also conducted sentiment and content analysis using Linguistic Inquiry Word Count (LIWC) and Provalis Word Stat and QDA Miner software. Using pre-tested dictionaries considered to be valid measures of psychological constructs (Tausczik & Pennebaker, 2010), I utilized LIWC to conduct sentiment analysis measuring word frequencies of constructs related to deservingness and political power of article text. Because these measures of psychological constructs are related but not specifically tailored to the theoretical definitions of deservingness and political power, I used Word Stat software to create dictionaries of words invoking political power and deservingness that are aligned with Ingram et al.’s (2007) construct definitions. I then apply codes to paragraphs that include the identified words to analyze the presence of these constructs using QDA Miner, a qualitative coding analysis software that functions in tandem with Word Stat (Jones, 2013). Sentiment Analysis using LIWC. LIWC compares document text to user-defined dictionaries to generate word counts and percentages of dictionary words found in that document (Pennebaker, et al., 2015), and sentiment analysis using LIWC and similar programs has previously been applied to study social constructions. For example, Bleich & van der Veen (2018) and Bleich et al. (2018) use a compilation of eight lexica to assess the tone of news 47 coverage. If an article had a more positive tone, that was considered to indicate a more positive construction for the target group of study and vice versa. I analyzed the word count data for each article’s text for articles with at least one paragraph11 (N=617) using LIWC 2015’s included measures of emotional tone and positive and negative emotion words. LIWC measures tone on a scale from 0-100, with 0 being a very negative tone and 100 being a very positive tone, with the mid-point of 50 being a neutral tone and serving as the dividing line between positive and negative. I also consider the percent of the article word count that contains positive emotion words and negative emotion words. LIWC’s dictionaries of emotions words have been tested and determined to match with human ratings of emotions in writing (Alpers et al., 2005; Tausczik & Pennebaker, 2010). While LIWC has been applied to study the deservingness dynamic of social constructions, it has not been directly applied to study the dimension of political power. The 2015 version of LIWC’s dictionaries included a measure for “power” as a psychological construct. This included elements of hierarchies and status comparisons (Pennebaker et al., 2015). I used this pre-tested dictionary as one measure of the percent of power words within an article. In keeping with Bleich & van der Veen (2018) and Bleich et al. (2018b), I assumed that attitudes reflected in the word frequency measures attitudes towards the main subject, or target population, the event the article covers focuses on. In this study, this is teachers on strike. Thus, I contend that if an article has a higher percentage of power words that teachers, as the main subject of the strikes, are being portrayed as more powerful. 11 LIWC (Pennebaker Conglomerates, 2023) indicates that analysis of texts with less than 50 words should be treated with a degree of skepticism, as the measures of the constructs are more accurate in longer texts. As such, all included articles have at least a paragraph of data, resulting in a dataset of articles with a minimum word count of 140. 48 Computer-assisted content analysis with Word Stat and QDA Miner. While LIWC includes a measure of psychological power, this does not seem to fully capture political power, such as voting power. Likewise, there are likely words that capture deservingness or undeservingness that are specific to teachers as targets that may not be captured in LIWC’s sentiment analysis. To address the potential lack of alignment between the pre-tested dictionaries and the constructs of interest, I created social constructions dictionaries that identify words that indicate political power and deservingness. Following Deng et al. (2018), I used a systematic and iterative procedure to build dictionaries reflecting high political power, low political power, deserving, and undeserving constructions of teachers. The dictionary building process was facilitated by Word Stat 9 software, which calculates the frequency of use of words across a dataset (Provalis Research, 2021). Word Stat can automatically exclude common words such as articles (the, a, an, etc.) that are likely to occur frequently as an artifact of language conventions, providing a list of words used, their frequency, and a full listing of in-context use of the words. Words can then be sorted into user-created categories. Using Word Stat software, I generated a word and phrase frequency list from a randomly selected 20% of my text analysis dataset, ensuring that all strikes and news sources were represented in the sample. I selected every fifth article from an alphabetical list of my data files and checked that my final subset had every strike event represented by at least 5 articles and that each news source was represented by at least 10 articles. This resulted in a dataset of 128 articles. Conceptually, to create the dictionaries, I followed a three-step process. First, I deductively identified words I anticipated being related to the constructs of political power and deservingness. Then, I reviewed the most frequent words and phrases from a subsection of the 49 dataset to identify inductively words that may be related to the dimensions of social constructions. To ensure my dictionaries were valid, I then applied the dictionaries to the full dataset and checked each dictionary word’s use in context to ensure it belonged in the dictionary, removing words that were not consistently used in ways that suggested either political power or deservingness. I describe this process in detail below and provide the step-by-step iterations of the dictionary in Appendix C. I began by sorting words deductively. Before running Word Stat’s word frequency analysis, I used Ingram et al.’s (2007) definitions of deservingness and political power to create a list of words and their synonyms that would be expected to be associated with either of these dimensions. This is in alignment with LIWC’s initial dictionary building procedure (Tausczik & Pennebaker, 2010). While this list is useful as a starting point, it is important that any dictionary be aligned with the text being analyzed (Deng et al, 2018). Thus, rather than include all the deductively derived words in the concept dictionaries, I first checked that these words and their use in the context of my corpus of news article data were in alignment with Schneider and Ingram’s use of political power and deservingness. Deductive words were included in the first iteration of the dictionary if they were both (1) present in the word frequency list and (2) was consistently used in context in a way that aligned with the social constructions’ concepts’ definitions. For a word or phrase to be included in the final dictionary, it needed to relate to the concepts in context in at least 80% of the excerpts (paragraphs) it appeared in, in keeping with prior literature (Bengtson & Xu, 1995; Bezem, 2021). This resulted in 49 political power words or phrases and 16 deservingness words or phrases. My deductive procedure likely would not identify all related words, particularly those related to the absence of deservingness (undeservingness) and political power (low power) and 50 those words that may be related to ways to show deservingness and power in the context of a strike. I built upon this deductive based using an inductive word and phrase sorting procedure. I reviewed the full list of frequent words and phrases from the 20% data subset to identify any words or phrases that might be related to social constructions. I checked each of these words or phrases in context to ensure that they met the 80% fit threshold before including the linguistic element in a dictionary. This process resulted in 90 high power words or phrases, 23 deserving words or phrases, 24 undeserving words or phrases, and 1 low power word (“concessions”). Because so few words suggested a low level of or lack of political power, I conceptualized an article showing teachers as having low political power if it has less reference to words or phrases from the high political power dictionary and removed the low political power dictionary from the analysis. I also inductively created dictionaries of words that referenced key stakeholder groups – teachers, unions, children, parents, political leaders, and administrators – to assess if the presence of different groups in text was related to the presence of words related to particular constructions. I then used Word Stat to calculate the frequencies of occurrence of the dictionary words and phrases across the full text analysis dataset (N=617). As a validity check, I reviewed each dictionary word or phrase in context to ensure that the words continued to reflect the intended constructs in the full dataset. Words or phrases that did not meet the 80% threshold were removed from the dictionaries. This resulted in final dictionaries with 73 high power words and phrases, 14 deserving words and phrases, and 18 undeserving words and phrases (see Appendix C for the process and final dictionaries). Once dictionaries were finalized, I used Word Stat’s frequency function to identify all the paragraphs that used words or phrases from each of the dictionaries. I applied codes to all excerpts identified by Word Stat, using the dictionary name as the code. For example, 460 51 excerpts contained at least one word from the deservingness dictionary, and all these segments were coded with the label “Deserving.” These codes then were exported to QDA Miner 6, where I could compare code applications, code frequencies, and code co-occurrences across the dataset or in segments (Provalis Research, 2020). Much like Dedoose, QDA Miner allows for the association of descriptive variables with each article, so code applications could be compared by strike event, location characteristics, news sources, and related variables of interest. Analytic Procedure For both the image and text content analyses, I labeled my content data using either a coding scheme (image data) or dictionaries (text data) and then compared the application of these labels within and across strike event and news source comparison groups by creating summary tables, bar graphs, and line graphs. I describe the coding and linguistic analysis processes below in the Image Context Analysis Coding Scheme section and Text Computer-Assisted Content Analysis and Sentiment Analysis section. To understand how social constructions of teachers were conveyed within the context of each strike event, I compared frequencies of social construction code applications, code co-occurrences, and word frequencies across the 7 cases. I then compared how constructions differed among individual events, between the early statewide strikes (West Virginia 2018, Kentucky 2018, and Oklahoma 2018) and the later citywide strikes (Los Angeles 2019, Denver 2019, and Chicago 2019), over time by comparing the Chicago 2012 and 2019 strikes, and across the variation in teacher demographics in each location. To compare across demographics, I group strike events by if they occurred in a location where the proportion of white teachers compared to non-white teachers was greater than the national average (West Virginia 2018 and Kentucky 2018), similar to the national average (Oklahoma 2018 and Denver 2018), and less than the national average (Los Angeles 2019 and Chicago 2019). In addition to 52 considering how constructions varied by the racial demographics of a place, I also compared how teachers were portrayed based on the racial composition of the teachers pictured in photographs. I also compared how teachers were constructed across news sources, first comparing the individual news sources, then in groups by political lean (Left-leaning, Center, and Right-leaning sources), and by type of news source (newspaper, television network producing online news, or online news site). While much of this analytic comparative process was conducted separately for the image and text data, I also considered how the image and text data varied in how they portrayed teachers. The online news articles exist with both images and text appearing together, complementing each other, and carrying meanings in their combination. As such, I created matrices comparing each article’s image social construction code counts to dictionary word frequency measures for related social constructions words. Because of the nature of the differing types of data, some research questions best lent themselves to exploration through the image data alone. Through the analytic process, an emergent finding was that strike signs seemed to play an important role in how teachers were constructed in images. As a result, I began to question how teachers themselves, through the creation of and choosing to display messages via strike signs, shaped policy images of themselves as a group. While the media has editorial discretion and chooses what to publish, photographers cannot photograph something that is not present at the event. Strike signs were a way that teachers could convey messages about the profession, schools, their demands, and their “why” – why they teach and why they are on strike, that could then be broadly shared beyond the picket line in images from news coverage. To understand how teachers shaped social constructions about themselves as a target population, I analyzed how social constructions varied 53 by strike sign theme, and how signs and constructions varied based on event and news source. Limitations There are several limitations regarding this study. First, as the study uses qualitative methodologies and descriptive statistics in its analysis, the findings are not causal and thus this study does not represent a test of Schneider and Ingram’s (1993, 1997, 2019) theory of social constructions of target populations, but an application of the theory and its concepts to a specific context. While the study is does not employ causal methods, I seek to enhance its generalizability through thoughtful sampling and through the analysis of a large volume of data to help ensure that the themes identified are more likely to be representative of cases beyond those studied (Maxwell, 2013). Additionally, as this study focuses on a specific movement of teacher strikes, it is possible that the findings may not be generalizable to strike events beyond the context studied. Computer-assisted text analysis is a powerful tool to analyze large volumes of text in ways that would present capacity challenges for the unassisted human. However, even at today’s stage of advanced technology, there are limitations to what a computer can accurately identify, particularly the nuances of tone. LIWC software notes that sarcasm is particularly hard for linguistic dictionaries to accurately identify the tone of, for example (Pennebaker et al., 2015). Words also have multiple meanings, and thus sometimes the inclusion of a particular word in a sentence or paragraph may not be an indicator of the concept being studies (Pennebaker et al., 2015). To help account for this, I make two strategic choices in my analysis. The first is that I use both LIWC and self-created dictionaries in Word Stat software to balance the rigorous validity testing of LIWC with the concept precision of the inductively created dictionaries. Comparing the results of both analyses allows me to identify potential issues with the computer- 54 assisted analysis. Further, for the content analysis of the text conducted using the social constructions dictionaries using Word Stat and QDA Miner, I verified that all words that remained in the final dictionaries were related to the intended concepts in at least 80% of the uses of the word in context. By reviewing each word’s use in this way, I can account for errors of inclusion such as a word being used multiple ways and sarcastic uses of words. As such, I have created smaller, more conservative dictionaries that may not include fewer words that could be associated with the social construction concepts in favor of accuracy. Beyond methodological limitations, I also understand that this work focuses solely on the news media’s social constructions of teachers. This study does not consider how other stakeholders, such as the broader public or policymakers, perceived teachers during these strike events, nor does this study consider how sources beyond online line written and still images portrayed teachers, such as video data or social media data. These are important players in the policy process and worthy sources of data regarding social constructions for future studies. 55 CHAPTER 4: FINDINGS REGARDING OVERALL SOCIAL CONSTRUCTIONS OF TEACHERS ON STRIKE DURING THE RED FOR ED MOVEMENT During the Red for Ed strikes12 (articles N= 660, images N=1619), teachers were frequently portrayed in images as an advantaged target population – seen as both politically powerful and as worthy of policy benefits. More than half (54.23%, N=878) of the images from the coverage of the Red for Ed strikes (N=1619) portrayed teachers as deserving, and over two- thirds (69.92%, N=1132) of the images showed teachers as having strong political power. Only 14.82% (N=240) of images showed teachers as undeserving and 2.22% (N=36) of images showed teachers as having weak political power. Looking at when social construction codes co- occurred in the same image, deserving and high political power co-occurred the most often at 33.72% (N=546) of the time. None of the other deservingness and power combinations occur together in more than 10% of the images. To visualize how the images from the Red for Ed dataset as a whole fit along the spectra of deservingness and political power, I plotted the percentage of images displaying these two characteristics on a coordinate plane to coincide with Schneider & Ingram’s (2019) social constructions matrix (see Figure 2). I conceptualized an X-axis where the value -100 is equivalent to 100% of images portraying teachers as deserving and 0% of the images portrayed teachers as undeserving. A value of 100 is equivalent to 100% of images portraying teachers as undeserving and 0% of images portraying teachers as deserving. Likewise, the Y-axis value of 100 corresponds with 100% of images displaying teachers with high power and 0% of the images displaying teachers with low power, and the value of -100 corresponding to 100% of images displaying low power and 0% of the images displaying high power. I used this scale to 12 The Red for Ed strike dataset includes all articles gathered from the 2018 and 2019 case study strike events. This portion of the dataset excludes the data from the Chicago 2012 dataset (articles N=168, images N= 383). 56 calculate the difference between the percent deserving and undeserving and the percent high power and low power. These differences became a set of coordinates with a measure of deservingness and power that allows for a visual representation of teachers’ average social construction. Figure 2. Overall Social Construction of Teachers during the Red for Ed Strikes 100% 80% HIGH POLITICAL POWER Teachers during the Red for Ed Strikes 60% 40% ADVANTAGED 20% CONTENDERS 100% 80% 60% 40% 20% 20% 40% 60% 80% 100% 20% DEPENDENTS DEVIANTS LOW POLITICAL POWER 40% 60% 80% 100% DESERVING UNDESERVING 57 In addition to identifying the approximate levels of deservingness and political power portrayed by the news coverage of the Red for Ed strikes through the visualization, I also compared how often deservingness codes co-occurred with political power codes (see Table 8). I found that teachers were shown as both deserving and having high levels of political power, or an advantaged construction, most often, with the two constructs co-occurring in just over a third (33.72%, N= 546) of the Red for Ed images. Teachers were shown as contenders, having high power and less deservingness, in 8.03% (N=130) of the images. Media images portrayed teachers as having low power very infrequently, with less than 1% of images showing teachers as dependents or deviants. These results provide additional evidence that during the Red for Ed strikes, media portrayed teachers as an advantaged target group most often. Table 8. Images with Co-Occurring Social Constructions Codes from Red for Ed Strike Coverage (N=1619) Deserving Undeserving n % n % High Political Power 546 33.72 130 8.03 Low Political Power 2 0.12 13 0.80 The image data portrays deservingness in ways the article text does not. Content analysis of the text using the social constructions dictionaries indicates that there were nearly equal numbers of articles that used language that described teachers as deserving of policy benefits and those that described teachers as undeserving, with 53.91% (N=262) of articles using deserving language and 51.44% (N=250) of articles using undeserving language. This near equal weight to deserving and undeserving language indicates that media likely tries to present multiple viewpoints within the text, presenting the teachers’ justification for a strike with the 58 inconveniences and challenges posed by closing schools. Much like with the image data, political power of teachers was nearly ubiquitous across articles, with 97.33% of the Red for Ed articles showing political power. These trends in social constructions presence hold when considering when social constructions words co-occur in the same paragraph as the word “teacher” and when considering the full dataset, inclusive of the Chicago 2012 strike data. The full results of the word frequencies are summarized in a series of figures in Appendix D. The balance of deserving and undeserving text found in the text content analysis aligns with LIWC sentiment analysis. Using positive and negative tone as a corollary for deserving and undeserving constructions (e.g., Bleich et al., 2018; Bleich & van der Veen, 2018), the Red for Ed articles on average have a slightly negative tone, or slight tendency towards undeservingness, with a rating of 37.78 on a scale from 0-100, with 0 being very negative and 100 being a very positive tone. However, there is some variation masked by averages, as LIWC’s tone measure within the Red for Ed dataset ranged from 1.00 to 97.66. Figure 3 shows the distribution of article counts for LIWC tone measures, highlighting how article tone tended to be slightly negative, though overall clustering towards the middle rather than either end of the spectrum. These findings are in line with literature that finds news media has a tendency towards negativity (Bode et al., 2020; Dunaway, 2013; Farnsworth & Lichter, 2011; Lengauer et al., 2012) and media effort towards at least perceived objectivity (Comstock & Scharrer, 2005). Overall, emotion words occurred infrequently in the article text, with positive emotion words occurring on average in 1.94% of the article words and negative emotion words occurring in 1.33% of the article text on average. Much like with the content analysis using the social constructions dictionaries showed that words showing high levels of political power were most common, LIWC’s linguistic analysis finds that words invoking power occurred more frequently than those 59 evoking emotion, with articles about the Red for Ed strikes averaging power words in 7.07% of their total word count. Figure 3 Distribution of LIWC Tone Measures, Text Analysis Red for Ed Dataset (N=486 articles) 100 90 80 Number of Articles (N=486) 70 60 50 40 30 20 10 0 ] 0] 0] 0] 0] 0] 0] 0] 0] 0] 0] 0] ] .7 0 6 .4 4 .1 1 .8 9 .5 7 .2 4 .9 2 .6 0 .3 8 .0 5 .7 3 .4 .1 0 ,8 ,1 ,2 ,3 ,3 ,4 ,5 ,6 ,7 ,7 ,8 ,9 01 00 ,1 [1 . 70 .4 0 .1 0 .8 0 .5 0 .2 0 .9 0 .6 0 .3 0 .0 0 .7 0 .4 0 (8 . (1 6 (2 4 (3 1 (3 9 (4 7 (5 4 (6 2 (7 0 (7 8 (8 5 (9 3 LIWC Tone Measure (0-100, negative-positive) Considered together, the political power exhibited by the strike process was particularly salient in the media coverage of the Red for Ed movement. While Kreitzer and Smith (2018) used public perceptions to label teachers as a dependent target population, this analysis shows that there are some circumstances in which teachers can wield political influence. Indeed, within the context of this corpus of news articles, only one word (“concessions”) was identified as invoking low levels of political power in at least 80% of its uses, and so I was unable to measure low political power using text. High political power words were present in all but 13 articles and more than two-thirds of the images showed teachers leveraging their political power. This is evidence of strikes and protests as a particularly powerful form of political speech that frames 60 teachers as capable of exerting political pressure. While the text analysis shows teachers portrayed as deserving and undeserving at similar rates, the images selected to accompany the text convey more nuanced messages about teachers. The even distribution of deserving and undeserving words in the text coupled with the tendency towards a slight negative tone reflect the style of news coverage in American media: the appearance of objectivity with an overall negative tone. However, the images show a noticeably more deserving view of teachers, with more than 3.5xs as many images showing teachers as deserving than as undeserving. In a media landscape that is filled with eye-catching photos, soundbites, and quick, easy-to-digest snippets of text, pictures may truly be worth a thousand words, and analyzing the content of those images provides information missed by reviewing the text alone. Teachers as Deserving The majority of images (N=878) during the Red for Ed strikes portrayed teachers as deserving, often by either showing teachers as worthy of economic benefits or respect or by affiliating teachers with children and occasionally by highlighting the importance of teachers to society broadly as providers of a public good. Nearly a quarter (24.26%, N=213) of the images that showed teachers as deserving did not invoke any of these three themes. Many of these images showed teachers on strike who were happy or celebrating or indicated support for teachers. During the 2019 Chicago teachers’ strike several Democratic presidential candidates joined teachers on the picket line, indicating that they saw the teachers’ demands as worthwhile and the teachers as deserving. Worthy of Economic Benefits and Respect Teachers were shown as worthy of economic benefits or of respect in 41.57% (N=365) of 61 deserving images. The issues of respect and economics are uniquely entwined, as pay and benefits are often equated to the valuation of work, with work that is seen as more deserving (for example, work society views as necessary, important, or requiring unique or advanced qualification) being more worthy of respect. A strike sign from the 2019 Chicago teachers’ strike said, “OVERWORKED! UNDERPAID! NO RESPECT! FAIR CONTRACT NOW!” (Krzaczynski, 2019), emphasizing how teachers viewed pay, respect, and collective bargaining contracts that outline working conditions as interconnected concerns. Sometimes, this theme was shown by emphasizing the challenges of teachers’ finances. For example, an article included a photo of a smiling Denver math teacher at her desk with brightly colored teaching materials and the heading “She’s one car breakdown away from living in a camper van” (CNN, 2019b), suggesting that as a working professional this teacher deserves to be in a better financial situation where she is debating the trade-offs of paying for transportation or housing. The media portrayed teachers as deserving through images and stories that highlighted teachers who struggled financially on their teaching salary alone, with the implication that a full-time professional career, like teaching, should provide sufficient income to pay bills. Teachers also pushed this message through signs that captured their economic frustrations. Some examples from the strike in Oklahoma include signs reading, “My 2nd Job Paid for my Sign!” (CNN, 2018c) and “Teacher Needs School Supplies! Anything Helps” (CNN, 2018a). In addition to focusing on personal finances, some images included strike signs with messages indicating teachers deserve other employment benefits, such as a pension or a fair contract. Other images focused on respect generally without invoking money: “KENTUCKY Teachers Deserve RESPECT” (CNN, 2018d). Worthy because Children are Deserving Just over a third (33.94%, N= 298) of the images showing teachers as deserving aligned 62 teachers with children in some way. Some images showed children striking with teachers, often carrying signs indicating that students would benefit from the policy gains teachers were striking for. In Los Angeles, a young girl was pictured wearing red and holding a sign with the message, “WE’RE NOT SARDINES. LOWER CLASS SIZES” (Schnell, 2019). Other images show teachers using strike signs to indicate that they were advocating on behalf of their students, or commonly “On Strike for My Students” (e.g., Associated Press, 2019b). Beyond the explicit association with students via strike signs, the media also would include images of smiling teachers working with students in the classroom as part of their strike coverage, reinforcing the ways teachers benefit students. By equating teachers’ demands to students’ needs, these images reinforce an image of teachers as selfless children’s advocates, rather than as greedy or self- serving. Worthy because Education is a Public Good While not nearly as prevalent as images that emphasized the economic worthiness of teachers or children’s worthiness, some images showed teachers as worthy because they provide a public good that benefits society broadly. Of the deserving images, 14.24% (N=125) depicted ways that educators serve the public good. This particular message was conveyed almost exclusively through strike signs teachers were pictured with, and included messages such as “Invest in the Future” (Zalubowski, 2018), “CAN YOU READ THIS? YOU’RE WELCOME” (Davis, 2019), and “Think EDUCATION’S expensive? try IGNORANCE” (Franklin, 2017). Article Language Connects Themes of Deservingness Together The themes of teachers as worthy professionals, working for the betterment of their students and the public good, were often connected in the text to show how teachers were deserving of the policy benefits they were on strike for. News articles focused on difficult 63 financial choices teachers were forced to make, and how those choices impacted students. For example, an article from The Washington Post (Balingit, 2018) used a vignette to show how low pay incentives quality teachers to leave some school systems: “Oklahoma’s 2016 teacher of the year, Shawn Sheehan, decamped for Texas last year, joining many other teachers who sought higher-paying jobs.” The implication is the students of Oklahoma lost out on the opportunity to learn from an excellent teacher because of poor teacher pay. Other articles emphasized that teachers are worthy of respect beyond an appropriate paycheck, sometimes by quoting strike supporters. Collinson (2018) quotes a tweet from Senator Elizabeth Warren: “Teachers in OK, KY, WV & all across the country change the lives of little girls and boys everyday,” Warren wrote in a Twitter thread. “They deserve our respect & support. I stand in solidarity with those teachers who are fighting to support their families & invest in their students. Their fight is our fight.” Here the media uses quotes to justify that teachers deserve respect and economic benefits because of the benefits they confer to their students. Language in the articles also highlighted the unique position teachers’ play in preparing students for the workforce. As opinion columnist Krugman (2018) noted, So we’re left with a nation in which teachers, the people we count on to prepare our children for the future, are starting to feel like members of the working poor, unable to make ends meet unless they take second jobs. In addition to linking teachers’ deservingness to the benefits their students already receive, some articles note that the policy benefits that teachers were on strike for also benefit students, like: “Our students deserve smaller class sizes. They deserve nurses. The deserve social workers. They deserve bilingual educators,” Perales said. “This is what we’re asking for 64 at the table, and we’re willing to strike to get them for our students.” (Andone, 2019) The collective article text tells a story of teachers as deserving because they have been selfless for too long – describing teachers who do everything they can to help their students (even going on strike) despite being compensated poorly and facing difficult working conditions. Teachers as Undeserving The images of worthy teachers contrast with the smaller proportion (N=240) of images that show teachers as undeserving. Often, teachers were presented as less deserving when children were shown to be being negatively impacted by strikes, particularly being shown out of class. For example, sometimes political leaders on the opposite end of negotiations from teachers were pictured spending time with children who were not in school, such as Chicago Mayor Lori Lightfoot sitting with students at a local YMCA during the strike (Lott, 2019). Other images showed empty classrooms (e.g., Flynn, 2018), with the absence of teachers being a key indicator of being less deserving. Other images of striking teachers include captions or are placed under headlines like “What will happen to 600,000 students?” to emphasize the displacement caused by the strike (Dovarganes, 2019). Occasionally, public school teachers were specifically shown as less deserving by including images of charter schools still in session while traditional public schools were on strike. During the Los Angeles teachers’ strike, the Wall Street Journal ran an article highlighting that charter schools remained open, showing charter teachers smiling and engaged in small group instruction as students conducted experiments (Zaucha, 2019). Similarly, the text analysis reflected the theme of teachers as undeserving when the inconveniences of a teacher strike, especially for children, are highlighted. Ten of 18 undeserving dictionary words reflected closing schools in some way, all of which were added through an inductive review of the language used in the news articles. In an article with broad 65 coverage of the early portion of the Red for Ed movement, Arizona Governor Doug Ducey was quoted as tweeting, “No one wants to see teachers strike. If schools shut down, our kids are the ones who lose out” (Karimi & Sutton, 2018). Another article quoted a parent who said, “I know there has been some frustration just from what I have heard – a lot of working parents. It’s hard to find child care last minute” (Hanna & Weisfeldt, 2018). Kentucky Governor Matt Bevin’s viral quote – “I guarantee you somewhere in Kentucky today, a child was sexually assaulted that was left at home because there was nobody there to watch them” (McLaren & Aulbach, 2018) – featured in 10 articles (2.06% of the Red for Ed text analysis dataset) and was a particularly strong criticism of teachers. Teachers were also sometimes shown as angry while protesting, portraying a more aggressive image that could be seen as less deserving when compared with images of teachers smiling, embracing students and parents, and otherwise appearing peaceable. For example, teachers were shown yelling and “boo[ing]” with their eyebrows narrowed as they filled the Kentucky State Capitol (Greene, 2018), in contrast to teachers smiling and waving outside the West Virginia capitol (CNN, 2018b). Other images showed political leaders who were opposed to the strike, often with a quote emphasizing their position. The Washington Post shared a photograph of Education Secretary Betsy DeVos with the caption that “DeVos said striking teachers should ‘serve the students that are there to be served’” (Martinez Monsivals, 2018). Similarly, photos of Oklahoma Governor Mary Fallin accompanied her comment that teachers’ demands were “kind of like a teenager wanting a better car” (Fox News, 2018), and Kentucky Governor Matt Bevin’s photo was shown with captions referencing his insinuation that teachers were responsible for child abuse that occurred during sickouts (McLaren, 2018). By showing images of people with political power 66 denouncing teachers, the news outlets shared ways that teachers were seen as unworthy of the benefits they sought, both by being greedy – like a teenager wanting a better car – and by harming children. Importantly, how teachers are associated with children impacts how teachers are viewed as deserving or undeserving. When teachers are either associated with the benefits they already confer to students through their work or as advocating for students, they are viewed as deserving. However, when teachers are depicted as harming students by striking, typically through their absence, they appear less deserving of policy benefits themselves. When pictured with children, teachers were pictured as deserving in 82.91% of images (N=97) and undeserving in only 10.26% of images (N=12). When students were pictured without teachers, teachers were depicted as undeserving in half of the images (50.98%, N=52). Positively associating with children, particularly as advocates and supporters of children, allows teachers to also affiliate with some of the traditional social constructions of children as deserving. Much like associating with children in a positive way allows teachers to adopt perceptions of deservingness, associating with unions seems to affiliate teachers with less deserving policy images. Teachers who visually identified with unions were depicted as undeserving more frequently than teachers who were not visually affiliated with unions. Nearly 18% (17.91%, N=60) of the images showing union teachers (N=335) were labeled undeserving, compared to 8.79% (N=81) of images showing non-union teachers (N=921). Who teachers align themselves with also relates to the perceived level of deservingness of themselves as a target group. Teachers as Politically Powerful As strikes are one way for teachers to display political power, it should not be surprising 67 that news coverage of teacher strikes largely shows teachers as having strong political power with 69.92% (N=1132) of images from the Red for Ed strikes showing high political power. Per Ingram, et al.’s (2007) definition of political power, one of the ways a group can display political resources is by being united and easy to mobilize. Teachers looked unified and organized most frequently when associating with the Red for Ed movement (N=775) or when associating with a teachers’ union (N=378). Interestingly, 23.23% of the images that conveyed teachers as having high political power showed teachers associating with both the Red for Ed movement and a teachers’ union, so associating with the Red for Ed movement did not necessarily result in teachers disassociating with unions. Approaching half (45.23%, N=512) of the images showing high political power for teachers did this by exclusively affiliating with the Red for Ed movement and not with teachers’ unions. The West Virginia teachers’ strike, which was the first of the wave of 2018 and 2019 strikes, showed unity by wearing red shirts often printed with messages such as “WV Teachers United [with an outline of the state]” or “55 United [with a state map of the 55 school districts]” (Hudson, 2018). Teachers striking in Arizona in 2018 used the #RedforEd commonly in mass- produced and other strike signs (e.g., York, 2018), and while the Arizona strike was not selected as a case study, photos from this strike often appeared in coverage of other teacher strikes as the news outlets tried to convey the scope of teacher protests nationally. As a result, signs with “#RedforEd” were commonly found in coverage of the teacher strikes, even if a particular strike event did not use the slogan itself. This became the name and phrase associated with the teachers’ movement, and teachers and their supports evoked the movement by wearing red, a tactic seen across the events. The text analysis using the social constructions dictionaries also highlights the emphasis 68 on the Red for Ed movement. “RedforEd” and “movement” were among the most frequently occurring words in the high political power dictionary, with “RedforEd” and similar words (e.g., “Red for Ed”) occurring 157 times and “movement” (and “movements”) occurring 309 times. Only the term “walkout” occurred more frequently from the high political power dictionary, showing how important the concept of the political movement was to conceptualizing teacher power (see Appendix D for frequencies of all dictionary words). Similarly, I find that when high political power dictionary words occurred within a paragraph, words referencing teachers also occurred in 72.4% (N=2439) of those paragraphs but words referencing unions only co-occurred with high political power words in 23.3% (N= 784) of paragraphs with high power words. The prevalence of language discussing teachers, rather than teachers’ unions, as powerful indicates that in the case of the 2018 and 2019 strikes much of the political power was derived from the political movement rather than from the organization of the unions. About a tenth (10.16%, N=115) of the images showing teachers as high power only associated teachers with a union and not with the Red for Ed movement, with most of these images coming from the Los Angeles and Chicago teacher strikes. These urban cities have strong teachers’ union presence, and the unions played a central role in organizing the strike activities. Often, these images were video stills or photographs of union leadership discussing the strike, such as United Teachers Los Angeles president Alex Caputo-Pearl addressing the press after the Los Angeles’ strike was called off (Reuters, 2019). Other images that captured teachers having political power through their affiliation with the union showed teachers holding printed signs with union logos and basic slogans, like those produced by the Chicago Teachers Union that read, “ON STRIKE for MY STUDENTS” and “FAIR CONTRACT NOW!,” and sometimes also showed teachers receiving the support of other unions, including the Service Employees 69 International Union (SEIU) that joined Chicago Teachers on picket lines with signs in their organization’s purple and yellow colors (Gress, 2019). The text also emphasized the union’s demands. For example, a Wall Street Journal article characterized the conflict in the Chicago 2019 strike as between Mayor Lori Lightfoot and the union: “Ms. Lightfoot said Wednesday that some of the union’s demands can’t be met because of funding, but that talks will go on. She also said that the union has continued to make demands during the negotiation process” (Hobbs, 2019). One author characterized one of the key issues of the Los Angeles teachers’ strike, charter regulation, as a “war between union leaders and charter school supporters” (Schallhorn, 2018), another example of using language to invoke unions as a target group rather than teachers. Words referencing unions appeared in most frequently in the articles from the Los Angeles (N=772 paragraphs) and Chicago 2019 (N=558) strikes, compared to other case studies. The next highest occurrence of paragraphs with union language was in the articles covering the West Virginia 2018 strike, with 328 coded paragraphs. Of the paragraphs including union dictionary words, 12.06% were from the Los Angeles coverage and 8.72% were from the Chicago 2019 coverage, compared to 5.12% from West Virginia. Likewise, 15.23% of paragraphs coded from the Los Angeles data and 16.36% of Chicago 2019 data included union language. Overall, we see that references to unions and related words were more common in the locations with high levels of union power. Images affiliating teachers with union power emphasized teachers less. A third of the images that only showcased power through unions, and not the Red for Ed movement, did not include teachers in them (33.04%, N=38). Conversely, images that exclusively pictured power through the Red for Ed movement almost all included teachers (97.27%, N=498), with only 2.73% of images not picturing teachers. The media images more clearly associated teachers with 70 the Red for Ed movement. While there was still a strong association between teachers and teachers’ unions, about a third of the time unions were shown as powerful without teachers. This suggests teachers and teachers’ unions are not synonymous target populations, and that the perceptions of the unions are not always the perceptions of teachers. Roughly a fifth (21.38%, N=242) of images showing teachers as high power did not associate teachers with either the Red for Ed movement or teachers’ unions. Typically, these images showed a large number of people on strike, often zoomed out to capture the size of the crowd (e.g., Davidson, 2019). In many of these images, it was difficult to determine what strike signs read, if people were wearing similar clothing, or other distinguishing features that may represent a united effort. Instead, these images capture Ingram et al.’s (2007) description of a powerful group as large. The effect of these images is two-fold. It suggests there are a large number of people in support of the teachers’ policy demands and suggests the impact of such a large work stoppage could be significant, both ways teachers can show their political power. Teachers as Politically Weak Few of the images from the Red for Ed teacher strikes depicted teachers as politically weak. Only 36 images (2.22% of the total images from the Red for Ed events) showed teachers as lacking political power. Typically, images that showed teachers as low power did this by showing the political leaders or administrators the teachers were bargaining with as powerful. For example, a video of teachers striking in Chicago was accompanied by the title “Mayor to teachers: Chicago won’t budge on money,” suggesting that teachers did not have the power to force the city to make significant changes (Associated Press, 2019a). Some images more subtly attributed power to political leaders over teachers by having political leaders as the key sources for updates on the strike’s status instead of teachers or union representatives. The Washington 71 Post used a photo of Governor Jim Justice at a press conference to accompany an article announcing the end of the West Virginia teacher strike (Raby, 2018). Notably, in 33 of the 36 images (91.67%) showing teachers as having weak political power, teachers were absent from the photos. Emphasizing how uncommon it was for news coverage to show teachers as lacking political power when on strike, the deductive and inductive review of word frequencies used to create the social constructions dictionaries only yielded one word that related to low political power in at least 80% of its occurrences – “concessions.” Because this word alone does not represent the concept of low political power, I eliminated the low political power dictionary from the text analysis. It is worth noting that only 13 articles from the Red for Ed time period did not contain words from the high political power dictionary, suggesting that even an absence of political power words was uncommon. Overall, teachers engaging in strikes were not described as politically weak, so strikes and protests are an opportunity for teachers to assert their power collectively as a target population. Relating Images and Text Data to Each Other Throughout this study, I primarily consider how the image and text data from news articles separately relate to social constructions of teachers. However, the images and text are published together and function jointly as an information source. News articles with images coded as having high political power or as deserving also tended contain words from the high political power and deserving dictionaries. Of those articles used in the text analysis that also contained images showing high political power (N=416), nearly all (98.80%, N=411) of the articles contained words from the high political power dictionary. The ubiquity of teachers being presented as having strong political power is evident across the media data. The deserving and undeserving social constructions dictionaries show some alignment 72 with the occurrence of deserving and undeserving images, though not nearly as strong of a relationship. The majority of articles with deserving images (N=343) also had text with words from the deserving dictionary (58.02%, N=199 articles). Articles without deserving images were less likely to include deserving language (42.66% or 61 of the 143 articles without a deserving photo). Initially, having an image that showed teachers as undeserving seemed to not correlate with article text that includes language that suggests teachers are undeserving. Less than half (48.89%, N=66) of the articles with undeserving images also had undeserving text, but slightly more than half (51.85%, N=182) of the images without an undeserving image had undeserving text. In both cases, articles had deserving text at about the same rate (with an undeserving image- 55.56%, N=75; without an undeserving image- 52.71%, N=185). However, some articles have both images that show deservingness and images showing undeservingness. To understand if the presence of image social constructions collectively were related to the presence of corresponding language about teachers, I considered how the deserving images alone, undeserving images alone, both deserving and undeserving images together, and neither deserving nor undeserving images relate to the article text content (see Table 9). Across all image conditions, the text contained deserving and undeserving language in roughly half of the articles, though there was a slight decrease in number of articles with deserving language when deserving photos were not present. Looking at the number of paragraphs, rather than articles, with deserving and undeserving language, allows us to consider how often language was used across the corpus in case articles with particular combinations of images used language of deservingness more frequently within their text. Paragraphs that occurred in articles with deserving photos had deserving language more frequently and paragraphs that occurred in articles without deserving photos had undeserving language more frequently (see Table 10). Overall, this provides evidence 73 that there is moderate alignment with the social constructions dictionaries and image coding schemes. Table 9. Comparison of Deservingness Image Codes to Text Analysis Social Constructions Dictionaries by Articles, Red for Ed articles in text analysis (N=486) Text Analysis Image Codes Co-Occurrence Dictionary Word Both Deserving Deserving Code Undeserving Neither Presence and Undeserving Present, Code Present, Deserving nor Codes Present Undeserving Deserving Code Undeserving Code Not Present Not Present Codes Present n % n % n % n % Deserving 51 63.75 148 56.27 24 43.64 37 42.05 Words Present Undeserving 42 52.5 138 52.47 24 43.64 44 50.00 Words Present Total Articles 80 263 55 88 In addition to enhancing the validity of the analysis, this also provides evidence that qualitative image content analysis provides insight beyond computer-assisted content analysis of the text alone when considering news media. The lack of variation in occurrence of deserving and undeserving words suggests that the style of news articles lends itself to presenting “both sides” of an issue, resulting in less clear portrayals of deservingness as worthiness is related to societal value-judgements. This aligns with Hess and Martin’s (2018) finding that titles of major newspapers’ articles regarding the 2018 teachers’ strikes largely had an impartial bias in their tone. The images, however, showed more variation in deservingness, with considerably more images showing teachers as deserving compared to undeserving. This shows how the adage “a picture is worth a thousand words” can be meaningful in the case of understanding subtle 74 messages about target groups in news media. The text alone suggests more of an ambiguous, neutral level of deservingness towards teachers, but the images provide strong evidence of teachers being presented as an advantaged target population during the Red for Ed strikes. Given the high volume of news information available to many Americans, images seen when skimming news articles may be particularly impactful. The evidence of a more positive construction of teachers in the images could have a greater impact on public opinion than the more neutral article text. Table 10. Comparison of Deservingness Image Codes to Text Analysis Social Constructions Dictionaries by Paragraph, Red for Ed articles in text analysis (N=486) Text Analysis Image Codes Co-Occurrence Dictionary Word Both Deserving Deserving Code Undeserving Neither Presence and Undeserving Present, Code Present, Deserving nor Codes Present Undeserving Deserving Code Undeserving Code Not Present Not Present Codes Present n % n % n % n % Deserving 107 8.37% 263 7.17% 38 6.71% 50 5.64% Language Present Undeserving 88 6.89% 247 6.73% 43 7.60% 67 7.55% Language Present Total 1278 3670 566 887 Paragraphs 75 CHAPTER 5: FINDINGS COMPARING SOCIAL CONSTRUCTIONS ACROSS STRIKE EVENT CASES When the data is disaggregated by strike event, there are similar trends in social constructions of teachers, with the media portraying teachers from all six Red for Ed strike events as an advantaged target population (see Figure 4). However, while all the events tended towards a more advantaged portrayal of teachers, there was variation among the constructions by strike event, particularly along the dimension of deservingness. Figure 4. Comparison of Social Constructions of Teachers by Strike Event 100% KY 18 DEN 19 80% WV 18 HIGH POLITICAL POWER CHI 19 ALL STRIKES LA 19 60% OK 18 40% ADVANTAGED 20% CONTENDERS 100% 80% 60% 40% 20% 20% 40% 60% 80% 100% 20% DEPENDENTS DEVIANTS LOW POLITICAL POWER 40% 60% 80% 100% DESERVING UNDESERVING 76 Differences in Social Constructions based on Strike Event Teachers were portrayed as deserving of policy benefits in more than half of the images of three strike events: West Virginia 2018, Oklahoma 2018, and Denver 2019. Images of these three events tended to emphasize that teachers were worthy of economic benefits or professional respect, with more than a third of the images of those strikes showing teachers as deserving in this way. For example, for both the Oklahoma and the Denver strikes, CNN ran feature articles that highlighted personal stories from teachers struggling financially, working multiple jobs, and seeking alternative income opportunities alongside photos of teachers at their second job or in their classrooms (Yan & Stix, 2018; Zdanowicz, 2019). Conversely, the three strikes where teachers were less frequently portrayed as deserving – Kentucky 2018, Los Angeles 2019, and Chicago 2019 – had images showing teachers as economically worthy or deserving of professional respect in less than a quarter of the media images, with images of both the Los Angeles and the Chicago strike showing teachers in this way in only roughly five percent of images (5.46%, N=16 and 4.60%, N=11, respectively). A similar, mirrored trend occurs with teachers being as undeserving. The teachers in West Virginia were portrayed as undeserving the least often, in only 3.90% of that strike’s images. The West Virginia strike was the first in the movement, so the public interest and support was likely higher than later in the movement when fatigue from the negative impacts of teacher strikes can reduce support. Also, West Virginia’s teacher pay was ranked 48th in the country (Vera, 2018), so the teachers call for increased compensation and health care benefits could feel particularly worthy in comparison to other states. Teachers in Oklahoma, Kentucky, and Denver were portrayed as undeserving roughly equally with just above 10% of each strike’s images showing teachers as unworthy of benefits. Over a quarter of images from the Los Angeles and the Chicago 2019 strike events, however, 77 depicted teachers as undeserving of policy benefits. Los Angeles and Chicago are locations that have high levels of union power (NCES, 2018; Winkler et al., 2012), and, particularly in Chicago, a history of strike events (Berlin, 2019). These are also more urban locations, and unlike Denver (another urban-citywide strike), Los Angeles and Chicago both have teacher populations with a higher proportion of the workforce identifying as non-white (U.S. Census Bureau, 2022). The image data shows nuanced differences where strikes in more urban areas, with more diverse teaching workforces, and higher levels of union power are portrayed more negatively by news media than strikes held in statewide contexts. Figure 5. Comparison of Average LIWC Tone Measures across Strike Events As seen in the overall data, the text analysis across strikes shows a more neutral construction of teachers. LIWC sentiment analysis rates the average tone of news articles as similar across all six strikes, with measures ranging from 31.80 (Kentucky 2018) to 42.47 (Denver 2019) on the 0-100, negative to positive tone scale (see Figure 5). The Kentucky 78 sickouts faced particularly stiff political opposition from Governor Matt Bevin and his negative comments received considerable media attention, which could be driving the more negative tone of the Kentucky news articles. Consistent with the image data, the Denver strike were typically viewed more positively. In addition to demographic differences in Denver and the other citywide case study strikes, it is important to note that the strike in Denver was short (3 days) compared to other events. Analysis of the full dataset showed that when children are shown as being harmed by the strike, teachers are shown as undeserving. A shorter strike is also a shorter pause in children’s education and care. In this case, Denver’s short strike could have allowed more focus on the ways teachers were deserving of the policy benefits they sought, instead of focusing on how children were harmed and families inconvenienced by the strike. Figure 6. Comparison of Average LIWC Emotion and Power Percent Word Counts by Strike Event 8 7.40 7.20 7.08 7.15 7.15 7.04 Average Percent of Article Word Count 7 6.68 6 5 4 3 1.98 2.14 1.97 1.99 1.87 1.89 1.87 2 1.68 1.49 1.22 1.21 1.35 1.15 1.12 1 0 WV 18 KY 18 OK 18 LA 19 DEN 19 CHI 19 Overall Strike Event Positive Emotions Negative Emotions Power While the overall tone of the articles is slightly negative, word count analysis of positive and negative emotion words shows that both are sparingly – roughly 1-2% of words in articles 79 were either positive or negative emotion words – and that for all six strikes positive emotion words were slightly more common than negative emotion words (see Figure 6). These trends were mostly consistent with the results from the social constructions dictionary-based content analysis (see Table 11), with more articles including language that from the deserving dictionary than from the undeserving dictionary across all six strikes. Statewide strikes had a larger proportion of their articles use deserving words than citywide strikes. While there are observable differences in deservingness in the text using the social constructions dictionaries, it is important to note that these differences are small. When looking at the difference in the number of paragraphs coded as deserving or undeserving for each strike, the largest difference was in the Los Angeles teacher strike coverage, where 27 more paragraphs contained deserving language than undeserving language across all 100 news articles analyzed. The smallest difference was found in the coverage of the Kentucky strike (5 paragraphs more were coded undeserving) and the Oklahoma strike (5 paragraphs more were coded deserving), where each strike’s coverage differed in the number of times terms from each deservingness dictionary was applied by only 5 paragraphs. Images showing high political power were more common across strikes, with at least 60% of the images from all six Red for Ed case study events portraying teachers as having strong political power. Likewise, the content analysis of the text finds that nearly all articles across all six events used high political power language (see Table 11) and similar frequencies of LIWC power words (see Figure 5). The Chicago 2019 strike coverage in particular included images of teachers as politically powerful, with over 80% of the images of that strike event showing teachers as politically powerful, and 80% of those images showing high power associating teachers with the teachers’ union. Similarly, nearly 75% of the images coded as high power from 80 the Los Angeles strike also associated teachers with teachers’ unions. Of the Red for Ed strike images coded as having high levels of political power and as associating with unions, 80.95% (N=306) of those images are from these two urban city strikes. The Los Angeles and Chicago 2019 cases also are in locations with a greater teachers’ union presence and where the teaching force is proportionately less white than the national average. These two strikes are also less frequently portrayed as deserving and more frequently portrayed as undeserving compared to the other case study Red for Ed strikes. Table 11. Summary of Article Counts of Applications of Social Constructions Text Dictionaries by Strike Event (N=486) Social Constructions Number of Articles containing Words from Dictionary Dictionary WV 18 KY 18 OK 18 LA 19 DEN 19 CHI 19 n % n % n % n % n % n % High Political 83 96.51 83 100 103 97.17 96 96.00 45 97.83 63 96.92 Power Deserving 45 52.33 54 65.06 64 60.38 48 48.00 20 43.48 29 44.62 Undeserving 52 60.74 52 62.65 60 56.60 36 36.00 20 43.48 28 43.08 Total Articles 86 83 106 100 46 65 It is worth noting that the theory of social constructions of target populations contend that policymakers design policies to reflect social constructions. If this were to hold, it would be expected that the strikes with the most deserving and most powerful constructions would likely result in the most benefits. As teachers were constructed as advantaged across all the strike 81 events, the theory predicts that teachers would receive more policy benefits and fewer policy burdens (see Figure 1). However, two strikes – Oklahoma and Kentucky 2018 – did have some policy losses. Teachers in Oklahoma walked out seeking a pay raise and additional education funding. While the message of economic worthiness seemed to resonate in the news coverage (see Chapter 4), it seemed to have less of an effect on policymakers. The Oklahoma state legislature passed a tax increase to both increase both teacher pay and school funding (Loewus, 2019), but they did so before the teacher walkout began and did not give additional concessions. The Kentucky 2018 sickouts similarly did not result in blocking the pension reform package teachers targeted in their protests (Loewus, 2019). This points to some of the complexities of understanding social constructions and the political agenda setting process. While the news media can be a potent force and is an important stakeholder in the policy process, the media alone does not dictate the agenda or the perceptions of policymakers. Given that both of these locations had Republican-controlled executive branches and legislatures (Ballotpedia, 2022f, 2022g, 2022h, 2022i) that did not support additional spending for education, it is also important to consider how partisan politics and political parties shape policymakers views of target populations. Differences in Social Constructions based on the Racial Composition of Images The differences in social construction evident in the two case study locations with a lower proportion of white teachers in their teaching work force allude to some of the trends in construction by the racial composition of the images: that images featuring white teachers showed teachers as deserving more often than images featuring teachers of color. First, it is worth noting that across strike events, there are eight times as many images that show a majority of white teachers compared to images that show a majority of teachers of color. This likely 82 reflects both the disproportionate number of white teachers compare to teachers of color and potentially some biases in perceptions of who is a teacher. Of the 560 images that show a majority of white teachers, 62.86% of these images show teachers as deserving. In comparison, 52.86% of the images showing a majority of teachers of color (N=70) show teachers as deserving. Images that consisted of a majority of white teachers tended to emphasize teachers being deserving of professional respect or economic benefits more often, while images of where the majority of teachers presented as people of color tended to emphasize striking for children (see Table 12). Images with a majority of white teachers also portrayed teachers as undeserving proportionately less frequently (8.75% of 560 images) than images with a majority of teachers of color (18.57% of 70 images). While images with a majority of white teachers and images with a majority of teachers of color showed teachers as having high political power at nearly the same proportionate frequency (roughly 90% of each coding category of image racial compositions showed teachers as having high political power), images with a majority of white teachers showed teachers has having high political power because of their affiliation with a teachers’ union far less often (16.61% of images with a majority white teachers) than those with a majority of teachers of color (61.43%). While there are differences in social constructions by the racial composition of the teachers pictured, these differences also align with variation in other location contexts. For example, the strikes where teachers were portrayed as more deserving – West Virginia, Oklahoma, and Denver – were also strikes where less than one percent of the media images showed groups with a majority of teachers of color. These locations all also have teaching populations with either an average (Denver) or higher than average (West Virginia and Oklahoma) percent of their teaching workforce identifying as white. Because so many of the 83 images from these strikes portrayed teachers as deserving, and there were so few images that portrayed teachers of color from these events, these particular events may skew the aggregate results so that it appears that white teachers are more deserving. As the racial demographics of the teaching force impact the demographics of which teachers it is possible to photograph, it is more difficult to determine if white teachers are being pictured as more deserving because of racial characteristics and stereotypes or because some characteristic of the events in places where white teachers are more prevalent resulted in more deserving depictions. The overall impression, however, remains the same. Teachers of color are depicted as deserving less frequently than white teachers. Table 12. Percent of Red for Ed Strike Images Coded for Teacher Racial Composition Co-Occurring with Social Construction Codes Social Constructions Racial Composition of Teachers <50% of >50% of Diverse Racial Teachers White Teachers White Composition Deserving 52.86% 62.86% 44.35% Associate with Children 24.29% 18.21% 11.29% Education as a Public Good 8.57% 15.00% 2.42% Worthy of Respect/Economic 17.14% 28.21% 8.87% Benefits Undeserving 18.57% 8.75% 23.39% High Power 91.43% 89.46% 94.35% Associate with #RedforEd 65.71% 65.54% 75.81% Associate with Union 61.43% 16.61% 81.45% Low Power 0.00% 0.54% 0.00% Total Images (N) 70 560 124 84 Looking at the two strike locations where teachers of color were more represented in the media images – Los Angeles and Chicago 2019 – complicates some of the associations between race and constructions. Images from both of these urban, city strikes showed teachers as deserving less frequently and undeserving more frequently compared to other strike locations. These two strikes also more frequently showed teachers affiliating with union power. However, when looking at just these two strike events, there are only small differences in the proportions of images that show teachers as politically powerful if the image shows a majority of white teachers (90.80%), a majority of teachers of color (98.08%), or a racially diverse group (94.78%). In regards to deservingness, for the Chicago 2019 a greater proportion of the photographs showing a majority of teachers of color showed teachers as deserving compared to photographs showing a majority of white teachers (51.43%, N=18 of the 35 images with a majority teachers of color showed teachers as deserving; 23.21%, N=13 of the 56 images with a majority white teachers showed teachers as deserving). Images with a majority white teachers also showed teachers as undeserving (N=9) slightly more often than images with a majority teachers of color (N=5). This relationship between race of teachers in the image and deservingness measures is unique to Chicago. While the images from the Los Angeles teacher strike follow the pattern of the majority of the data set – images with a majority of white teachers showed teachers as deserving more frequently than images with a majority of teachers of color (18 images both deserving and with majority white teachers vs. 8 images both deserving and majority teachers of color) and vice versa – the differences between frequency of occurrences is considerably smaller. This suggests that in urban areas with a greater representation of teachers of color, there is less difference in how teachers are portrayed by racial characteristics. This also provides evidence that the less-deserving constructions of the Los Angeles and Chicago teacher 85 strikes may be driven by factors beyond race, such as union power. It is difficult to determine precisely how perceptions of race impact social constructions of teachers. There is a complex interplay of other contextual factors, including the relative power of the teachers’ union, the urbanicity of the place, and the representation of teachers of varying racial backgrounds in the news coverage, that all seem to relate to how teachers on strike are viewed. Because these characteristics are endogenous and influence each other, it is important to consider that associations of social constructions of teachers with the racial composition of images are not causal and are likely explained by a combination of contextual characteristics. While this analysis does not determine which individual factor has the greatest impact on social constructions of teachers, it does show clearly that there are differences in social constructions between teachers in states that are more rural and suburban contexts with higher proportions of their teaching force identifying as white and teachers in urban cities with smaller proportions of their teaching force identifying as white. Inherently, place and demographic characteristics are linked, and so the social constructions are associated with characteristics of both place and race. Differences in Social Constructions of Teachers on Strike over Time One possibility is that the construction of teachers on strike as an advantaged population is linked to the association with the Red for Ed movement and the particular moment in history. To better understand how the Red for Ed movement is influencing social constructions and potential policy images of teachers, I compare the coverage of the 2019 Chicago teacher strike to the 2012 Chicago teacher strike that predates the Red for Ed movement. First, it is worth noting that the 2012 strike received more media coverage than the 2019 strike by six of eight news sources (Breitbart and USA Today had more articles and images covering the 2019 strike). One potential explanation is that the 2012 strike was considered more newsworthy as a rare strike 86 event when compared with the 2019 strike, coming at the end of over a year of teacher strikes. Teachers during the 2019 strike are portrayed as deserving with almost the same frequency (30.54% of images, N=239) as they were in the 2012 strike (29.97% of images, N= 377), though they are portrayed as undeserving slightly less often. This is consistent with the text analysis data, that shows nearly equal proportions of news articles with deserving language in both the 2012 and 2019 strikes (45.80% of 2012 articles and 44.62% of 2019 articles), but slightly less use of undeserving words in the coverage of the 2019 strike, with only 43.08% of articles containing undeserving words in the 2019 coverage compared to 54.96% of articles in 2012. A quarter of images (25.94%, N= 239) from the 2019 strike showed teachers as undeserving compared to a third of images portraying teachers as undeserving from the 2012 strike (33.42%, N=377). Interestingly, the media portrayed teachers as deserving in the same ways and in similar proportions in both 2012 and 2019, with the most common way to show teachers as deserving being associating them with children, with themes of teachers as worthy of respect or economic benefit and highlighting education as a public good appearing less frequently. It is also worth noting that images of both Chicago strikes had the lowest proportions of images with deserving portrayals of teachers and the highest proportions of images with undeserving portrayals. It is possible that the consistent levels of deservingness may be because Chicago teachers could be viewed as one of the least deserving groups of teachers, and that if there was another recent strike in an area with Red for Ed strikes that comparing strikes over time in a different location may yield different results. However, one of the reasons that the Red for Ed strikes were so newsworthy was because they spread across states where teacher strikes were not just uncommon, but often illegal, making it difficult to compare constructions in other cases. 87 Noticeably different is how often teachers were portrayed as having high levels of political power. During the 2012 Chicago teachers’ strike, 62.54% of images showed teachers having high political power, compared to 81.17% of images from the 2019 strike. Results from the text analysis also find that high political power words were only applied to 86.26% of the 2012 articles, compared to 96.92% of the 2019 articles. During the 2012 strikes, there were more images of political leaders, parents, and even Tony Danza (e.g., CNN, 2012) with captions relating to these other stakeholders’ views of the teachers’ strike. Teachers were included in photographs less frequently in the 2012 strike coverage than in the 2019 strike coverage or in the dataset as a whole, with only 66.84% of the 2012 images (N=377) showing teachers and 77.82% of images from the 2019 strike (N=239) including teachers (similar to the 75.22% of all images, N=2002). As over 90% of the images associated with strong political power for teachers pictured teachers, including images of teachers less frequently and also including on other stakeholders in coverage in 2012 seems to have limited the frequency at which teachers were shown to have political power in this earlier strike. The news media’s editorial discretion to focus on multiple stakeholders and not just teachers impacted the overall collective social construction of teachers during this event. The coverage of the later strike, coming at the end of the Red for Ed movement, focused more on teachers. The coverage of the 2012 and 2019 Chicago teacher strikes provide evidence to place teachers on the border between advantaged and contenders in Schneider and Ingram’s (2019) typology. Using the differences in proportions of deserving and undeserving, high power and low power code applications visualization strategy, the Chicago 2019 strike falls barely over the midpoint of deservingness to the advantaged category and the Chicago 2012 strike is just over the midpoint the other direction in the contender category (see Figure 7). The difference between 88 them, however, is minimal, with most other locations varying noticeably more from each other in terms of deservingness (Figure 4). When looking at when codes co-occurred in the same image, deserving and high power codes (advantaged construction) occurred together more frequently in the 2019 coverage compared to the 2012 coverage, but the undeserving and high power codes (contender construction) occurred together more frequently as well. As a result, it is difficult to conclusive say that the social constructions of teachers in Chicago differed greatly from 2012 to 2019, but there are slight differences in how media constructed teachers that suggest there were changes over time, and these may be related to the Red for Ed movement. Figure 7. Comparison of Social Constructions of Teachers over Time, Chicago Strikes 100% CHI 2019 80% HIGH POLITICAL POWER CHI 2012 ALL STRIKES 60% 40% ADVANTAGED 20% CONTENDERS 100% 80% 60% 40% 20% 20% 40% 60% 80% 100% 20% DEPENDENTS DEVIANTS LOW POLITICAL POWER 40% 60% 80% 100% DESERVING UNDESERVING 89 CHAPTER 6: FINDINGS COMPARING SOCIAL CONSTRUCTIONS BY NEWS SOURCE Much like social constructions of teachers varied subtly by strike event, social constructions varied based on the news source. Overall, most news sources’ collective coverage of the Red for Ed teacher strikes portrayed teachers as an advantaged target population, both worthy of policy benefits and politically powerful. More than 60% of the images from each of the eight news sources showed teachers as having high political power, and less than 10% of each source’s images showed teachers as politically weak. Differences in Social Constructions based on Political Lean of Sources For all sources except Breitbart, images that portrayed teachers as deserving appeared in a larger proportion of the news sources’ images than images that portrayed teachers as undeserving. Breitbart portrayed teachers as contenders in their news coverage, a less-deserving construction than other sources. Using the coordinate plane visualization process outlined in Chapter 4, we can notice the relative similarity of constructions of teachers across news sources and the subtle differences within those portrayals (see Figure 8). These relative placements of the social constructions of teachers by each news source align with the frequency of co-occurrence of deservingness and political power codes within the same image (see Table 13). Seven of eight sources (all except Breitbart) had images with the deserving and high-power codes occurring together most often (representing an advantaged construction). The next most frequent construction of teachers was that of the contender, with images coded as both undeserving and high power occurring the most frequently in Breitbart’s coverage and with the second highest frequency in all other news sources’ coverage. Because images showed teachers as low power so infrequently, the low power code co-occurred with both 90 deservingness codes less than 2% of the time for all sources. As with the coordinate plane visualization using individual code frequencies to approximate the social construction of teachers, Breitbart also had the undeserving and high political power codes co-occur together most frequently (23.53% of Breitbart images with a social construction code, N=12). Figure 8. Comparison of Social Constructions of Teachers by News Source 100% Slate Breitbart USA Today 80% Washington Post Fox News HIGH POLITICAL POWER ALL SOURCES CNN 60% New York Times Wall Street Journal 40% ADVANTAGED 20% CONTENDERS 100% 80% 60% 40% 20% 20% 60% 100% 40% 80% 20% DEPENDENTS DEVIANTS LOW POLITICAL POWER 40% 60% 80% 100% DESERVING UNDESERVING 91 Table 13. Comparison of Social Construction Code Co-Occurrences by News Source for Red for Ed Strikes (N=1,619) News Source Social Construction Categories Advantaged: Contender: Dependent: Deviant: Deserving + Undeserving + Deserving + Undeserving + High Power High Power Low Power Low Power n % n % n % n % Breitbart 3 5.88 12 23.53 0 0 0 0 Fox News 28 40.58 10 14.49 0 0 1 1.45 Wall Street Journal 18 20.93 7 8.14 0 0 1 1.16 USA Today 215 36.01 34 5.70 0 0 8 1.34 Washington Post 72 42.11 10 5.85 2 1.17 2 1.17 New York Times 74 41.11 13 7.22 0 0 1 0.56 CNN 129 31.16 43 10.39 0 0 0 0 Slate 6 46.16 1 7.69 0 0 0 0 All Sources 546 34.47 130 8.22 2 0.13 13 0.82 There was a larger degree of variation by news source in the frequency of portrayals of deservingness than in portrayals of high political power. Breitbart’s coverage, in particular, portrayed teachers as undeserving of policy benefits, with teachers portrayed as undeserving in 32.08% of its images (N=53) and as deserving in only 9.43% of its images. Undeserving images within the Breitbart data often included images of angry teachers yelling and appearing aggressive. Breitbart also used images with handmade signs from teachers expressing their reasons for striking less frequently than almost all other sources. Handmade signs appeared in 28.81% of strike images in Breitbart coverage (only The Wall Street Journal included images of handmade signs less frequently, in only 20.30% of its images). Conversely, 65% of the photos included by Slate had handmade strike signs, which contained homemade strike signs with 92 messages like “We [heart] our students” and “I work to make a living. I teach to make a difference.” All the photos with handmade signs included by Slate showed teachers overall as deserving. Overall, Left-leaning news sources portrayed teachers as deserving more often. All four Left and Left-leaning sources (CNN, New York Times, Slate, Washington Post) portrayed teachers as deserving more often than the full set of images from the Red for Ed (54.23% of images showed teachers as deserving, N=1,619), while the Center (USA Today, Wall Street Journal) and Right-leaning (Breitbart, Fox News) sources portrayed teachers as deserving proportionately less frequently than the full Red for Ed sample. While Left-leaning sources’ images were coded as deserving at relatively similar rates, how teachers were pictured as deserving differed by source. The Washington Post and CNN portrayed teachers as deserving because they were worthy of respect or economic benefits most frequently. In addition to photographs with teachers holding signs calling out low pay or articles spotlighting the financial struggles of teachers, The Washington Post and CNN emphasized the economic worthiness of the teachers’ demands by including graphs and charts often comparing teachers’ wages overtime and relative to other similar professions (e.g., Merrill, 2018; The Washington Post, 2019). Conversely, The New York Times and Slate included more images that showed teachers as advocating for students and students’ needs. Often, teachers associated themselves with students’ needs through strike signs messaging or students aligned themselves with teachers by joining picket lines, but the news media also chose to run feature articles highlighting school conditions that directly impact students. For example, The New York Times published an article with eleven photos of dilapidated school supplies and facilities with images of teachers trying to make improvements on their own, without district support (Sedgwick, 2018). A summary of the 93 frequency of social construction code applications relative to the total number of images for each source can be found in Table 14. Table 14. Summary of Frequency of Code Applications of Social Constructions Codes of Red for Ed Strikes, N=1619 News Source Social Construction Categories Total Deserving Undeserving High Power Low Power n % n % n % n % n Breitbart 5 9.43 17 32.08 44 83.02 0 0 53 Fox News 39 53.42 16 21.92 52 71.23 1 1.37 73 Wall Street Journal 31 34.44 20 22.22 56 62.22 7 7.78 90 USA Today 299 49.18 80 13.16 464 76.32 17 2.80 608 Washington Post 118 68.21 15 8.67 119 68.79 5 2.89 173 New York Times 110 58.82 31 16.58 127 67.91 6 3.21 187 CNN 268 63.66 60 14.25 259 61.52 0 0 421 Slate 8 57.14 1 7.14 11 78.57 0 0 14 All Sources 878 54.23 240 14.82 1132 69.92 36 2.22 1619 Both Right-leaning news sources (Breitbart, Fox News) and one Center-bias source (The Wall Street Journal) portrayed teachers as undeserving more than 20% of the time. These sources emphasized the political power of policymakers at odds with the teachers. However, they also highlighted the stereotype of the “militant” teachers’ union that stems from strikes in New York and Newark in the 1960s-1970s that were particularly disruptive, intertwined with complex debates of the Civil Rights Movement, and occasionally became violent (Goldstein, 2014), often through depictions of raised fists, a political symbol that has been associated with unity, but also socialism, social justice and in particular fighting racism, strength, and force (Joseph, 2016; Korff & Peterson, 1992). Breitbart’s coverage took the subtle association a step further by 94 Table 15. Summary of Article Counts of Applications of Social Constructions Text Dictionaries by News Source (N=486) Social Constructions Number of Articles containing Words from Dictionary Dictionary Slate CNN NYT WP USA WSJ Fox Breitbart n % n % n % n % n % n % n % n % High Political 12 92.31 95 97.94 69 100 95 96.94 89 97.80 48 97.96 39 100 26 86.67 Power Deserving 11 84.62 51 52.58 45 65.22 55 56.12 53 58.24 22 44.90 12 30.77 11 36.67 Undeserving 11 84.62 45 46.39 41 59.42 50 51.02 53 58.24 19 38.78 18 46.15 11 36.67 Total Articles 13 97 69 98 91 49 39 30 95 labeling strike organizers as “socialists” and in particular highlighting the role of teacher Noah Karvelis, an Arizona teacher involved in spearheading the movement in his state, and his links to a “Marxist magazine” (Leahy, 2019). The attempts to link teachers to socialists was a clear effort to frame them negatively. Though more subtly, the text analysis supports the trends in deservingness found in the image analysis. Left-leaning news sources (CNN, New York Times, Slate, Washington Post) and one of the center-bias sources (USA Today) used words that invoked deserving constructions in most of their news articles (see Table 15). Right-leaning sources Breitbart and Fox News used words associated with undeserving constructions in at least as many news articles as used words with deserving constructions, suggesting a slightly negative portrayal of teachers compared to Left-leaning sources. Figure 9. Comparison of Average LIWC Emotion and Power Percent Word Counts by News Source 96 Figure 10. Comparison of Average LIWC Tone Measures across News Sources Similar trends to the text analysis of deservingness were found in the LIWC word frequency analysis of positive and negative emotion words (see Figure 9) and tone (see Figure 10). Variation in the LIWC word frequencies and tone measures are overall small, with the differences in the average percent of word count of different types of emotion words differing by less than 1% across all news sources. USA Today, as a center-bias source, might be expected to have a slightly less deserving construction of teachers than both the text and image data may suggest. However, while in 2019 USA Today was ranked as tending towards the center by AllSides Media Bias Ratings, in later years this has shifted to a Left-leaning bias rating since 2021(AllSides, 2023). It is possible that the coverage of the Red for Ed movement may be an example of how USA Today’s coverage has shifted over time. USA Today, unlike the other news sources, also owns many local newspapers, including in Oklahoma and Kentucky, so covered those strikes in greater detail. As coverage of statewide strikes generally portrayed teachers as more deserving, this could be upwardly biasing the average tone of the articles and frequency of 97 deserving constructions. Likewise, The Wall Street Journal’s coverage’s proportion of deserving and undeserving dictionary words is more similar to Right-leaning news sources despite its center-bias rating. While the news coverage for The Wall Street Journal is rated as center-biased, the opinion section is rated Right-leaning and has been since 2018 (AllSides, 2023). Thus, the coverage from The Wall Street Journal, which includes opinion pieces and news pieces, likely reflects this slight Right-leaning tendency in the less deserving portrayal of teachers. It seems counterintuitive that both Breitbart and Slate both used deserving and underserving words in equal proportions in their news coverage (see Table 15). However, as online-only news sources, rather than television network media companies or newspapers, these sources have unique use of voice. Breitbart, for example, has a sarcastic writing style. LIWC text analysis notes that computer-assisted language analysis cannot differentiate sarcasm from other writing styles, so measures particularly of positive words may be skewed upward for this source. For example, one Breitbart article noted “[f]ailed Democrat presidential candidate and U.S. Sen. Bernie Sanders of Vermont has tweeted that he stands in ‘solidarity’ with the Los Angeles teachers’ union” (Berry, 2019). The word “solidarity” is included in the deserving dictionary, as in more than 80% of the times it was used in context it referred to teachers in a deserving way. The use of quotation marks, a relatively common stylistic choice of Breitbart, helps to indicate a sarcastic and negative tone. The inclusion of “[f]ailed Democrat presidential candidate” to describe Senator Bernie Sanders also serves to discredit his view of the striking teachers. Slate, on the other hand, often qualifies statements that could portray teachers as undeserving, essentially softening the blow and potentially skewing measures of undeservingness upward for this source. A Slate article begins one paragraph with the sentence, “More than 22,000 West Virginia teachers have been on strike since Thursday, and every public school in the state is 98 closed” (Mathis-Lilley, 2018). While this sentence emphasizes schools being “closed” (a word included in the undeserving dictionary because of the potential negative impacts of closures), the paragraph goes on to justify the strike by explaining that West Virginia teacher pay ranked 48th in the nation in one ranking and was considerably lower than neighboring states. The difference in writing style and tone for the online news sources deviates from traditional newspaper articles’ objective approach. As a result, the measures of deservingness for Breitbart and Slate likely do not fully reflect the intended tone of the news sources. Across all eight sources, teachers were portrayed as having high political power, with the online-only news sources Breitbart and Slate’s images showing political power in roughly 80% of their images. While these online-only sources overall published fewer articles and images, they had a clear and consistent message about teachers, albeit with opposite views on if teachers are deserving of policy benefits. Breitbart also associated teachers with teachers’ unions when depicting political power more frequently than any other source. The other seven sources more commonly depicted political power through associating teachers with the Red for Ed movement. The high levels of political power in news articles were also consistently found in the text analysis using social constructions dictionaries (see. Table 15). Center-biased and Right-leaning news sources tended to have slightly higher average percent word counts using LIWC’s measure of power than Left-leaning news sources (see Figure 10). Differences in Social Construction by News Source by Event Context In addition to being differences generally in how news sources portrayed teachers on strike, there were also differences by news source in how particular strikes were portrayed. Notably, Breitbart and The Wall Street Journal portrayed teachers as contenders during the citywide 2019 Red for Ed Strikes (Los Angeles, Denver, and Chicago 2019 strikes). Fox News 99 approaches a neutral dimension of deservingness, with only one more image showing teachers as deserving than images showing teachers as undeserving during the citywide strikes. Across the board, the coverage of the statewide strikes from 2018 (West Virginia, Kentucky, and Oklahoma) portrayed teachers as deserving more frequently. Sources were split on whether the statewide strikes or citywide strikes were associated with teachers appearing politically powerful more or less often, but all sources had a higher percentage of their images showing teachers as politically strong over politically weak. The New York Times, USA Today, and The Wall Street Journal all published images of teachers appearing politically powerful more frequently during the statewide strikes, while CNN, Fox News, and The Washington Post published more images of teachers as politically powerful more frequently during the citywide strikes. Figures 11 and 12 show the relative social construction of teachers by each of the news sources, comparing the statewide strikes to the citywide strikes.13 Overall, I find that partisan biases do seem to be related to differing constructions of teachers. Left and Left-leaning news sources tend to portray teachers as deserving of policy benefits more often in their coverage of the Red for Ed strikes. Right-leaning news sources, and Breitbart in particular, portrayed teachers as undeserving more frequently. While there was less variation in the proportion of sources’ images that portrayed teachers as having high levels of political power, Right-leaning and Center-bias tended to show teachers as having political power slightly more often and did this through association with teachers’ unions more frequently than Left-leaning sources. 13 Because some strikes are covered less by some sources, I decided that there needed to be at least 10 total images to analyze a particular source’s coverage of a particular strike. This precluded some strikes and some sources from this cross-characteristic analysis. As a result, comparisons by strike event were conducted by groups of strikes sharing similar characteristics: statewide strikes from 2018 where unions have less political power compared to citywide strikes from 2019 where unions have more political power. Even within these groups, there was minimal coverage of statewide strikes by Breitbart and citywide strikes by Slate, and so coverage of these groups was not included in comparisons. 100 Figure 11. Comparison of Social Constructions of Teachers participating in Citywide Strikes in 2019 by News Source 100% Breitbart Washington Post CNN 80% Fox News ALL SOURCES HIGH POLITICAL POWER USA Today 60% New York Times Wall Street Journal 40% ADVANTAGED 20% CONTENDERS 100% 80% 60% 40% 20% 20% 40% 60% 80% 100% 20% DEPENDENTS DEVIANTS LOW POLITICAL POWER 40% 60% 80% 100% DESERVING UNDESERVING 101 Figure 12. Comparison of Social Constructions of Teachers participating in Statewide Strikes in 2018 by News Source 100% USA Today Slate 80% ALL SOURCES Washington Post HIGH POLITICAL POWER Fox News 60% New York… CNN Wall Street Journal 40% ADVANTAGED 20% CONTENDERS 100% 80% 60% 40% 20% 20% 40% 60% 80% 100% 20% DEPENDENTS DEVIANTS LOW POLITICAL POWER 40% 60% 80% 100% DESERVING UNDESERVING 102 CHAPTER 7: FINDINGS REGARDING TEACHER AGENCY IN INFLUENCING SOCIAL CONSTRUCTIONS VIA STRIKE SIGNS The media has editorial discretion over what images they choose to publish, however, the media is constrained by the context of the event – they cannot publish images of something that did not happen (with the exception of digital alteration). Thus, how teachers protest, specifically what they wear, what they say, how the act, and in the signs they carry, is a part of the images and thus the overall portrayal of themselves as a group that the teachers control. Strike signs can be particularly powerful as this provides a way for teachers to thoughtfully frame their views on policy issues. The media cannot change the message written on their signs, so this is a way teachers can influence what issues are being seen and how they are being discussed. This is a key element of agenda-seeding (Wasow, 2020), or using the media attention from a protest event to elevate issues from less powerful target groups to the policy agenda. In addition to providing media attention, teacher strikes require a response from policymakers, as typically there are policy demands that need to be negotiated on in order for teachers to call an end to the strike. Thoughtful framing of those policy demands in the media influences what the public knows about the issues and likely how the public views strikes and teachers by proxy. To understand how teachers helped shape the media portrayal of themselves, I analyzed the strike sign themes that appeared in the article images. Across the Red for Ed strikes, 964 images (59.54%) included visible strike signs. Of those images with signs, nearly three-quarters (73.34%, N=707) included handmade signs, meaning that teachers crafted the messages on the signs, rather than a union or other group printing the signs en masse. Handmade signs highlight the teacher agency in shaping messages about themselves and the strike. Many of these signs were creative, visually appealing, and even 103 witty, making the signs interesting. Table 16. Summary of Strike Sign Themes in Red for Ed Strike Images (N=964) Sign Theme Image Counts and Percent Strike Signs Handmade (Percent of images with signs) (Percent of signs with strike theme) n % n % Invoke Children 380 39.41 251 66.05 Invoke Union 329 34.13 139 42.25 Invoke Economics 438 45.43 352 80.37 Invoke Societal Needs 197 20.44 183 92.89 Invoke Fairness 145 15.04 103 71.03 Invoke Frustration 285 29.56 266 93.33 Six strike sign themes were identified inductively from the images: signs invoked children, unions, economic concerns, societal needs, fairness, and frustration (see Table 16). As noted in the analysis of the overall social constructions of teachers (see Chapter 4), teachers tend to associate with two traditionally well-defined target populations: children and unions. These themes occurred at similar frequencies, with signs invoking children occurring in 39.42% (N=380) of the strike sign images and signs invoking unions occurring in 34.13% of strike sign images (N=329). Two thirds (66.05%, N=251) of the images with strike signs invoking children included handmade signs, while handmade signs were found in only 42.25% of images with signs invoking unions. This highlights that many union signs were mass-produced by the unions themselves. These signs often carried simple messages – “ON STRIKE for BETTER SCHOOLS” or “ON STRIKE for a FAIR CONTRACT” – in bold black letters bordered by union logos (Wood, 2019). Some of mass-produced union signs did invoke children, commonly “ON STRIKE for MY STUDENTS” (Boyle, 2019), however, in general teachers were sharing 104 more substantive messages to associate themselves with children. Signs included messages saying, “You Can’t Put Students First If You Put Teachers Last,” (Ciaglo, 2019), “I TEACH MY STUDENTS HOW TO USE THEIR VOICES…NOW I’M USING MINE!” (Wallace, 2018), and “FIGHTING FOR THE FUTURE OF OUR KIDS” (Crisp, 2018a). While the message is similar to the simple “striking for students” signs, these handmade signs use more emotional language to convey that teachers needs are aligned with children’s needs. The four other sign themes arose largely from the handmade signs, highlighting the role of teacher autonomy in shaping the messaging. Two themes appeared almost exclusively with handmade signs: invoking frustration (93.33%, N=266) and invoking societal needs (92.89%, 183). Teachers expressed frustration sometimes through calling out political leaders in their signs. In Oklahoma, a teacher attending a rally at the Capitol carried a sign reading, “MY STUDENTS KNOW WHAT INTEGRITY IS, DO YOU?!” (Carter, 2018), while in Kentucky a teacher held a sign saying, “BEVIN BETTA HAVE MY $$” with images referencing a popular song. Teachers also carried signs reminding politicians that they were also voters, with either the explicit or implied threat that teachers rallied together could vote policymakers out of office if they did not comply with demands. The tone of frustration in many ways complimented signs that emphasized how teachers benefit society. While these signs lack the slightly negative or attacking tone that some of the signs expressing frustration carried, pointing out how teachers benefit society broadly often conveyed a sense that teachers felt under appreciated. Common sign messages showcasing how teachers address societal needs included, “Teaching made your Job Possible” (Orgocki, 2018) and “CAN YOU READ THIS? YOU’RE WELCOME” (CNN, 2019a). Other signs focused on the impact to the future of a particular state, city, or their students, implying that the teachers were asking for policies that were necessary for future 105 societal prosperity. Another common theme that was strong associated with the Red for Ed strikes was economic worthiness and need, with this theme occurring most frequently (N=438, 45.43% of images with strike signs) and the majority of the time conveyed via handmade signs (80.37% of signs invoking economics appeared in images with handmade signs). While some mass-produced signs invoked themes of economic issues by focusing on contracts, the handmade signs were more personalized. For example, a photograph of teachers picketing in Denver prior to the strike showed one high school teacher holding a sign made with black marker and the side of a cardboard box that read, “I shouldn’t have to worry about Lesson Plans AND being able to PAY RENT” (Sangosti, 2019). Messages about not being able to afford housing or similar needs on a teachers’ salary were more poignant than “ON STRIKE for a fair contract” union-produced signs from Chicago’s 2012 strike (Green, 2012). This theme of economic deservingness was particularly emphasized in the early Red for Ed strikes – West Virginia, Oklahoma, and Kentucky – where almost all images of strike signs included handmade signs. Teachers could drive some of the messaging by referencing personal financial challenges, and this messaging was picked up by the news media that spotlighted teachers’ stories of economic hardship in feature articles (e.g., Weir, 2018). While this is not causal evidence, teachers were making and displaying signs with messages highlighting their economic challenges and worthiness of increased pay and funding before news stories began highlighting their struggles. Fairness is a concept closely linked to deservingness and worthiness. If something is unjust or unfair, those who are wronged may be perceived as worthy of policy benefits that would correct or at least address in some ways the inequity. Fairness or a lack of respect was invoked in 15.04% of strike sign images, with over two-thirds of those images including 106 handmade signs. As noted, union-made signs often referenced “fair contracts” (Wood, 2019) and “Unfair Labor Practices” (Rezin Garcia, 2019) but again the handmade signs used more emotional words to convey a similar sentiment that teachers deserved to be treated fairly. Signs included messages like, “KEEP YOUR PROMISES #PROTECT PERA #Teachers’LivesMatter” (Weisfeldt, 2018), suggesting that it was unfair to teachers to takeaway retirement (or other) benefits that past policy had promised. More poignant were signs that teachers deserved greater respect such as, “I AM SMART. I AM KIND. I AM IMPORTANT. (apple images) I AM A TEACHER” (Strupp, 2018), a play on affirmations for a child in a recent movie, emphasizing that teachers are important and suggesting that policy or society have not treated them that way. Sign Themes Connections to Social Constructions Notably, these themes regularly intermingled, and together told the story of teachers feeling neglected, economically disadvantaged, and frustrated, despite the work they did for both children and the benefit of society. This story became the Red for Ed movement’s common narrative, reinforced by teachers’ handmade signs and personalized messages. Throughout the images with signs, two associations between sign messages and social constructions emerged: (1) deservingness was often associated with signs invoking economics and signs invoking children and (2) political power was often associated with signs invoking unions and frustration, in addition to economics and children (see Table 17). 107 Table 17. Co-Occurrence of Strike Sign Themes and Social Constructions Codes in Images with Strike Signs (N=964) Social Constructions Strike Sign Theme Codes Codes Invoke Invoke Invoke Societal Invoke Invoke Union Invoke Fairness Children Economics Needs Frustration n % n % n % n % n % n % Deserving 245 25.41 129 13.38 288 29.88 157 16.29 90 9.34 155 16.08 Undeserving 42 4.36 55 5.71 36 3.73 13 1.35 12 1.24 32 3.32 High Political 326 33.82 321 33.30 385 39.94 160 16.60 131 13.59 264 27.39 Power Low Political 1 0.10 1 0.10 0 0 0 0 0 0 1 0.10 Power 108 Strike Sign Themes and Deservingness Strike signs invoking both economics and children co-occurred with teachers being depicted as deserving in more than a quarter of the images with strike signs (29.88%, N=288 and 25.41%, N=245, respectively). Different strike events seemed to be driving these two associations and provide two separate justifications for teachers being worthy of policy benefits. Images with strike signs invoking economic themes and showing teachers as deserving were frequent in the coverage of the West Virginia 2018, Oklahoma 2018, and Denver 2018. These three events were also the three with the most deserving constructions of teachers (see Figure 4) overall. A co-occurrence between a social construction and a strike sign theme was considered frequent if the two codes co-occurred in at least 25% of the images with strike signs for a particular strike event. The coverage from the Oklahoma strike, in particular, associated signs with economic themes and deservingness together, with over half of the images with strike signs including both codes (55.41%, N=87). Across all three strikes, this association was driven largely by handmade signs, as over 98% of the images with strike signs included handmade signs. This emphasizes the role teachers had in emphasizing the story of being worthy of economic benefit and how this is linked with frequent portrayals of deservingness. Signs invoking children were also commonly associated with deservingness, especially among the citywide strikes – Los Angeles, Denver, and Chicago 2019. Strike signs invoking children and constructions of deservingness most frequently associated together in Denver (41.43% of event strike sign images, N= 29) and least frequently in Chicago (21.02% of event strike sign images, N=37). This follows the pattern found in Figure 4, where I find Denver teachers to be portrayed as deserving most often, followed by Los Angeles teachers and then Chicago teachers. Notably, the Los Angeles and Chicago strike events also have a high 109 prevalence of union signs, including mass-produced signs that invoke children such as the ubiquitous but simple “ON STRIKE FOR OUR STUDENTS” (e.g., Associated Press, 2019b; Brown, 2019). The majority of strike sign images in Los Angeles that included both deserving codes and invoked children in signs were also handmade (59.70%, N=40) and nearly half of the signs in Chicago had all three codes co-occurring (48.65%, N=18), providing evidence that a handmade message, even within the same theme, may help teachers appear more deserving. It is also worth noting the mediating role associating with unions seems to have on deservingness. The two strikes with high levels of union power, Los Angeles and Chicago 2019, also had signs referencing unions most often, with 70.05% of strike sign images in Los Angeles invoking unions (N=131) and 78.41% of strike sign images in Chicago (N=138). This is a notable difference from the other strike events, where signs invoking unions occurred in less than 20% of the events’ strike sign images. The most common strike sign theme to be associated with undeservingness both within strike events and overall was signs invoking unions. In Los Angeles and Chicago, more than two-thirds of images with signs invoking children also had signs invoking unions. Signs that invoke children are generally associated with deserving constructions, but when those signs are also invoke unions they are more likely to portray teachers as undeserving. For example, across the Los Angeles strike event coverage, 15 images included signs invoking children that were associated with undeserving constructions of teachers. Ten of those 15 images also include signs that invoke unions, and all but two of the images with signs that invoke children and unions do so with the same sign – ON STRIKE FOR OUR STUDENTS with a United Teachers of Los Angeles union logo. The Chicago 2019 coverage tells a similar story, with 16 strike sign images both including signs that invoke children and show teachers as undeserving and 9 of those images also including signs that invoke unions. In 110 fact, while teachers appeared deserving 64.47% of the time when an image contained signs associating with children, these signs were the third most likely theme to be associated with constructions of undeservingness. This seems to be driven by the tendency of unions to explicitly associate their demands with children’s needs, rather than teachers. There is a difference in perceptions of deservingness when a sign is handcrafted and appears to be from an individual teacher compared to when a similar message is shared with a mass-produced sign with a union logo. Teachers’ unions have been criticized as putting adult interests before children (Finn & Petrilli, 2013; Moe, 2001), so it is possible that unions associating with children does not resonate as authentic in the way a handmade sign from an individual teacher does with perceptions of deservingness. The case of the Kentucky 2018 strike provides interesting variation in strike sign themes and association with deservingness. While overall the images of strike signs invoked frustration (59.41% of event strike sign images, N=60) and economic themes most frequently (46.56%, N=47), signs that invoked how teachers and education benefit society were most frequently associated with deservingness (co-occurring in 29.70% of event strike images, N=30). The societal benefits theme was also the only strike sign theme to co-occur with teacher deservingness in more than a quarter of strike sign images. Overall, teachers in the Kentucky strike event were constructed as deserving less frequently than teachers in the other statewide strikes. This may be in part because while signs focusing on benefits to society and the public good functions of teaching were most commonly associated with deservingness (79.70% of signs invoking societal benefits co-occurred with the deserving code), the argument that teachers deserved greater economic benefits was stronger or more clearly associated with deservingness. In the Kentucky strikes, signs invoking economics and signs invoking society appeared almost 111 equally as frequently (46.53% and 43.53% of event images with strike signs, respectively), while in the West Virginia and Oklahoma strikes that were otherwise similar along many characteristics, signs invoking economic themes occurred noticeably more frequently14. Also likely contributing to the slightly less positive construction of Kentucky teachers was the focus on strike signs with messages suggesting teachers were frustrated. More than 10% (11.23%) of the total images with signs that invoked frustration were also images where teachers were constructed as undeserving, the second most frequent strike sign theme associated with undeserving images after association with unions. The high prevalence of signs invoking frustration in strike sign images in Kentucky (59.41%) may have resulted in a more negative portrayal of teachers. This suggests that there is a balance point in considering how the tone of strike signs influence how deserving a target group is portrayed; while some frustration is warranted and can be associated with arguments of economic deservingness that were typically associated with teachers being seen as deserving, too much focus on frustration seems to be less likely to be viewed as deserving. Strike Sign Themes and Political Power Overall, strike signs co-occurred in images that showed teachers as politically powerful in just over half (52.38%, N=848) of the images of the Red for Ed strikes. This makes intuitive sense as one way of visually showing political power is by striking, and it is logical that many photos of strikes would include strike signs. The presence of the signs, rather than what they say, is indicative of strong political power. However, some message themes were common in these images. Of the Red for Ed images where strike signs invoked unions, 97.57% (N=321) were 14 In the coverage of the Oklahoma strike events, signs that invoked economics occurred in 64.33% of the event strike sign images while signs invoking societal benefits occurred in 46.50% of these images. In West Virginia, this gap was larger as signs invoking societal benefits were infrequently pictured. Of the images with strike event signs, 36.43% referenced economic themes and 2.33% referenced benefits to society. 112 associated with high political power. Interestingly, this theme occurred only slightly more frequently in images that showed high political power through association with unions (78.12%, N=257) than in images where high political power was shown through association with the Red for Ed movement (73.25%, N=230). While the majority (84.80% of Red for Ed strike signs with the invoking union theme, N=279) of strike signs invoking unions were from urban, citywide strike events coordinated by the local union, these strikes also closely affiliated themselves with the Red for Ed movement. Notably, there was relatively little variation in portrayals of teacher political power across strikes, regardless of whether teachers affiliated with unions or not (see Figure 4). This suggests that affiliation with the Red for Ed movement was a way to show political power, with or without the backing of the teachers’ unions. It is also possible that the unions strategically borrowed the power and deservingness of the Red for Ed movement and intentionally played up the association to build on the momentum of the earlier, grassroots events. Images that showed teachers and having strong political power and affiliating with the Red for Ed movement most frequently (16.92%, N=274) included strike signs with economic themes. Themes of benefits to society and invoking frustration were also more commonly associated with Red for Ed political power (6.18% and 11.49% of Red for Ed images, respectively), rather than union political power (1.54% and 2.72%, respectively). The strike signs provided a narrative for the movement: despite all the ways teachers and education benefit children and society broadly, teachers have been economically disenfranchised and they are frustrated about it, so they are demanding policy change. In addition to providing a clear narrative, the fact that this is more clearly associated with the political movement rather than unions provides evidence that teachers as a target population can be constructed differently that 113 teachers’ unions as organizations and target populations themselves. Images with handmade strike signs also co-occurred in images that showed political power through the Red for Ed movement more frequently (26.31% of Red for Ed images, N=426) than they co-occurred in images showing political power through unions (8.09% of Red for Ed images, N=131). If teachers want to be thought of as the target group of a policy, rather than the target being teachers’ unions, it seems that they may be able to portray high political power by associating with a social movement, rather than a union. This may be beneficial if teachers collectively seek to distance themselves from societal constructions of unions as contenders and instead reframe themselves as more deserving of benefits. Handmade signs are something teachers have autonomy over that seem to be closely linked to how media is portraying teachers as a group, and thoughtful messaging can allow the teachers to influence the narratives about themselves. 114 CHAPTER 8: DISCUSSION This study contributes theoretically to the understanding of social constructions of target populations and to the literature on the social constructions of teachers. Previous studies have found teachers to be typically constructed as either less deserving (e.g., Cochran-Smith & Lytle, 2006; Crespo, 2017; Goldstein, 2011) or as lacking political power (Kreitzer & Smith, 2018). Media coverage of the 2018 and 2019 Red for Ed strikes portrayed teachers as an advantaged group, both worthy of policy benefits and wielding sufficient political power to influence the policy conversation. One potential explanation for this is that strikes can function much like protests and social movements, and that these particular moments in time can be windows for redefining policy issues and policy ideas. This research connects social constructions with agenda-seeding (Wasow, 2020), which posits that the media attention from a protest can result in the issues from the protest reaching the policy agenda, by also showing how media coverage of a political activity may influence how the group participating in the political activity is portrayed. In turn, the media portrayals can both reflect and shape societal views of the target group, which may also influence how an issue is addressed in the policy agenda. Strikes as a form of political protest are being used as a tool to shape the public impression of teachers and issues related to teachers, providing support and urgency for many of the policies teachers walked out over. Future research should consider whether this pattern of teachers as an advantaged group exists in other context beyond teachers’ strikes and if the construction is persistent over time. Worthy by Association: Constructions through Associations with Other Target Groups Also, this analysis provides evidence that target groups with less established or more varied social constructions, like teachers, can affiliate with other groups to associate with those 115 target groups’ more established constructions. Teachers were constructed by association, much like the adage of “guilt by association” suggests that who you surround yourself with, and how they are perceived, reflects back on yourself. For example, one way for teachers to show political power was by affiliating with the teachers’ union. As an organized interest group with financial resources and political connections, unions have access to policymakers and the political knowledge to navigate the policy process. This is not a new strategy; teachers in the Chicago 2012 strike were associated with unions in close to two-thirds (61.54%, N=232) of the event images. However, there are trade-offs to this association, as the two Red for Ed strikes with the most frequent images showing teachers as undeserving – Los Angeles and Chicago 2019 – were also the same events that associated with unions most frequently. While teachers could associate with unions’ political power, they also were associated with the unions’ historical construction as underserving when they visually affiliated with the union. Similarly, teachers associated themselves with children’s deservingness. Unlike affiliating with unions, however, associating with children did not risk associating with children’s lack of political power. There were no images across the Red for Ed dataset that showed both teachers as deserving by affiliating with children and teachers as lacking political power. This is likely because the act of striking is a strong display of political power, potentially offsetting the less-powerful construction of children. Broadly, the implications of these findings are two-fold. First, this is evidence that affiliating with children and children’s needs is an effective political strategy for a group hoping to be portrayed positively. From a theoretical perspective, this provides evidence that how target groups are perceived in relationship to other target groups is important for understanding social constructions. Affiliations with other target groups and alliances could change how policymakers and the public view a particular target group and potential policy solutions that would benefit or 116 harm that group. Agency in Constructions: Teachers as Active Participants in their own Portrayal Furthermore, there is evidence that teachers, particularly through their use of strike signs, actively played a role in shaping the narrative about themselves. In framing themselves as deserving, highlighting economic concerns, and emphasizing their connection to children and the public good through handmade strike signs, teachers provided visuals to tell their own story in media coverage. While the media controls which images they publish, they do not control the strike signs teachers’ carry. There was some evidence that the image constructions were reflected in the article text, and future analysis considering how which policy issues were included in the text how frequently and how those discussions are frames could provide additional evidence of how teacher strike signs relate to media coverage of the events broadly. Often, social constructions literature considers how stereotypes from society are projected onto a target group, often speaking of social constructions as labels placed on target groups from outside the group itself (Schneider & Ingram, 2005a). This study builds on the limited evidence that target groups can actively engage in efforts to shift social constructions of their particular group (e.g., DiAlto, 2005; Hudson & Gonyea, 2012), indicating opportunities for groups that have traditionally been viewed as less deserving or less powerful by society and policymakers to have some agency in changing how their group is perceived. Race, Place, and Union Strength as Interwoven Strands of the Same Thread: Teacher Constructions Vary based on the Complexities of Context Comparisons of events showed that constructions of deservingness varied by event, even when teachers were on strike for the same reasons, at the same time, and striking in similar ways. Teachers on strike in locations with a higher percentage of non-white teachers and higher union 117 power were portrayed as underserving more frequently, and those strikes that occurred in less urban locations with teacher populations that had proportions of white teachers greater than or equal to the national average tended to have coverage that showed teachers as deserving more frequently. Analysis of the racial composition of teachers in images indicates that white teachers were portrayed as deserving more frequently, in alignment with research on social constructions of women accepting welfare assistance (e.g., Bensonsmith, 2005; Constance-Huggins, 2011; Hancock, 2003) that show in an American context, race and associated racial stereotypes are deeply connected to social constructions. While the multiple case studies allowed for comparisons by contexts, it is important to note that the characteristics of urbanicity, union strength, and teacher racial demographics often varied together across strike events. This qualitative study considers how these contexts work in tandem to impact deservingness; future work could consider which of these factors has a larger impact on media portrayals and public perceptions of teachers. A carefully designed survey experiment varying race, urban/suburban/rural contexts, and associations with unions independently could provide causal evidence and measures of significance to further untangle the relationships and separate influences of these contexts on social constructions of teachers. Emerging Partisan Views of Teachers: News Source Matters In a time of increasing political bias in news sources, where readers of one news source tend to be ideologically different from the readers of another news source, it is important to note that how teachers are being presented by media is different depending on the news source. It is especially worth noting that Left-leaning news sources portrayed teachers as deserving in images more frequently than right-leaning news sources and Right-leaning news sources portrayed teachers as undeserving more frequently than left-leaning news sources, even though almost all 118 news sources portrayed teachers as deserving overall. Given the increasing partisan political divide in both news media and education politics, it is likely that the differences in how Left and Right-leaning news sources portray teachers have grown, with potentially diverging characterizations of teachers. Future research tracking constructions of teachers and other educational stakeholders in news media over time should consider the political biases of news sources and how that may influence what is published. In the case of coverage of teacher strikes, pictures truly were “worth a thousand words.” While there was some variation in tone and in the use of language invoking social constructions of teachers, in the case of the Red for Ed strikes the variation in constructions was most apparent in the content of the images shared by the media. Ingram and Schneider (2005) note that social constructions can be subtle, and this study provides evidence that visual data, including media images, can be a rich source of information about subtle social attitudes that can otherwise be masked by neutral language. Sympathetic Teachers on Strike: Strikes as a Social Movement for Education Policy Change The Red for Ed strikes as a social movement are an example of the potential power teachers can collectively have as a target population, especially when considered in light of the policy gains teachers received both for themselves and the education system broadly in 2018 and 2019. This study, and its finding that teachers were generally portrayed as both deserving and as having political power, provide evidence that strategic strikes and protests by teachers are one- way teachers can influence the conversation around education policy issues. Since 2019, there has been significant change in the education policy landscape. The Covid-19 pandemic brought an abrupt end to the Red for Ed movement, though teacher strikes have resumed following the 119 re-opening of schools (Will, 2022). Going forward, it will be important to examine how the news media portrays teachers in an increasingly politicized and polarized climate, and to consider ways that teachers are able to actively participate in the education policy process. As key stakeholders in the education system, it is important to consider ways teachers can be perceived as worthy of policy benefits and how teachers can effectively marshal and exert political power to influence policy. For those who believe teachers’ voices should be reflected in educational policy, the Red for Ed movement should be considered an exemplar of strategic political involvement, where teachers were able to shape the narrative around policy issues. 120 REFERENCES AllSides.com. (2023). AllSides Media Bias Ratings. https://www.allsides.com/media-bias/media- bias-ratings Alpers, G.W., Winzelberg, A.J., Classen, C., Roberts, H., Dev, P., Koopman, C., et al. (2005). Evaluation of computerized text analysis in an Internet breast cancer support group. Computers in Human Behavior, 21, 361-376. https://doi.org/10.1016/j.chb.2004.02.008 Amrein-Beardsley, A. (2017, Aug. 18). VAMboozled! The “Widget Effect” report revisited. National Education Policy Center. https://nepc.colorado.edu/blog/widget- effect Amundson, K. & Zajicek, A. (2018). A case study of state-level policymakers’ discursive co- constructions of welfare drug testing policy and gender, race, and class. Sociological Inquiry, 88(3), 383-409. https://doi.org/10.1111/soin.12208 Andone, D. (2019, Oct. 3). More than 30,000 Chicago school staff will strike on October 17 if their demands aren’t met, union says. CNN. https://www.cnn.com/2019/10/03/us/chicago-teachers-union-strike/index.html Associated Press. (2018). [Photograph of Democratic state Rep. Rick Rand speaks against a bill to change the state’s troubled pension system, March 29, 2018 in Frankfort, Ky.] Fox News. https://www.foxnews.com/politics/kentucky-lawmakers-pass-pension-overhaul- despite-teacher-protests Associated Press Domestic. (2019a). Mayor to teachers: Chicago won’t budge on money [Video]. USA Today. https://www.usatoday.com/videos/news/nation/2019/10/17/mayor- teachers-chicago-wont-budge-money/4013448002 Associated Press. (2019b). [Photograph of striking teachers picketing outside George Leland Elementary School on Chicago’s West Side]. Associated Press. https://www.foxnews.com/us/chicago-teachers-strike-300000-students-without-classes (2nd) Bain, J., Weishaar, H., Semple, S., Duffy, S., & Hilton, S. (2017). Vulnerable children, stigmatised smokers: the social construction of target audiences in media debates on policies regulating smoking in vehicles. Health, 21(6), 633-649. Balingit, M. (2018, Mar. 30). ‘It just hurts my heart’: Low pay, big classes are the plight of Oklahoma teachers. The Washington Post. https://www.washingtonpost.com/local/education/it-just-hurts-my-heart-low-pay-big- classes-are-the-plight-of-oklahoma-teachers/2018/03/30/e5e10eb8-2c88-11e8-b0b0- f706877db618_story.html Ballotpedia. (2022a). Chicago Public Schools, Illinois. https://ballotpedia.org/Chicago_Public_Schools,_Illinois 121 Ballotpedia. (2022b). Denver Public Schools, Colorado. https://ballotpedia.org/Denver_Public_Schools,_Colorado Ballotpedia. (2022c). Kentucky House of Representatives. https://ballotpedia.org/Kentucky_House_of_Representatives Ballotpedia. (2022d). Kentucky State Senate. https://ballotpedia.org/Kentucky_State_Senate Ballotpedia. (2022e). Los Angeles Unified School District, California. https://ballotpedia.org/Los_Angeles_Unified_School_District,_California Ballotpedia. (2022f). Oklahoma House of Representatives. https://ballotpedia.org/Oklahoma_House_of_Representatives Ballotpedia. (2022g). Oklahoma State Senate. https://ballotpedia.org/Oklahoma_State_Senate Ballotpedia. (2022h). West Virginia House of Delegates. https://ballotpedia.org/West_Virginia_House_of_Delegates Ballotpedia. (2022i). West Virginia State Senate. https://ballotpedia.org/West_Virginia_State_Senate Baumgartner, F. R. & Jones, B. D. (2009). Agendas and instability in American politics (2nd ed.). University of Chicago Press. Bazeley, P. (2013). Qualitative data analysis: Practical strategies. Sage. Berry, S. (2019, Jan. 10). Los Angeles teachers’ union postpones strike to avert court-ordered delay. Breitbart. https://www.breitbart.com/politics/2019/01/10/los-angeles-teachers- union-postpones-strike-avert-court-ordered-delay/ Bengston, D., & Xu, Z. (1995). Changing National Forest Values: a content analysis - Research Paper NC-323. http://www.nrs.fs.fed.us/pubs/rp/rp_nc323.pdf Bensonsmith, D. (2005). Jezebels, matriarchs, and welfare queens: The Moynihan Report of 1965 and the social construction of African-American women in welfare policy. In A. L. Schneider & Ingram, H. M. (Eds.), Deserving and entitled: Social constructions and public policy (pp. 243-260). State University of New York Press. Berlin, J. (2019, Oct. 31). 102 days on strike: Take a look back at the 11 teacher strikes since 1969. Chicago Tribune. https://www.chicagotribune.com/news/ct-met-viz-chicago- teacher-strikes-timeline-20190920-htmlstory.html Bezem, P. (2021). School inspection in the United States: Potential for school reform and lasting change (Doctoral dissertation, Michigan State University, East Lansing, MI). Bleich, E., Nisar, H., & Vazquez, C. (2018a). Investigating state hierarchies with media analysis: Muslims, Jews, and Catholics in The New York Times and The 122 Guardian headlines, 1986-2014. International Journal of Comparative Sociology, 59(3), 239-257. https://doi.org/10.1177/0020715218775142 Bleich, E., Souffrant, J., Stabler, E., & van der Veen, A. (2018b). Media Coverage of Muslim Devotion: A Four-Country Analysis of Newspaper Articles, 1996–2016. Religions, 9(8), 247. https://doi.org/10.3390/rel9080247 Bleich, E. & van der Veen, A. M. (2018). Media portrayals of Muslims: A comparative sentiment analysis of American newspapers, 1996-2015. Politics, Groups, and Identities, 9(1), 20-39. https://doi.org/10.1080/21565503.2018.1531770 Boris, E. (2007). On cowboys and welfare queens: Independence, dependence, and interdependence at home and abroad. Journal of American Studies, 41(3), 599- 621. https://doi.org/10.1017/S002187580700401X Boyle, C. (2019). [Photograph of Chicago Public school teachers picketing early in the morning]. Chicago Sun Times via the Associated Press. https://www.usatoday.com/picture- gallery/news/nation/2019/10/17/chicago-public-school-teachers-strike-after-failure- negotiate-contract/4006540002/ Broockman, D. E., & Kalla, J. L. (2022, August 3). The impacts of selective partisan media exposure: A field experiment with Fox News viewers [Working paper]. Brown, F. J. (2019). [Photograph of teacher in red poncho marching with a UTLA sign]. Getty Images. https://www.breitbart.com/politics/2019/01/15/rebecca-friedrichs-strikes-about- more-money-teachers-unions-far-left-politics Bulkley, K. & Gottlieb, J. (2017). Policy images of teachers: How influential actors construct images of teachers. Teachers College Record, 119(4), 1-34. Byrne, D. (2009) Case-based methods: Why we need them; what they are; how to do them. In D. Byrne & C. Ragin (Eds.), Handbook of case-based methods (pp. 1-10). Sage. Cammett, A. (2014). Deadbeat dads & welfare queens: How metaphor shapes poverty law. Boston College Journal of Law & Social Justice, 34, 233-265. Carter, J. P. (2018). [Photograph of teachers rallying at the state Capitol on Tuesday in Oklahoma City]. Slate. https://slate.com/news-and-politics/2018/04/democrats-should- do-more-to-support-striking-teachers.html Cassese, E. C., & Barnes, T. D. (2019). Intersectional motherhood: Investigating public support for childcare subsidies. Politics, Groups, and Identities, 7(4), 775-793. https://doi.org/10.1080/21565503.2018.1441035 Choi, Y., & Chung, I. H. (2016). Voice effects of public sector unions on turnover: Evidence from teacher contracts. Public Personnel Management, 45(2), 213-233. https://doi.org/10.1177/0091026016645063 123 Ciaglo, M. (2019). [Photograph of Denver Public Schools teachers and members of the community picketing at sunrise outside South High School]. Getty Images. https://www.usatoday.com/picture-gallery/news/education/2019/02/11/thousands-walk- out-denver-teachers-strike/2838974002/ CNN. (2012, Sept. 11). Tony Danza on the Chicago teacher strike [Video]. CNN. https://www.cnn.com/videos/bestoftv/2012/09/11/exp-point-tony-danza.cnn CNN. (2018a, March 5). Teacher stands by the road to raise class funds [Video]. CNN. https://www.cnn.com/2018/03/05/us/oklahoma-teachers-possible-strike-trnd/index.html CNN. (2018b). [Photograph of happy teachers in front of West Virginia’s state capitol building]. CNN. https://www.cnn.com/2018/03/09/us/teachers-union-movement/index.html CNN. (2018c). [Photograph of Melissa Knight, teacher, demonstrating outside the state capitol in Oklahoma]. CNN. https://www.cnn.com/us/live-news/teachers-walkout-oklahoma- kentucky/index.html CNN. (2018d). [Photograph of teachers in Kentucky protesting at the state capitol calling for higher wages and respect]. CNN. https://www.cnn.com/us/live-news/teachers-walkout- oklahoma-kentucky/index.html CNN. (2019a). [Photograph of second grade teacher Fiona Engler with her protest sign]. https://www.cnn.com/2019/01/17/us/why-los-angeles-teachers-are-striking- trnd/index.html CNN. (2019b). [Photograph of teacher Katie McOwen smiling in her classroom]. CNN https://www.cnn.com/2019/02/10/us/denver-teacher-strike-multiple-jobs/index.html Cocca, C. E. (2002). From" welfare queen" to" exploited teen": Welfare dependency, statutory rape, and moral panic. NWSA Journal, 56-79. Cochran-Smith, M., & Lytle, S. (2006). Troubling images of teaching in no child left behind. Harvard Educational Review, 76(4), 668-697. https://doi.org/10.17763/haer.76.4.56v8881368215714 Collinson, S. (2018, Apr. 3). From teens to teachers, activism stirs the electorate. CNN. https://www.cnn.com/2018/04/03/politics/protests-teachers-trump-oklahoma- kentucky/index.html Constance-Huggins, M. (2011). A review of the racial biases of social welfare policies. Journal of human behavior in the social environment, 21(8), 871-887. https://doi.org/10.1080/10911359.2011.588531 Cowen, J. M., & Strunk, K. O. (2015). The impact of teachers’ unions on educational outcomes: What we know and what we need to learn. Economics of Education Review, 48, 208- 223. https://doi.org/10.1016/j.econedurev.2015.02.006 124 Crespo, J. (2017). The Social Construction of Teachers and the Teaching Profession among Florida State Legislators from 1984 to 2015 (Doctoral dissertation, Florida Atlantic University, Boca Raton, FL). Crisp, J. (2018a). [Photograph of rally attendees listening to a speaker in Frankfort, with a teachers holding a sign reading, “Fighting for the future of our kids” in the center of the image]. Courier Journal. https://www.courier-journal.com/picture- gallery/news/politics/2018/03/12/teachers-and-others-rally-big-against-pension-bill-in- frankfort/32875905/ Crisp, J. (2018b). [Photograph of rally attendees listening to a speaker in Frankfort, with children and other rally attendees standing above the heads of the crowd]. https://www.courier- journal.com/picture-gallery/news/politics/2018/03/12/teachers-and-others-rally-big- against-pension-bill-in-frankfort/32875905/ D’Amico Pawlewicz, D. (2020). Blaming teachers: Professionalization policies and the failure of reform in American history. Rutgers University Press. Davey, M. (2012, Sept. 10). Teachers’ strike in Chicago tests mayor and union. The New York Times. https://www.nytimes.com/2012/09/11/education/teacher-strike-begins-in-chicago- amid-signs-that-deal-isnt-close.html?searchResultPosition=3 Davidson, B. (2019). [Photograph of people marching through the streets of downtown Los Angeles in the pouring rain]. Breitbart. https://www.breitbart.com/education/2019/01/14/photos-thousands-of-l-a-teachers- march-on-strike/ Davis, J. (2019, Jan. 17). Teacher: “I’m not being as effective as I could be” [Video]. CNN. https://www.cnn.com/2019/01/17/us/why-los-angeles-teachers-are-striking- trnd/index.html deLeon, P. (2005). Review: Social construction for public policy. Public Administration Review, 65(5), 635-637. http://www.jstor.com/stable/3542530 Deng, Q., Hine, M., Ji, S., & Sur, S. (2019). Inside the black box of dictionary building for text analytics: a design science approach. Journal of International Technology and Information Management, 27(3), 119–159. DiAlto, S. J. (2005). From “problem minority” to “model minority”: The changing social construction of Japanese Americans. In A. L. Schneider & Ingram, H. M. (Eds.), Deserving and entitled: Social constructions and public policy (pp. 81-103). State University of New York Press. Di Carlo, M. (2014, Mar. 4). Revisiting The Widget Effect. Albert Shanker Institute. https://www.shankerinstitute.org/blog/revisiting-widget-effect Dorst, C. (2018, March 2). [Photograph of striking teachers demonstrating outside the West Virginia Capitol in Charleston on Monday]. Charleston Gazette-Mail. 125 https://www.nytimes.com/2018/03/02/opinion/west-virginia-unions- teachers.html?searchResultPosition=9 Dovarganes, D. (2019). [Photograph of Los Angeles teachers rallying on December 15 to demand more school staffing, smaller class sizes and higher salaries]. CNN. https://www.cnn.com/2019/01/14/us/los-angeles-teachers-strike/index.html Education Next. (2020). Education Next annual poll: Trends through 2020 [Interactive graphic]. https://www.educationnext.org/ednext-poll-interactive-trends-through-2020-public- opinion/ Farris, E. M., & Silber Mohamed, H. (2018). Picturing immigration: how the media criminalizes immigrants. Politics, Groups, and Identities, 6(4), 814-824. https://doi.org/10.1080/21565503.2018.1484375 Finn, C. E., Jr., & Petrilli, M. J. (2013). The failures of U.S. education governance today. In P. Manna & McGuinn, P. (Eds.), Education governance for the twenty-first century: Overcoming the structural barriers to school reform (pp. 21-35). Brookings. Flynn, A. (2018). [Photograph of the empty cafeteria at University High School in Morgantown]. The New York Times. https://www.nytimes.com/2018/03/02/us/west-virginia-teachers- strike.html Foster, C. H. (2008). The welfare queen: Race, gender, class, and public opinion. Race, Gender & Class, 15(3-4), 162-179. https://www.jstor.org/stable/41674659 Frankenberg, E. (2009). The demographic context of urban schools and districts. Equity & Excellence in Education, 42(3), 255-271. https://doi.org/10.1080/10665680903032294 Franklin, R. D. (2017). [Photograph of protesters gathered at the Arizona School for the Arts]. The Washington Post. https://www.washingtonpost.com/politics/why-republicans-are- now-bragging-about-their-investments-in-education/2018/03/25/1faed770-26c0-11e8- b79d-f3d931db7f68_story.html Fox News. (2018). [Photograph of Oklahoma Governor Mary Fallin from the National Governors Association]. Fox News. https://www.foxnews.com/us/oklahoma-teachers- demands-like-a-teenager-wanting-a-better-car-governor-says Gallup. (2020, Dec. 17). In depth: Topics A-Z: Education. https://news.gallup.com/poll/1612/education.aspx. Goldstein, D. (2018, Apr. 2). Teachers in Oklahoma and Kentucky Walk Out: ‘It Really is a Wildfire.’ The New York Times. https://www.nytimes.com/2018/04/02/us/teacher-strikes- oklahoma-kentucky.html?searchResultPosition=4 Goldstein, D. (2014). The teacher wars: A history of America’s most embattled profession. Doubleday. 126 Goldstein, R. A. (2011). Imaging the frame: Media representations of teachers, their unions, NCLB, and education reform. Educational Policy, 25(4), 543-576. https://doi.org/10.1177/0895904810361720 Greene, S. (2018). [Photograph of protesters chanting and booing as they fill the Kentucky State Capitol]. USA Today. https://www.usatoday.com/picture- gallery/news/nation/2018/04/02/teachers-protest-around-the-country/33479911/ Gress, J. (2019). [Photograph of picketers outside Sullivan High School on the second day of a teachers’ strike in Chicago]. The Wall Street Journal. https://www.wsj.com/articles/chicago-teachers-strike-for-second-day- 11571433491?mod=searchresults&page=1&pos=12 Groeling, T. (2013). Media bias by the numbers: Challenges and opportunities in the empirical study of partisan news. Annual Review of Political Science, 16, 129-151. https://doi.org/10.1146/annurev-polisci-040811-115123 Han, E. S., & Maloney, T. N. (2021). Teacher unionization and student academic performance: Looking beyond collective bargaining. Labor Studies Journal, 46(1), 43-74. https://doi.org/10.1177/0160449X19883373 Hancock, A. M. (2003). Contemporary welfare reform and the public identity of the “welfare queen.” Race, Gender & Class, 10(1), 31-59. http://www.jstor.com/stable/41675059 Hanna, J. & Weisfeldt, S. (2018, Apr. 15). Colorado teachers plan their own Capitol rally Monday. CNN. https://www.cnn.com/2018/04/15/us/colorado-teacher-day-of- action/index.html Hernández, L. E. (2016). Race and racelessness in CMO marketing: Exploring charter management organizations” racial construction and its implications. Peabody Journal of Education, 91(1), 47-63. https://doi.org/10.1080/0161956X.2016.1119566 Hertel-Fernandez, A., Naidu, S., & Reich, A. (2021). Schooled by strikes? The effects of large- scale labor unrest on mass attitudes toward the labor movement. Perspectives on Politics, 19(1), 73-91. https://doi.org/10.1017/S1537592720001279 Henderson, M. B., Houston, D., Peterson, P. E., & West, M. R. (2019, Aug. 20). Public support grows for higher teacher pay and expanded school choice. Education Next. https://www.educationnext.org/school-choice-trump-era-results-2019-education-next- poll/ Henig, J. R. (2013). The end of exceptionalism in American education: The changing politics of school reform. Harvard Education Press. Hess, F. & Martin, RJ. (2018, Aug. 27). How did major newspapers cover the 2018 teacher strikes? Education Next. https://www.educationnext.org/how-did-major-newspapers- cover-2018-teacher-strikes/ 127 Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to decline in firms, organizations, and states. Harvard university press. Hobbs, T. D. (2019, Oct. 16). Chicago cancels Thursday public-school classes as strike begins. The Wall Street Journal. https://www.wsj.com/articles/chicago-teachers-on-brink-of- strikeand-pay-isnt-only-demand-11571223600 Hudson, C. (2018). [Photograph of teachers in Charleston, WV reacting to the news that they would receive a 5 percent raise]. The New York Times. https://www.nytimes.com/2018/03/07/briefing/winter-storm-texas-stormy-daniels.html Hudson, R. B. & Gonyea, J. G. (2012). Baby Boomers and the shifting political construction of old age. The Gerontologist, 52(2), 272-282. https://doi.org/10.1093/geront/gnr129 Ingram, H. & Schneider, A. (1991). The choice of target population. Administration & Society, 23(3), 333-356. Ingram, H. M. & Schneider, A. L. (2005). Introduction: Public policy and the social construction of deservedness. In A. L. Schneider & Ingram, H. M. (Eds.), Deserving and entitled: Social constructions and public policy (pp. 1-34). State University of New York Press. Ingram, H., Schneider, A. L., & deLeon, P. (2007). Social construction and policy design. In P. A. Sabatier (Ed.), Theories of the policy process (2nd ed.). Westview Press. James, E. A., & Rashid, M. (2013). “Welfare queens” and “teen moms”: how the social construction of fertile women impacts unintended pregnancy prevention policy in the United States. Policy, Politics, & Nursing Practice, 14(3-4), 125- 132. https://doi.org/10/1177/1527154413510408 Janus v. American Federation of State, County, and Municipal Employees, 585 U.S.__ (2018). https://www.oyez.org/cases/2017/16-1466 Jones, A. L. (2013). Shared Spaces or Separate Meanings? Mediated Communication as a Globalizing Force. (Doctoral dissertation, George Washington University, Washington, DC) Joseph, P. E. (2016, May 18). The many meanings of a fist. Chronicle of Higher Education, 62(37), B4-B5. https://www.chronicle.com/article/the-many-meanings-of-a-fist/ Jurkowitz, M., Mitchell, A., Shearer, E. & Walker, M. (2020, Jan. 24). U.S. Media polarization and the 2020 election: A nation divided. PEW Research Center. https://www.pewresearch.org/journalism/wp- content/uploads/sites/8/2020/01/PJ_2020.01.24_Media-Polarization_FINAL.pdf Karimi, F. & Sutton, J. (2018, Apr. 20). Arizona teachers vote for statewide walkout. CNN. https://www.cnn.com/2018/04/20/us/arizona-teachers-walkout-vote/index.html Karp, S., & Sanchez, A. (2018, Summer). The 2018 wave of teacher strikes: A turning point for 128 our schools? Rethinking Schools, 32(4). https://rethinkingschools.org/articles/the-2018- wave-of-teacher-strikes/ Korff, G. & Peterson, L. (1992). From brotherly handshake to militant clenched fist: On political metaphors for the worker’s hand. International Labor and Working-Class History, 42, 70-81. http://www.jstor.org/stable/27672022 Kreitzer, R. J. & Smith, C. W. (2018). Reproducible and replicable: An empirical assessment of the social construction of politically relevant target groups. PS: Political Science and Politics, 51(4), 768-774. https://doi.org/10.1017/S1049096518000987 Krugman, P. (2018, Apr. 23). We don’t need no education. The New York Times. https://www.nytimes.com/2018/04/23/opinion/teachers-protest-education-funding.html Krzaczynski, K. (2019). [Photograph of a striking teacher holding a sign as she walks a picket line]. USA Today. https://www.usatoday.com/picture- gallery/news/nation/2019/10/17/chicago-public-school-teachers-strike-after-failure- negotiate-contract/4006540002/ Leahy, M. P. (2019, Feb. 28). #RedforEd leader in Marxist magazine: Teachers must embrace community organizing. Breitbart. https://www.breitbart.com/politics/2019/02/28/redfored-leader-in-marxist-magazine- teachers-must-embrace-community-organizing/ Lee, J. & Liebelson, D. (2012, Sept. 20). Map: Wave of new teacher strikes hits Illinois. Mother Jones. https://www.motherjones.com/politics/2012/09/map-teacher-strikes-across-united- states/ Lieberman, R. C. (1995). Social construction (continued). American Political Science Review, 89(2), 437-441. Limbert, W. M., & Bullock, H. E. (2009). Framing US redistributive policies: Tough love for poor women and tax cuts for seniors. Analyses of Social Issues and Public Policy, 9(1), 57-83. https://doi.org/10.1111/j.1530-2415.2009.01198.x Lipsky, M. (1968). Protest as a political resource. American political science review, 62(4), 1144-1158. Lott, J. (2019). [Photograph of Mayor Lori Lightfoot of Chicago at a Y.M.C.A. with students affected by the teachers’ strike on Thursday]. The New York Times. https://www.nytimes.com/2019/10/17/us/chicago-schools-strike.html Loewus, L. (2019, Mar. 4). Where Have There Been Teacher Strikes and Protests? (Map). Education Week. https://www.edweek.org/ew/section/multimedia/map-where- have-there-been-teacher-strikes.html. Lyon, M. A. (2022). In the strike zone: New data to contextualize the recent surge in teacher strikes [brief]. Urban Institute. https://www.urban.org/sites/default/files/2022- 129 05/In%20the%20Strike%20Zone%20New%20Data%20to%20Contextualize%20the%20 Recent%20Surge%20in%20Teacher%20Strikes.pdf Lyon, M. A. (2021). Heroes, villains, or something in between? How “Right to Work” policies affect teachers, students, and education policymaking. Economics of Education Review, 82. https://doi.org/10.1016/j.econedurev.2021.102105 Lyon, M. A. & Kraft, M. A. (2021). Elevating education in politics: How teacher strikes shape Congressional election campaigns. (Working Paper). Retrieved from https://sites.google.com/brown.edu/melissa-lyon/research?authuser=0 Martinez Monsivals, P. (2018). [Photograph of Education Secretary Betsy DeVos speaking]. The Washington Post. https://www.washingtonpost.com/news/education/wp/2018/04/09/betsy-devos-to- oklahoma-teachers-serve-the-students/ Mathis-Lilley, B. (2018, Feb. 26). Every public school in West Virginia has shut down as teachers’ strike enters second week. Slate. https://slate.com/news-and- politics/2018/02/west-virginia-schools-all-closed-during-teachers-strike.html Maxwell, J. A. (2013). Validity: How might you be wrong? In Qualitative research design: An interactive approach (3rd ed., (pp. 121-138). SAGE publications. May, P. J. (1992). Policy learning and failure. Journal of Public Policy, 12(4), 331-354. www.jstor.org/stable/4007550 McCombs, M., & Valenzuela, S. (2021). Setting the agenda: Mass media and public opinion. John Wiley & Sons. McGinty, E. E., Kennedy-Hendricks, A., Choksy, S., Barry, C. L. (2016). Trends in news media coverage of mental illness in the United States: 1995-2014. Health Affairs, 35(6), 1121- 1129. https://doi.org/10.1377/hlthaff.2016.0011 McLaren, M. (2018, Apr. 13). Kentucky governor says children were abused because teachers rallied [Video]. USA Today. https://www.usatoday.com/story/news/nation- now/2018/04/13/kentucky-governor-children-were-sexually-assaulted-because-teachers- protesting/516783002 McLaren, M. & Aulbach, L. (2018, Apr. 13). Bevin: Children were sexually assaulted and tried drugs because teachers were protesting. USA Today via The Courier Journal. https://www.courier-journal.com/story/news/politics/ky-legislature/2018/04/13/matt- bevin-kentucky-governor-children-sexually-assaulted-teacher-protests/516569002/ Mercier, A. (2019, Oct. 24). The top 25 U.S. daily newspapers of fall 2019. Muck Rack Blog. https://muckrack.com/blog/2019/10/24/the-top-25-us-daily-newspapers-of-fall-2019 Merriam, S. B. & Tisdell, E. J. (2016). Qualitative research: A guide to design and implementation. Jossey-Bass. 130 Merrill, C. (2018). [Series of charts that show why America’s teachers are fired up and can’t take any more]. CNN. https://www.cnn.com/2018/04/13/health/teachers-pay-profession- charts-stats-trnd/index.html Mettler, S. & Sorelle, M. (2017). Policy feedback theory. In C. M. Weible & P. A. Sabatier (Eds.), Theories of the policy process (4th ed., pp. 103-134). Mitchell, A., Jurkowitz, M., Oliphant, J. B., & Shearer, E. (2021, Feb. 22). How Americans navigated the news in 2020: A tumultuous year in review. PEW Research Center. https://www.pewresearch.org/journalism/wp- content/uploads/sites/8/2021/02/PJ_2021.02.22_News-Pathways-2020- Lookback_FINAL.pdf Moe, T. (2001). The problem of union power. In Special interest: Teachers unions & America’s public schools. Washington, D.C.: Brookings Institution Press. Morin, A. (2020). How to help your child learn to read facial expressions. https://www.understood.org/en/friends-feelings/common-challenges/picking-up-on- social-cues/helping-your-child-learn-to-read-facial-expressions. Muck Rack (2023). New York Post. https://muckrack.com/media-outlet/nypost National Center for Education Statistics. (2018). Percentage of public school teachers who indicated that they were members of a union or an employee association similar to a union, by state: 2017-18. National Teacher and Principal Survey. https://nces.ed.gov/surveys/ntps/tables/ntps1718_20111201_t1s.asp Nicas, J. (2012, September 17). Chicago parents protest teachers’ strike [Video]. Wall Street Journal. https://www.wsj.com/articles/SB10000872396390443995604578002180453638800 Nittler, K. (2019, Mar. 28). Collective bargaining and teacher strikes [blog]. National Council on Teacher Quality. https://www.nctq.org/blog/Collective-bargaining-and-teacher-strikes Ogrocki, S. (2018). [Photograph from March 27, 2018 showing teacher Adrien Gates picketing with other educators in Norman, OK]. Associated Press. https://www.usatoday.com/story/news/2018/04/02/teachers-strike-spacex- cosby/468876002/ Osborne-Lampkin, L., Cohen-Vogel, L., Feng, L. & Wilson, J. J. (2017). Researching collective bargaining agreements: Building conceptual understanding in an era of declining union power. Educational Policy, 32(2), 152-188. https://doi- org.proxy2.cl.msu.edu/10.1177/0895904817745378 Ostrom, E. (1990). Governing the commons: The evolution of institutions for collective action. Cambridge University Press. Pearson, M. (2012, Sept. 19). Wins, losses and draws in Chicago 131 school strike. CNN. https://www.cnn.com/2012/09/19/us/illinois-chicago-teachers- strike/index.html. Pennebaker Conglomerates, Inc. (2023). How it works. https://www.liwc.app/help/howitworks Pennebaker, J.W., Boyd, R.L., Jordan, K., & Blackburn, K. (2015). The development and psychometric properties of LIWC2015.Austin, TX: University of Texas at Austin. DOI: 10.15781/T29G https://www.liwc.app/static/documents/LIWC2015%20Manual%20- %20Development%20and%20Psychometrics.pdf Pierce, J. J., Siddiki, S., Jones, M. D., Schumacher, K., Pattison, A., & Peterson, H. (2014). Social construction and policy design: A review of past applications. Policy Studies Journal, 42(1), 1-30. https://doi.org/10.1111/psj.12040 Provalis Research. (2020). QDA Miner 6: User’s Guide. https://q9j3s8w6.rocketcdn.me/Documents/QDAMiner6.pdf Provalis Research. (2021). Word Stat 9 Text Analytics Software: User’s Guide. https://q9j3s8w6.rocketcdn.me/Documents/WordStat9.pdf Raby, J. (2019). [Photograph of West Virginia Gov. Jim Justice announcing a deal was reached to end a statewide teachers strike]. The Washington Post. https://www.washingtonpost.com/news/education/wp/2018/02/27/west-virginia-teachers- expected-to-return-to-classrooms-thursday-as-gov-jim-justice-announces-deal Ravitch, D. (2006). Why teacher unions are good for teachers-and the public. American Educator, 30(4), 6. https://www.aft.org/periodical/american-educator/winter-2006- 2007/why-teacher-unions-are-good-teachers-and Reilly, K. (2019, Jan 10). The history of American teacher strikes- and where Los Angeles fits in. Time. https://time.com/5498897/history-american-teacher-strike-los-angeles/ Reuters. (2019, Jan. 22). L.A. teacher’s union president: “Public education desperately needs attention” [Video]. The Washington Post. https://www.washingtonpost.com/video/national/la-teachers-union-president-public- education-desperately-needs-attention/2019/01/22/87d060cf-6c4d-4b45-91fb- bcc4ddcba65d_video.html Rezin Garcia, A. (2019). [Photograph of striking Chicago Teachers Union members from McAuliffe Elementary School rally on a corner]. Associated Press. https://www.usatoday.com/story/news/education/2019/10/17/chicago-teachers-union- strike-new-contract/4000656002 Ritchie, J. & Ormstrom, R. (2014). The applications of qualitative methods to social research. In J. Ritchie, J. Lewis, C. M. Nicholls, & R. Ormston (Eds.), Qualitative research practice: A guide for social science students & researchers (2nd ed., pp. 27-46). Sage. Rose, M., & Baumgartner, F. R. (2013). Framing the poor: Media coverage and US poverty 132 policy, 1960–2008. Policy Studies Journal, 41(1), 22-53. Sabatier, P. (1999). Theories of the Policy Process (1st ed.). Westview Press. Saldaña, J. (2016). The Coding Manual for Qualitative Researchers (3rd ed). SAGE publications. Sanes, M. & Schmitt, J. (2014, Mar.). Regulation of public sector collective bargaining in the states. Center for Economic and Policy Research. https://cepr.net/documents/state-public- cb-2014-03.pdf Sangosti, R. J. (2019). [Photograph of Denver teachers picketing before school January 15]. Denver Post. https://www.cnn.com/2019/02/06/us/denver-teachers-strike/index.html Schallhorn, K. (2018, Apr. 13). Oklahoma teachers strike for higher pay, education funding. Fox News. https://www.foxnews.com/us/oklahoma-teachers-strike-for-higher-pay-education- funding Schneider, A. & Ingram, H. (1993). Social construction of target populations: Implications for politics and policy. The American Political Science Review, 87(2), 334- 347. https://doi.org/10.2307/2939044 Schneider, A. L. & Ingram, H. (1997). Policy design for democracy. University Press of Kansas. Schneider, A. L. & Ingram, H. M. (Eds.) (2005a). Deserving and entitled: Social constructions and public policy. State University of New York Press. Schneider, A. L. & Ingram, H. M. (2005b). A response to Peter deLeon. Public Administration Review, 65(5), 638-640. https://www-jstor-org.proxy1.cl.msu.edu/stable/3542531 Schneider, A. L., & Ingram, H. M. (2019). Social constructions, anticipatory feedback strategies, and deceptive public policy. Policy Studies Journal, 47(2), 206-236. https://doi.org/10.1111/psj.12281 Schnell, L. (2019). Hooper Teacher Rally 001 [Photograph]. USA Today. https://www.heraldmailmedia.com/story/news/2019/01/14/our-armageddon-los-angeles- teachers-strike-in-nations-2nd-largest-school-distric/116469362/ Schram, S. F. (2005). Putting a Black face on welfare: The good and the bad. In A. L. Schneider & Ingram, H. M. (Eds.), Deserving and entitled: Social constructions and public policy (pp. 261-286). State University of New York Press. Schreiner, B. (2019, Aug. 16). Kentucky finds 1,000 teachers violated law during ‘sickouts.’ ABC News. https://abcnews.go.com/US/wireStory/kentucky-finds-1000- teachers-violated-law-sickouts-65021856 Sedgwick, J. (2018, Apr. 16). 25 year old textbooks and holes in the ceiling: Inside America’s 133 public schools. The New York Times. https://www.nytimes.com/2018/04/16/reader- center/us-public-schools-conditions.html Sensoy, Ö. & DiAngelo, R. (2017). Is everyone really equal? An introduction to key concepts in social justice education (2nd ed.). Teachers College Press. Shanahan, E. A., Jones, M.D., McBeth, M. K., & Radaelli, C. M. (2017). The narrative policy framework. In C. M. Weible & P. A. Sabatier (Eds.), Theories of the policy process (4th ed., pp. 173-214). Silber Mohamed, H., & Farris, E. M. (2020). ‘Bad hombres’? An examination of identities in US media coverage of immigration. Journal of Ethnic and Migration Studies, 46(1), 158- 176. Smith, M. & Davey, M. (2019, Oct. 31). Chicago Teachers’ Strike, Longest in Decades, Ends. The New York Times. https://www.nytimes.com/2019/10/31/us/chicago-cps-teachers- strike.html. Smith, K. B. & Larimer, C. W. (2017). The public policy primer (3rd ed.). Westview Press. Spencer, L., Ritchie, J., Ormstrom, R., O’Connor, W., & Barnard, M. (2014). Analysis: Principles and process. In J. Ritchie, J. Lewis, C. M. Nicholls, & R. Ormston (Eds.), Qualitative research practice: A guide for social science students & researchers (2nd ed., pp. 269-294). Sage. Stone, D. (2012). Policy paradox: The art of political decision making (3rd ed.). W.W. Norton & Company. Tausczik, Y. R., & Pennebaker, J. W. (2010). The Psychological Meaning of Words: LIWC and Computerized Text Analysis Methods. Journal of Language and Social Psychology, 29(1), 24–54. https://doi.org/10.1177/0261927X09351676 Turner, C., Lombardo, C., & Logan, E. B. (2018, Apr. 25). Teacher walkouts: A state by state guide. NPR. https://www.npr.org/sections/ed/2018/04/25/602859780/teacher-walkouts- a-state-by-state-guide. U.S. Census Bureau. (2022). 2014-2018 EEO Tables (ACS 5-Year Data) [Data set]. https://www.census.gov/acs/www/data/eeo-data/eeo-tables-2018/ U.S. Department of Education. (2009, July 2). Partners in reform: Remarks of Arne Duncan to the National Education Association [Press release of speech]. https://www2.ed.gov/news/speeches/2009/07/07022009.html U.S. Department of Education. (2015, November) Fundamental change: Innovation in America’s schools under Race to the Top. https://www2.ed.gov/programs/racetothetop/rttfinalrptfull.pdf. Van Green, T. (2021, Aug. 20). Republicans increasingly critical of several major U.S. 134 institutions, including big corporations and banks. Pew Research Center. https://www.pewresearch.org/fact-tank/2021/08/20/republicans-increasingly-critical-of- several-major-u-s-institutions-including-big-corporations-and-banks/ Vera, A. (2018, Mar. 5). Why are West Virginia teachers on strike? Take a look at their salaries. CNN. https://www.cnn.com/2018/02/27/health/west-virginia-teachers-salaries- trnd/index.html Wallace, A. (2018). [Photograph of Amber Epperson, an elementary speech language pathology teacher participating in the Oklahoma City strikes]. The Republic. https://www.usatoday.com/picture-gallery/news/politics/arizona- education/2018/04/02/teachers-in-oklahoma-fight-for-student-funding/33495867/ Washington Post. (2019). [Line graph of teacher earnings relative to comparable worker]. The Washington Post. https://www.washingtonpost.com/us-policy/2019/02/14/with-teachers- lead-more-workers-went-strike-than-any-year-since/ Wasow, O. (2020). Agenda seeding: How 1960s Black protests moved elites, public opinion, and voting. American Political Science Review, 114(3), 638-659. http://dx.doi.org/10.1017/S000305542000009X Weir, B. (2018, Apr. 2). Oklahoma’s rich energy fields not enough for state’s teachers. CNN. https://www.cnn.com/2018/04/02/us/oklahoma-teachers-energy-taxes/index.html Weisburg, D., Sexton, S., Mulhern, J., & Keeling, D. (2009, June 8). The widget effect: Our national failure to acknowledge and act on differences in teacher effectiveness, 2nd ed. The National Teachers Project. https://files.eric.ed.gov/fulltext/ED515656.pdf. Will, M. (2022, Sept. 19). Why teachers are going on strike this fall–and what could come next. Education Week. https://www.edweek.org/teaching-learning/why-teachers-are-going-on- strike-this-fall-and-what-could-come-next/2022/09 Wilson, A. E., Parker, V. A., & Feinberg, M. (2020, August). Polarization in the contemporary political and media landscape. Current Opinion in Behavioral Sciences, 34, 223-228. https://doi.org/10.1016/j.cobeha.2020.07.005 Winkler, A. M., Scull, J., & Zeehandelaar Shaw, D. (2012, Oct. 29). How strong are U.S. teacher unions? A state-by-state comparison. Thomas B. Fordham Institute. https://fordhaminstitute.org/national/research/how-strong-are-us-teacher-unions-state- state-comparison Wood, R. (2019). [Photograph of teachers and support staff at Steven Tyng Mather School in Chicago]. USA Today. https://www.usatoday.com/picture- gallery/news/nation/2019/10/17/chicago-public-school-teachers-strike-after-failure- negotiate-contract/4006540002/ Yan, H. & Stix, M. (2018, Mar. 31). These teachers work up to 6 jobs. Now they’re fed up and ready to walk out. CNN. https://www.cnn.com/2018/03/31/us/oklahoma-teacher- 135 profiles/index.html York, M. (2018). [Photograph of teachers at Humphrey Elementary in Chandler, Arizona, participating in a statewide demonstration]. CNN. https://www.cnn.com/2018/04/25/us/teacher-walkouts-and-rallies-arizona-colorado- north-carolina-utah/index.html Zalubowski, D. (2018). [Photograph of man holding #RedforEd sign in a crowd of teachers protesting]. The New York Times. https://www.nytimes.com/2018/04/26/briefing/bill- cosby-michael-cohen-golden-state-killer.html Zaucha, A. (2019). [Photograph of Lauren Moody teaching a science class at Math and Science Prep]. The Wall Street Journal. https://www.wsj.com/articles/charter-schools-stay-open- as-striking-teachers-decry-their-growth-11548082825 Zdanowicz, C. (2019, Feb. 11). Denver is so expensive that teachers have to get creative to make ends meet. CNN. https://www.cnn.com/2019/02/10/us/denver-teacher-strike-multiple- jobs/index.html 136 APPENDIX A: 2018 AND 2019 RED FOR ED TEACHER STRIKE AND PROTEST TIMELINE Table A1. Timeline of Teacher Strikes and Protests during the Red for Ed Movement in 2018 and 2019 Location Dates Outcome West Virginia (1) 2/22/18-3/7/18 5% pay raise for teachers Increase in per pupil funding; did not stop 3/30/18, 4/2/18, Kentucky (1) passage of pension reform bill teachers 4/13/18 opposed Increased school funding and teacher pay via Oklahoma 4/2/18-4/12/18 tax increase passed before strike with no additional conditions after strike Some legislative gains for Colorado Education Colorado 4/26/18-4/27/18 Association backed programs, did not get changes to state employee retirement fund 20% teacher pay raise over next 3 years, Arizona 4/26/18-5/3/18 increased funding for support staff, new textbooks, upgraded tech and infrastructure Pueblo City School 2.5% cost of living adjustment and new health 5/7/18-5/11/18 District, CO insurance contribution from the district Promised up to 6.2% teacher raises before North Carolina 5/16/18 protest, but no additional increases due to activity Washington (state, 14 districts collectively Several districts provided significant pay raises, 8/2018-9/2018 & independently on such as 14.4% raises in Tacoma strike) Chicago, IL- Acero Pay raises for teachers and paraprofessionals, Schools charter 12/4/18-12/10/18 reduced class sizes, sanctuary for undocumented network students 137 Table A1. (cont’d) Location Dates Outcome 6% pay raise, reduced class sizes, more Los Angeles, CA 1/14/19-1/21/19 school support staff Los Angeles- Accelerated More job protections, improvement to health 1/15/19-1/25/19 Schools charter insurance network 5% pay raise for teachers and proposed Virginia 1/20/19 increases in school funding Chicago, IL- Pay raises for teachers and paraprofessionals, Chicago 2/5/19-2/18/19 reduced class sizes, paid paternal leave for International teachers charter network Denver, CO 2/11/19-2/14/19 Across-the-board teacher pay raise Stopped passage of an education bill that West Virginia (2) 2/18/19-2/19/19 would allow for charter schools and vouchers in the state Oakland, CA 2/21/19-3/1/19 11% pay raise plus one-time pay bonus Several legislative bills were not passed this 2/28/19, 3/6/19, session after sickouts; ongoing conflict over Kentucky (2) 3/7/19, 3/12/19 the Governor subpoenaed records of teachers who called in sick Chicago, IL- Reduced class sizes, additional school Chicago Public 10/16/19-11/1/19 support staff, 16% salary increase over 5 Schools years Note. Bold typeface indicates cases selected for study. Sources. Loewus, 2019 with supplements from Aguilar, 2018; Honeycutt Spears, 2019; McLaren, 2019; Pearson, 2012; Perez & Moreno, 2016; Smith & Davey, 2019; Turner, Lombardo, & Logan, 2018; Will, 2018a, 2018b, 2019a, 2019b; Yan, 2018 138 APPENDIX B: IMAGE CODING SCHEME Table B1. Chart of Social Constructions Codes Code Definition Examples I. Deserving Positive constructions of teachers that show teachers as • Teacher in a worthy of policy benefits or as benefitting general welfare banana (Ingram, et al., 2007, p. 101) costume Includes images that depict activities such as: dancing in the • Teachers working with children street with • Teachers hugging children sign • Parents/children/others support teachers by joining strike • Caption • Teachers celebrating policy wins mentioned • Teachers dancing while on strike activities to Includes teachers displaying positive emotions: keep energy • Happiness- smiling, looking relaxed, laughing up Setting includes: • Teachers celebrating • American flag, particularly foregrounded or carried/worn by teachers the end of a strike • Children present as either supporting teachers or • Teachers in positively associated with teachers (such as in a a classroom classroom with a teacher leading a lesson) with students Signs or captions focused on the following messages: • Students • Teachers deserving more pay making • Teachers on strike for students, to benefit students protest signs • Teacher working conditions as harming students • Taco truck • Teachers deserving respect feeding LA • Education funding as insufficient/inequitable teachers on • Referencing education as benefitting society as a whole the picket line Signs portraying one of the following common archetypes: • Sign with a • Superhero teacher- Either explicitly referencing teachers list of all the as superheroes, or more subtly, referencing teachers roles a teacher doing it all, teachers as heroically helping students, or fills similar. Similar to Bulkley & Gottlieb’s (2017) “counselor, “Individual Great Teachers Can Overcome All nurse, parent, Obstacles” policy image coach” etc. • Teachers combatting greater societal issues- Based on • “Teacher, Bulkley & Gottlieb’s (2017) “Profession of Teaching because Struggling Against Difficult Circumstances” policy superhero image, references to teachers working in adverse isn’t a job circumstances against poverty, trauma, and other social title” problems 139 Table B1. (cont’d) Code Definition Examples A. Borrowing Images that depict teachers as on strike "for the from Children children/students," with students, or otherwise associating the strike with benefits for children B. Ed as Image includes references to teachers serving society or the Public Good greater public good. This may include references to teaching the leaders of tomorrow, teaching critical skills, or teachers as fulfilling patriotic duty (or images associating patriotism with teachers). C. Respect or Image includes references to teachers deserving greater • “I paid for Economically respect, to teachers deserving greater pay or benefits, or to this sign with Worthy teachers needing to supplement their income my 2nd job” II. Negative constructions of teachers that show teachers as • Mayor Undeserving unworthy of policy benefits or as not benefitting/harming Lightfoot general welfare (Ingram, et al., 2007, p. 101) talking with a Includes images that depict actions such as: small group of • Teachers yelling or confronting people in an aggressive elementary manner students • Someone other than the classroom teacher engaging • Adult with students (politicians, substitutes, parents taking off monitoring a work to chaperone) gym activity • Kids alone somewhere with captions or context • Child and indicating they should be in school parent walking Includes teachers displaying negative emotions: hand-in-hand • Anger- Faces with narrowed eyebrows, furrowed up past picket brows, frowning or open mouths, flushed faces, “steam line to enter coming out of the ear” school Caption or image indicates frustration with teachers from • Kids at the other stakeholders: computer alone • Parents or others express frustration with the strike. • Parent in a References to the strike as harming/neglecting children coffee shop or being inconvenient. The caption may explicitly working with a reference this, the faces of stakeholders other than student teachers may show frustration (less intense anger), or • Caption the context of the photo may evoke the idea of children mentions alone or parents being burdened. parental Setting includes: challenges of • Empty classrooms or other school spaces, closed strike school door, suggesting teachers have abandoned children • Students in alternative spaces other than traditional classrooms (i.e., at home with parent, with the Boys & Girls Club or similar organization, in a large auditorium for class) 140 Table B1. (cont’d) Code Definition Examples II. Or signs/images/captions portraying one of the following Undeserving common archetypes: (cont’d.) • Not working full-time- References to the narrative that teachers have summers off, leave work at 3, or otherwise work less than others; may be used to justify less pay or benefits • Ineffective teachers- From accountability rhetoric, references to failing teachers, teachers unable to improve student outcomes. Similar to Bulkley and Gottlieb’s “Dysfunctional structures of teaching trump teacher quality” • Militant union- Beyond solidarity, though possibly including solidarity, images of anger, violence, disruption, or defiance associated with the union. This is inspired by Goldstein (2014) discussion of unions during the 1960-70s. III. High Teachers being shown as politically powerful- images that • Teachers Power evoke high levels of political resources: large, united, easy walking en to mobilize, skilled, focused on issues of concern to it, and masse down accustomed to voting, contacting public officials (Ingram, et Chicago streets al., 2007, p. 101) • Teachers Includes images that depict actions that suggest a unified, gather in a organized group effort such as: State Capital • Teachers marching and protesting Building • Teachers chanting • Caption • Teachers holding hands or showing other forms of shares chant mutual support • Several • Teachers or union holding a press conference teachers with • Teachers wearing red (or other union/matching clothes) mouths open, • Teachers present and engaged with possibly some speakers/rally/negotiations with fists Images that suggest power in numbers: raised • Karen Lewis • Images with more than 20 teachers/supporters, or where the frame is filled with teachers to suggest a walking away large group from a Images that invoke union power: microphone with cameras • Includes union leaders around her • Includes signs/apparel/or other indications of union • Teachers in organization red shirts • References contract negotiation • Drone photo • A large pile of strike signs to be picked up, produced of a sea of by union people 141 Table B1. (cont’d) Code Definition Examples III. High Image or signs invoke voting power: • In WV, blue Power • Signs threatening to vote politicians out of office or union shirts (cont’d.) frustration with officials • Teachers Images that invoke archetypes of: sitting in rows • Strong women- References to strong women or looking towards embracing both strength and femineity a presenter • Union solidarity- Examples of union members banding during together, such as hand holding or forming a picket presentation of line; wearing red clothes might fall under this contract • Militant union- Beyond solidarity, though possibly • A large group including solidarity, images of anger, violence, in a state capital disruption, or defiance associated with the union. This where you can is inspired by Goldstein’s (2014) discussion of unions see few distinct during the 1960-70s. faces • “Schools out. Justice Cut the Bull Out” • “I’m here more than Jim” • Signs referencing students as voters • “Rosie the educator” • Other non- teacher union members joining the strike • Teachers marching and holding hands • “United” signs A. Borrowing Images that invoke union power specifically, either Images that from Union through clear union presence, logos, captions, or other invoke union cues that reference the union power: Images that invoke archetypes of: • Includes union • Union solidarity- Examples of union members banding leaders together, such as hand holding or forming a picket • Includes line; wearing red clothes might fall under this signs/apparel/or indications of union 142 Table B1. (cont’d) Code Definition Examples A. Borrowing Images that invoke archetypes of: • References from Union • Militant union- Beyond solidarity, though possibly contract (cont’d.) including solidarity, images of anger, violence, negotiation disruption, or defiance associated with the union. This • A large pile of is inspired by Goldstein’s (2014) discussion of unions strike signs to during the 1960-70s. be picked up, produced by union Image or signs invoke voting power: • Signs threatening to vote politicians out of office or frustration with officials B. Borrowing Images where teachers seem to be borrowing power • Signs that say from specifically from the Red for Ed movement. #RedforEd #RedforEd • A sea of red clothing • In Arizona, Arizona Educators United is not a union –this is a group of teachers organizing the Red for Ed movement in this state IV. Low Power Teachers being shown as lacking political power. Often • Mayor the absence of teachers or the presence of others with Lightfoot power. talking at a Images showing non-teachers & supporters as powerful: podium • Political leader holding a press conference who is teachers leaving opposed to the strike without meeting Images that suggest a small strike or low levels of demands teachers participating: • Images with fewer than 20 teachers that also suggest the few people in the image are not part of a larger group/crowd; sparse participation 143 Table B1. (cont’d) Code Definition Examples IV. Low Power Images that show teachers disappointed by strike results, • Image with (cont’d.) see captions for details caption that Images that invoke the following archetypes: suggests • Missionary teacher- Teachers as serving others for the numbers of public good, but not necessarily deserving money or teachers other compensation; inspired by Goldstein’s (2014) participating historic depiction of early female teachers in the US dwindling, as missionaries of morals and democratic ideals teachers leaving without meeting demands 144 Table B2. Chart of Image Context Codes Code Definition Examples I. Role What is the role or identity of the people/stakeholders pictured? Roles determined by context and caption. Select all that apply. A. Teacher Adults participating in strikes who appear to be • Teacher group marching teachers; teachers working in classrooms; caption with signs labels subjects as teachers; Adult protesting • Teachers chanting without being identified as another group • Teacher sitting at a kitchen table looking reflective, and identified in caption 1. Union Teachers are identified as union members, • Teacher wearing a CTU member wearing badges/clothes referencing unions, hat carrying signs with union logo • Teacher carrying a sign with union logo • Teacher handing out signs with union logo to other teachers • Teachers standing behind union leader at a press conference- like the support on stage B. Student School-age children; may be participating in the • Child with a backpack strike, may be in out-of-school or in-school • Group of children in a activity; may be with parents gym while school closed • Child making a protest sign C. Parent Adults who appear to be parents or guardian; may • Adult in business jacket be dropping off children, holding hands with walking with child with children, etc. backpack • Adult sitting with child on laptop in coffee shop • Adult holding child on picket line with sign D. Union Adult who is in leadership role and identified • CTU president Karen Leader with union; may be identified by name in caption; Lewis may be wearing or holding union gear; may be • UTLA president Alex leading a press conference, holding a megaphone Caputo-Pearl leading chants, handing out union materials, etc. 145 Table B2. (cont’d) Code Definition Examples E. Political Adult who is an elected official or political See categories below Leader candidate; may be identified by name in caption, may be visibly recognizable 1. Local A mayor, school board member, or other local • Mayor Rahm Emmanuel elected official 2. State A governor, state legislator, or other state-level • WV Governor elected official • KY Governor 3. National The President, Senators, Representatives, and • Elizabeth Warren national political office candidates (esp. • Bernie Sanders Presidential candidates); may be identified in caption or visibly recognizable F. District or Superintendents, principals, or others who are not • Superintendent of School elected officials but who are also not teachers; district Admin typically administrators dealing with the logistical • Principal acting as fallout of a strike who may or may not be substitute in classroom involved in negotiation; likely identified in caption G. Union, Adult who is either wearing union gear that is not Examples: Non-Teacher the teaches’ union or caption indicates they are a • SEIU- often a union for union member other than the teachers’ union support personnel at schools • UAW members • Grad Student Unions H. No People The photo does not include images of people • A closed school door • An empty classroom • A pile of strike signs II. Gender What is the apparent gender balance in the photo? TEACHERS ONLY A. Single One person (either alone/within a group) presents female as female; The goal is to label both individuals in portrait shots and individuals who stand out as the lone representative in a large group. B. Single One person (either alone/within a group) presents male as male; The goal is to label both individuals in portrait shots and individuals who stand out as the lone representative in a large group. C. Majority Over 50% of the group presents as female, or female female individuals are foregrounded D. Majority Over 50% of the group presents as male, or male male individuals are foregrounded E. Diverse The group seems to present an equal balance of gender binary gender presentations 146 Table B2. (cont’d) Code Definition Examples F. Can’t Unclear from the photo what genders are • Sea of people that you determine represented in what proportions cannot see faces in gender III. Race What is the apparent racial balance of the photo? TEACHERS ONLY A. Single One person (either alone or within a group) POC presents as a POC The goal is to label both individuals in portrait shots and individuals who stand out as the lone representative in a large group. B. Single One person (either alone or within a group) white presents as white The goal is to label both individuals in portrait shots and individuals who stand out as the lone representative in a large group. C. Majority Over 50% of the group presents as POC, or POC POC individuals are foregrounded D. Majority Over 50% of the group presents as white, or white white individuals are foregrounded E. Diverse The image contains a range of people of different race races/ethnicity. There is not a clear majority. F. Can’t Unclear from the photo what racial identities are • Sea of people that you determine represented in what proportions cannot see faces in race IV. Strike Which strike is actually depicted in the photo? Pictured This may differ from the primary strike discussed in the article. Check captions carefully. A. CHI 12 Chicago Teacher Strike 2012 Sept 9-19, 2012 B. WV 18 West Virginia Teacher Strike 2018 Feb 21-Mar 8, 2018 C. OK 18 Oklahoma Teacher Strike 2018 April 1-13, 2018 D. KY 18 Kentucky Teacher Strike 2018 Mar 29-Apr 14, 2018 E. LA 19 Los Angeles Teacher Strike 2019 Jan 13-22, 2019 F. DEN 19 Denver Teacher Strike 2019 Feb 10-15, 2019 G. CHI 19 Chicago Teacher Strike 2019 Oct 16-Nov 1, 2019 147 Table B2. (cont’d) Code Definition Examples H. Other Any other teacher strike from 2018 i.e., North Carolina, 2018 ed Arizona, Colorado strike I. Other 2019 Any other teacher strike from 2019 i.e., West Virginia 2019, ed strike Kentucky 2019, Oakland 2019 J. Historic ed Image is not of current (as of news coverage) • Image of Al Shanker and strike strike, but from news archives. Often references a UFT leaders on strike in similar strike event. May be in black & white. NYC in 1968 • Images from the historic Strikes in education other than Chicago 2012, any 1960s FL teachers’ in 2018 or 2019 walkout K. Non-ed Image depicts a strike, but not a teacher strike • GM workers on strike strike 148 Table B3. Chart of Strike Sign Codes Code Definition Examples I. Handmade Signs seem to be handmade, may utilize a variety of art supplies and bright colors; often includes witty sayings or imagery; may invoke elementary classroom bulletin board II. Invoke Sign language or imagery invokes children. • “Fighting for our Children Students” Includes signs reflecting issues of: • “You can’t put students • Student needs- Sign relates strike demands to first if you put teachers how it will benefit the kids/learning. last” • Working conditions impacting students- ex. • “Fight for the schools "Our students deserve smaller class sizes" students deserve” • Funding impacting children- "For each child, • “Students, because CO spends $2700 less than the national average" you’re mine, I walk the line” III. Invoke Signs reference the union, either directly with the • Signs with the image of Union union name, or indirectly with labor union the raised fist imagery • Signs with CTU on them in Chicago IV. Invoke Signs the invoke any issue related to money: • “Keep great teachers in Economics • Bread & butter- The slogan or image on the sign Denver. Pay them.” is related to wages, pension, work hours, etc. • “3% that’s not a raise, • Working Conditions- Sign relates to working that’s an insult” conditions, infrastructure, overcrowding, lack of • “Teachers need more supplies, support personnel etc. than apples” • Sign focuses specifically on money for schools, • “Will teach for budget cuts, lack of money in the system insurance” • “my second job paid for this sign” • “Pack em deep, teach em cheap” • “For each child, CO spends $2700 less than the national average” • “teachers just want to have funds” • “school funding, not corporate tax cuts” • “starving the schools, feeding the rich” • “Unfair labor practices” • “Striking for a fair contract” 149 Table B3. (cont’d) Code Definition Examples V. Invoke Sign references the benefits of teachers to society • “If you can read this, Society as a whole you’re welcome” • “Teachers made your job For example, signs that include possible” • References to the future- "Fund our future," • “Fund our future” "Dinosaurs extinct, Teachers endangered" • “Teachers, we work for • Education as a public good- Sign expresses the people” importance of schools, state deserves better, we • “Fund/Care About Our owe it to the future, deserve better, ex. “Teachers, Future” we work for the people” & “Keep the public in • “Keep the public in our our public education” public education” • Role of teachers in educating society- teachers • “Dinosaurs Extinct, shape lives/careers, ex. "If you can read this, Teachers Endangered” you're welcome" & "Teachers made your job possible" VI. Invoke Signs indicate that the status quo is unfair/unjust • “Unfair labor practices” Fairness in some way. • “Striking for a fair contract” This includes unfair to teachers: • “Arm me with a fair • Signs invoking need for respect- Sign relates to contact” teachers deserving respect, being a profession, • “Teachers are worth teachers shape lives, etc. more” • Signs invoking social justice- Signs referencing • “We deserve better” social justice language and ideas; participation of • “school funding, not or reference to civil rights leaders corporate tax cuts” Note that WV Governor during 2018 strike is • “starving the schools, name Jim Justice. feeding the rich” Social Justice: • Caption- “Albert Shanker, center, president of the United Federation of Teachers, and the civil rights leader Bayard Rustin, right, outside City Hall during the 1968 New York City teachers’ strike.” • “Time for real Justice for Public Employees” 150 Table B3. (cont’d) Code Definition Examples VII. Invoke Includes signs with threats, anger, or other • “We’ve had enough” Frustration indication that teachers are frustrated. • “Schools out. Justice Cut the Bull Out” For example: • “I’m here more than • "Fed Up" signs- Signs referencing teachers Jim” reaching a breaking point • “Don’t make me use my • Political frustration- Signs referencing teacher voice” politicians doing a poor job, being responsible • “Don’t make me go for the state of schools, political all WVa on you” ineptitude/corruption, • “3% that’s not a raise, or politics otherwise negatively impacting that’s an insult” education • “Teacher burnout is why • Aggressive signs- Sign conveys anger or we have this turnout!” aggression on the part of the teacher 151 APPENDIX C: STAGES OF THE DICTIONARY BUILDING PROCESS Table C1. Political Power Theoretical Definition and Deductive Words to Examine Ingram et al., 2007 definition (p. Deductive suggested Present in 20% sample & 101) dictionary terms applicable 80% of time in context? “The political power of the target group is one dimension indicating the extent of its political resources, such as whether it is… large, large Present, not valid in context hundred(s) Included thousand(s) Included many Not present countless Present, not valid in context crowd(s) Included crowded Present, not valid in context numerous Not present united, unite/united Present, not valid in context unify/unified Present, not valid in context together Not present link/linked Not present collective/collectively Present, not valid in context easy to mobilize, mobilize/mobilized Not present assemble/assembled Not present marshal/marshalled Not present organize/organized Included organization Present, not valid in context driven to Present, not valid in context catalyst/catalyze/catalyzed Not present plan/planned/planning Present, not valid in context wealthy, campaign contributions Present, not valid in context money Present, not valid in context cost Present, not valid in context skilled, skilled/skillful Not present knowledge/knowledgeable Not present experienced Present, not valid in context understanding Present, not valid in context strategic/strategy Present, not valid in context 152 Table C1. (cont’d) Ingram et al., 2007 definition Deductive suggested Present in 20% sample & (p. 101) dictionary terms applicable 80% of time in context? well-positioned, well-positioned Not present important Present, not valid in context key Present, not valid in context focused on issues of concern to it, focus/focused Present, not valid in context concentrate/concentrated Not present engage/engaged Present, not valid in context and accustomed to voting, vote/voting Included contacting public officials, [and so voter Present, not valid in context on].” November Present, not valid in context polls Present, not valid in context elections Present, not valid in context 153 Table C2. Deservingness Theoretical Definition and Deductive Words to Examine Ingram et al., 2007 definition Deductive suggested Present in 20% sample and (p. 101) dictionary term applicable 80% of time in context? “It refers to the valence or the positive and negative social construction of the group… as more-or-less worthy and worthy Included deserving and as contributing worthwhile Not present more-or-less to the general deserve Included welfare” deserving Not present value/valued/valuable Not present respect Included respected Present, not valid in context appreciate/appreciated Not present admire/admired Not present beloved Not present esteemed Not present needed/necessary Present, not valid in context overworked Not present underpaid Included unworthy Not present undeserving Not present disrespect Included disrespected Not present unimportant Not present trivial Not present unnecessary Not present blame Included blamed Not present 154 Table C3. Lists of Initial Deductive and Inductive Words and Phrases derived from 20% sample of corpus- Political Power High Political Power Deductive Words or Phrases crowd organize thousand vote crowds organized thousands votes hundreds organizers voting organizing High Political Power Inductive Words or Phrases across the country join rallied teachers demanded across the nation marched rallies teachers gathered across the state marchers rally teachers rallied activism movement rebellion teachers walk activists movements RedforEd teachers walked advocate picket thousands of teachers tens of thousands of teachers advocates picket line threatened thousands of people agenda picket lines union’s demands walk battle picketed revolt walk off the job chanting picketing revolts walked demand pickets schools across the state walked off the job demand higher pay plan to walk statewide strike walkout demanded power statewide walkout walkouts demanding powerful stoppage wave of educator demands protest stoppages wave of teacher demonstration protested striking wave of teacher protests demonstrations protesters teachers across wearing red demonstrators protesting teachers across the work stoppage fight protest state work stoppages fighting teachers demand hundreds of teachers Low Political Power Inductive Words or Phrases concessions 155 Table C4. Lists of Initial Deductive and Inductive Words and Phrases derived from 20% sample of corpus- Deservingness Deserving Deductive Words or Phrases deserve respect underpaid worth Deserving Inductive Words or Phrases crumbling grassroots improvements supported dignity great teachers improving supporters dignity and respect improve make ends meet supportive dilapidated improved public support sympathetic dozens solidarity teacher of the year Undeserving Deductive Words or Phrases blame disrespect Undeserving Inductive Words or Phrases bad teachers closed fail shut bully closing failing shuttered canceled classes closures propaganda shutting canceling class criticized schools closed shutting down schools child care disappointed scramble teenager wanting a better car sexually trouble 156 Table C5. Final Social Constructions Words and Phrases Dictionaries, after validity check on full sample High Political Power Words and Phrases activism picket teachers demanded activists picket line teachers demands battle picket lines teachers gathered chanting picketed teachers rallied crowd picketing teachers vote to strike crowds pickets teachers voted demand higher pay plan to walk teachers walk demanding protest teachers walked demands protested tens of thousands of teachers demonstration protesters thousands demonstrations protesting thousands of people demonstrators protests thousands of teachers fight rallied threatened fighting rallies union’s demands hundreds of thousands rally walk marched rebellion walk off the job marchers Red for Ed walked off the job movement RedforEd walkout movements revolts walkouts organizers schools across the state wave of educator organizing statewide strike wave of teacher protests statewide walkout wearing red stoppage work stoppage stoppages work stoppages striking Deserving Words and Phrases crumbling disrespect supporters deserve improvements sympathetic dilapidated make ends meet teacher of the year dignity respect underpaid dignity and respect solidarity Undeserving Words and Phrases bad teachers closed sexually blame criticized shut canceled classes failing shuttered canceling class propaganda shutting down schools child care schools closed teenager wanting a better car scramble trouble 157 APPENDIX D: WORD STAT DICTIONARIES WORD FREQUENCY ANALYSIS Table D1. Distribution of Key Words across the Red for Ed text analysis dataset (N=486) Dictionary Word and Phrase Counts Frequency (word count) Cases n % n % High Political Power 5018 1.13 473 97.33 Deserving 509 0.11 262 53.91 Undeserving 480 0.11 250 51.44 Teachers 8804 1.98 486 100 Children 3582 0.80 445 91.56 Union 3439 0.77 428 88.07 Political Leaders 1961 0.44 378 77.78 Parents 680 0.15 274 56.38 Administrators 329 0.07 138 28.40 158 Table D2. Distribution of Key Words across the full text analysis dataset (N=617) Dictionary Word and Phrase Counts Frequency (word count) Cases n % n % High Political Power 5690 1.02 586 94.98 Deserving 604 0.11 323 52.35 Undeserving 613 0.11 322 52.19 Teachers 10,858 1.94 617 100 Children 4750 0.85 571 92.54 Union 4700 0.84 552 89.47 Political Leaders 2683 0.95 495 80.23 Parents 979 0.18 350 56.73 Administrators 395 0.07 178 28.85 Figure D1. Distribution of Deserving Dictionary Words and Phrases (frequency) 159 Figure D2. Distribution of Undeserving Dictionary Words and Phrases (frequency) Figure D3. Distribution of Most Frequent (occurred more than 50 times) High Power Dictionary Words and Phrases 700 589 600 500 400 Frequency 303 300 257 249 234 232 225 217 209 193 200 169 158 156 148 141 118 111 104 103 96 88 82 100 68 66 65 65 63 62 54 54 54 52 0 W M KO O A L VE U T M DPR N T O EM ST ATEE ST W RI ND A S LK PR IN O TEK RA S ST G W D KOA L FI LL G Y HT PI EM U T H ND O A TS US G IN CK A N DET D S LI EM S N E RE A N DF D FI RED G O HT TH PR PI OW NG A LK TE ED S I O USA ND S CK TIN A OF TI E ET G C N LI TE AC M BA SV H IS ER D EM O R NS A LL TR AT STT LE IE W O O M ION RK RC SA ST HE O P D RG PA R O A LLA GE NI ZI N G RG A CK NI D PI Z ER IE S PR ET I O PR E S O T TE E DT NG ST ER S High Power Dictionary Words and Phrases 160