Conduit or contributor? The role of media in policy change theory

Transcription

1 DOI /s y Conduit or contributor? The role of media in policy change theory Elizabeth A. Shanahan Æ Mark K. McBeth Æ Paul L. Hathaway Æ Ruth J. Arnell Ó Springer Science+Business Media, LLC Abstract The policy change literature is contradictory about the role the media plays in policy change: a conduit for policy participants, with media accounts transmitting multiple policy beliefs of those involved in policy debates or a contributor in the policy process, with media accounts supplying consistent policy beliefs with congruent narrative framing strategies to construct a policy story. The purpose of this study is to empirically test whether the role of the media is that of a conduit or contributor in the policy change process. This study tests whether there are differences in policy beliefs and narrative framing strategies between local and national print media coverage of two contentious policy issues in the Greater Yellowstone Area between 1986 and 2006, that of snowmobile access and wolf reintroduction. In the Greater Yellowstone Area policy arena, local media accounts are believed to be aligned with the Old West Advocacy Coalition, whereas the national media accounts are thought to be part of the New West Advocacy Coalition. With a methodology informed by narrative policy analysis, one hundred seventy five local and national print newspaper accounts were content analyzed to determine whether these media accounts were policy narratives, with embedded policy beliefs and narrative framing strategies. The results indicate that there are statistical differences between local and national media coverage for five of the seven hypotheses. Media accounts are generally policy stories, suggesting that the media s role is more of a contributor than a conduit in the policy change process. E. A. Shanahan (&) Montana State University, Bozeman, USA shanahan@montana.edu M. K. McBeth R. J. Arnell Idaho State University, Pocatello, USA mcbemark@isu.edu R. J. Arnell arneruth@isu.edu P. L. Hathaway Jacksonville State University, Jacksonville, USA hathpaul@isu.edu

2 Keywords Media Policy change Narrative policy analysis Content analysis Greater Yellowstone Advocacy coalition New West Old West While there is growing attention in the policy sciences literature centering on the media (Howland et al. 2006; Terkildsen et al. 1998), policy scholars have not adequately differentiated the role of the media in the policy change process. Some theorize that the media is a conduit for policy actors, whereby media accounts transmit multiple policy preferences in the coverage of policy debates (e.g., Baumgartner and Jones 1993; Iyengar 1997). Others assert that the media is a contributor in the policy process, a source of a particular policy preference (Sabatier and Jenkins-Smith 1993). Policy scholars have focused most of their attention on the role of interest groups, elected officials, and citizens in describing the actors involved in the policy change process and have left the role of the media empirically unexamined. In this study, we test whether the role of the media is that of a conduit or contributor in the policy change process. Through the integration of narrative policy analysis (NPA) with that of Advocacy Coalition Framework (ACF) policy change theory, we seek to add to the policy sciences literature by determining whether the media construct policy stories that purport, at its core, policy beliefs with congruent narrative framing strategies. Two recent studies inform the methodological framework of this work. First, McBeth et al. (2005) develop a methodology based on NPA that systematically measured core policy beliefs in the policy narratives of interest groups in different advocacy coalitions. Second, McBeth et al. (2007) determine that these interest groups also use similar political narrative strategies, depending on whether they understand themselves to be winning or losing in the policy debate. The narrative strategies in this study are anchored in the media effects framing literature and tested not for their presence in the story but for their alignment with the narrative s policy beliefs. Thus, in this study, we develop a methodology informed by NPA to analyze media narratives for policy beliefs and narrative framing strategies. The central research question of this study is: Does the media serve as a conduit for policy actors, reporting multiple policy preferences in newspaper accounts or as a contributor in the policy debate, constructing policy stories that harbor consistent policy beliefs and narrative framing strategies? The import of this study lies at the heart of how Lasswell (1971) and others have advanced the study of policy sciences: [to refine] the intellectual tools needed to support problem-oriented, contextual, and multi-method inquiry in the service of human dignity for all (Brunner 1997, p. 191). This tripartite focus cast within the context of advancing democracy anchors this study. First, policy analysts are necessarily bounded, as the policy issues they investigate empirically may exclude important factors that shape policy dilemmas (Brunner 1991). This study examines to what extent it is important for policy analysts to account for the role of the media as a source of influence and power in the policy change process. Second, GYA policy issues in this study are examined in the context of a socially, politically, and economically changing western region, with divergent policy values creating a wicked policy environment. Our research question tests how the GYA s contentious policy context is played out in local and national media accounts. Third, this study represents the convergence of epistemologies discussed at length in Brunner (1991, pp ) to include the study of contextually communicated problem definitions through content analysis, a mode of inquiry heralded by Lasswell (see Janowitz

3 1968) in the policy sciences. Finally, this study builds on Lasswell s belief that policy knowledge and understanding normatively should improve the practice of democracy (1951, p. 15). As Brunner (1997, p. 208) suggests, citizens are capable of informed, sound judgment when they have access to reliable facts and interpretations prepared by experts. The nature of how the media the primary disseminator of information to citizens portrays facts and expert opinion is crucial in understanding the nature of what informs their opinions. If indeed, the media is a contributor of policy stories, such a situation most likely increases divisiveness, conflict, and policy stalemate and the ideals of democracy based on discourse, information, and rational persuasion are lost. In this regard, the media is a policy marketer contributing to policy intractability, with policy solutions that offer no long term resolution of the issue (McBeth and Shanahan 2004). In sum, this study meets the goals and standards originally set out by Lasswell and others in the development of the policy sciences. To address the research question, we compare local and national print media coverage of two Greater Yellowstone Area (GYA) contentious policy debates to ascertain whether there are, in fact, differences in policy beliefs (e.g., federalism, view of human and nature relationship) and differences in supporting narrative framing strategies (e.g., source cues, problem definition) in media representation of these policy issues (see Fig. 1). If the media is a mere conduit of policy preferences, then the divergent policy beliefs and narrative framing strategies of the two competing GYA advocacy coalitions will appear across both the local and national media articles, revealing no differences; conversely, if the local and national media accounts do show differences in policy beliefs and narrative framing strategies, then these outlets are likely contributing to the competing advocacy coalitions in the GYA policy arena. We begin with a discussion of the theoretical anchors that inform our work. We first examine how traditional policy change theories and NPA literature have dealt with the GREATER YELLOWSTONE AREA POLICY SUBSYSTEM Conceptual Framework for the Media as a Contributor New West Advocacy Coalition (NWAC) Old West Advocacy Coalition (OWAC) Membership includes environmental groups, tourists, Membership includes all terrain vehicle groups, extractive Yellowstone National Park employees, and national media outlets. industries, ranchers, farmers, and local media outlets. This NWAC coalesces around these core policy beliefs: This OWAC coalesces around these core policy beliefs: 1. national theory of federalism 1. compact theory of federalism 2. biocentric view of human nature 2. anthropocentric view of human nature 3. expertise: conservation-solution science 3. expertise: technology-solution science The NWAC advocates for The OWAC advocates for Policy Alternative A Policy Alternative B (ex: reintroduction of the wolf; (ex: no wolves in the GYA; no/limited access of snowmobiles) access for snowmobiles to YNP) The NWAC s congruent narrative framing strategies: The OWAC s congruent narrative framing strategies: 1. New West (anti-snowmobile and pro-wolf) source cues 1. Old West (pro-snowmobile and anti-wolf) source cues 2. New West (anti-snowmobile and pro-wolf) descriptors 2. Old West (pro-snowmobile and anti-wolf) descriptors 3. framing format 3. framing format 4. contextual problem definition (e.g., environmental) 4. contextual problem definition (e.g., economic) End result: policy intractability, as the core values of these coalitions divide the publics with incomplete and policy driven narratives Conceptual Framework for the Media as a Conduit The national and local media outlets portray all policy beliefs and use varied framing strategies. Policy Alternatives A and B End result: policy learning in a democratic context, as policy alternatives are reported to a citizenry, capable of informed opinion Fig. 1 Greater Yellowstone area policy subsystem

4 question of the role of the media. We then focus on the media effects literature to elucidate the elements of framing in a media story. Finally, we discuss the significance of utilizing the contentious policy environment of the GYA as a case study for the investigation of the role of the media. Contemporary policy theory, narrative policy analysis and media effects literature Contemporary policy change theory is led by Kingdon (2003), Sabatier and Jenkins-Smith (1993), and Baumgartner and Jones (1993). A less mainstream, but still highly influential, strand of policy theory is narrative policy analysis (NPA), led by Stone (2002) and Roe (1998). We begin with a careful examination of how these policy theories understand the role of the media in the policy change process. In turn, we consider the extensive literature of media effects that investigates the various ways in which the media frames stories. Media and policy change theory In the mainstream policy change literature, the media s role is empirically underdeveloped and contradictory. The media is, at times, identified as a conduit for political actors and policy entrepreneurs in the policy subsystems to disseminate policy beliefs and, at other times, acknowledged as a contributor, pursuing policy beliefs in its own right to influence policy change. Baumgartner and Jones (1993, pp ) identify the media as a conduit, reflecting policy arguments in the policy subsystems. They contend that the result is a lurching of positive to negative tones on a specific issue in media accounts. The authors descriptively demonstrate the relationships between how an issue was covered (positive or negative) and what the focus was (economic, environmental, or governmental topic). The media is thus understood as a conduit for policy entrepreneurs, who use the media to transmit their issue definition in order to mobilize groups and citizens to achieve policy success (see also Paletz 1998, pp ). The Advocacy Coalition Framework (Sabatier and Jenkins-Smith 1993) briefly identifies the role of the media as both a conduit a resource for members to influence policy outcome (p. 227) and a contributor a member of competing advocacy coalitions (p. 183). In their extensive work, the careful, systematic coding of the advocacy coalitions did not include the media, and yet the authors anecdotally recognize both roles of the media in the policy arena. Clearly, there is a need for empirical examination to determine whether the media plays the role of conduit in the policy debate or contributor as an advocacy coalition. Kingdon (2003) identifies the media as a conduit, rather than a contributor in the policy arena. The media report what is going on in government... rather than having an independent effect on governmental agendas, (Kingdon 2003, p. 59). Given this assumption of the role of the media, he purports a diluted effect of the media on the policy agenda for two reasons (pp ). First, the press has a tendency to cover stories for a short period of time. Second, the media s propensity to feature dramatic stories means that the policy highlighted tends to be at the end of the policy making process, rather than the beginning. Thus, according to Kingdon, the agendas were already set, and the media had minimal impact. Each of the major policy change theorists cast the media as having an important role in the policy arena. However, there is neither consensus on the nature of its role nor are there

5 any empirical analyses of this role. These differences in assumptions about the role of the media have not been disputed in the conventional policy literature, as they simply have not been examined. To test the nature of the role of the media, we rely on Sabatier and Jenkins- Smith s (1993) ACF theory that purports that advocacy coalitions are united by policy beliefs. McBeth et al. (2005) found that policy beliefs are embedded in interest group narratives and are a source of empirical study. Thus, any study that examines the question of whether the media is an active contributor to an advocacy coalition would have to empirically test for policy beliefs embedded in media accounts. Media and narrative policy analysis Similar to their mainstream counterparts, narrative policy analysis theorists tend to include discussions of the media in the policy change process in a tangential manner. While the media is not brought center stage as a policy actor, both Roe (1998) and Stone (2002) allude to the media as conduits not contributors in the policy change arena. In his discussion of what constitutes policy narratives, Roe (1998) primarily examines the constructs not the purveyors of such stories. When he identifies the media in his case studies (e.g., pp ), he casts the media as a conduit of policy change, akin to the Baumgartner and Jones (1993) perspective. Certain groups were successful at getting the media to transmit their policy perspective, which, in turn, affected public opinion and policy outcome. Similarly, Stone (2002) does not directly address the role of the media in policy change, and she also portrays the media as conduit rather than as contributor. She argues that the media is a medium used by various groups in indoctrination (pp ) and a key element used by assorted groups to promote their causal stories (p. 203). The fact that Roe and Stone have not substantively explored the role of the media in policy controversies represents a significant gap in the NPA literature. To test the nature of the role of the media, we rely on Stone s (2002) contention that participants in the policy process engage in constructing political narratives that are calculated arguments, filled with narrative strategies to define a problem in order to capture support for the group s proposed policy solution. For example, groups cast different entities in their policy stories as their heroes (those who can help make the desired policy outcome occur), villains (those who are trying to stop the desired policy outcome from occurring) and victims (those who suffer from the undesirable policy outcome). Similarly, policy stories are filled with symbols, colorful adjectives, and metaphors to describe policy issues in favor of a particular policy outcome. If newspaper articles are also policy stories, then they should reflect a strategic use of narrative strategies that align with policy beliefs. Media effects literature The media effects literature serves as a framework for examining media accounts for policy beliefs and narrative framing strategies. Media effects are addressed widely in contemporary communication, public opinion, and political science literature under general headings such as information dissemination, priming, agenda setting, and framing. While the effects of information dissemination (see Graber et al. 1998; Iyengar 1997; Lippman reprinted in Graber 1990; McCombs 1997), priming (Iyengar and Kinder 1987; Iyengar and Simon 1993; Miller and Krosnick 2000; Roskos-Ewoldsen et al. 2002), and agenda setting (e.g., Iyengar 1990; McCombs and Shaw 1972) are well documented in the

6 literature, for the purposes of testing for core policy beliefs and narrative strategies in media accounts, we focus on the media effect of framing. Media scholars define framing as the media s power to influence how people comprehend issues by using interpretive frames to cover the issue (Carpini 2005; Entmen 1993; Gamson and Modigliani 1987; Iyengar 1991; Takeshita 2005); it is a thematic shorthand, structuring policies and intimating political solutions (Terkildsen et al. 1998, p. 47). While many of the relevant media effects studies make references to the policy agenda, there is a definitive lack of connection in this body of literature with mainstream policy change theory. Here, we focus on two major elements to framing in our research design: linguistics (source cues, symbolic language, and problem definition (framing format). Source cues and adjectives employed by the journalist are linguistic devices that serve to frame a policy issue (Terkildsen et al. 1998). Who the media uses as sources of information and how they characterize the issue influences the perspective of the policy at hand. For example, a policy issue can remain localized through exclusive media use of local sources. Conversely, an issue is expanded as the media cites national source cues. Gamson and Modigliani (1987) and Stone (2002) also identify the power of symbols in framing policy debates, as symbols can evoke powerful emotions that serve to either expand or contain the scope of the policy issue. Iyengar and Simon (1997, p. 250) argue that episodic or thematic frames influence problem definition, which informs the parameters of the policy debate. The episodic frame involves concrete or specific instances that illustrate an issue whereas the thematic puts issues in a general or abstract context to present evidence. The authors discover that the news media overwhelmingly relies on episodic framing (see also Collins et al. 2006). This conclusion may be partially explained by the desire for storytelling, which intrigues the reader with individual characters strategically portrayed as heroes, villains, and victims. Ultimately, the type of media framing format influences how recipients attribute responsibility for problems. It is through the use of elements of media framing (source cues, symbolic language, and framing format) that we test whether media accounts are policy stories, harboring internally consistent policy beliefs with accompanying narrative framing strategies. The case study To test whether the role of the media is a conduit or contributor, we use policy controversies in the Greater Yellowstone Area (GYA) as our case study. There are two reasons for situating this study in the GYA policy arena. First, the area receives an abundance of press coverage from both local and national media outlets. The local press coverage emerges from the communities in the GYA, which encompasses Yellowstone National Park (YNP) and the surrounding lands. These GYA gateway communities are located in expansive rural states with economic, cultural, and social stakes in the decisions made by the National Park Service. The national coverage is generated by broad interests in policy outcomes for YNP, the world s first national park. Second, the contentious nature of the policy debates and alternating victories and losses between the two opposing advocacy coalitions in the GYA means that there is no policy monopoly. In other words, the two dominant advocacy coalitions in the GYA are both significantly powerful and have the resources to successfully get their message out to a variety of different groups and citizens. Thus, in this highly contentious and competitive policy environment, if the media were

7 mere conduits, then we would expect to read competing policy narratives, reflective of the divergent policy beliefs of opposing advocacy coalitions. Who are these two opposing advocacy coalitions in the GYA? We refer to one as the Old West Advocacy Coalition and the other as the New West Advocacy Coalition. Previous studies have indicated that these two advocacy coalitions are comprised of various interest groups, elected officials, and citizens (McBeth et al. 2005; Morris and McBeth 2003). Broadly speaking, they constitute New West-centered (in-migrants, urban, professionals, service based economy, environmentalism) and Old West-centered (long-term residents, rural, working class, natural resource-based economy) values and represent deep cultural, social, and political divisions (Tierney and Frasure 1998; Wilson 1997). Importantly, the Old West Advocacy Coalition and the New West Advocacy Coalition are present across GYA issues, although their membership may change slightly with each policy issue. For example, both the Blue Ribbon Coalition and the Idaho Anti-Wolf Coalition are part of the larger Old West Advocacy Coalition but contribute only to the snowmobiling and wolf-reintroduction policy issues, respectively. Similarly, the Greater Yellowstone Coalition and the Wolf Fund are part of the larger New West Advocacy Coalition but contribute primarily to snowmobiling and wolf-reintroduction issues, respectively. In accordance with the advocacy coalition framework (ACF), these two advocacy coalitions are internally united by their own, respective policy core beliefs, which are, in turn, applied by members to a specific policy issue. Recent research (McBeth et al. 2005) has teased out policy core beliefs from the Old West and New West coalitions interest group narratives. In the contentious policy arena of the GYA, ACF core policy beliefs fall on the following continua: (1) nationalism compact theories of federalism, 1 (2) conservation biology technical fix centered science, 2 and (3) biocentric anthropocentric view of the relationship between humans and nature. 3 The Old West Advocacy Coalition leans toward these policy beliefs: compact theory of federalism, technical fix-centered science, and an anthropocentric view of humans and nature; the New West Advocacy Coalition adheres more closely to these policy beliefs: nationalism theory of federalism, conservation biology-centered science, and a biocentric view of humans and nature (McBeth et al. 2005). In this study, we test for Old West and New West policy value and narrative framing strategy differences between local and national newspapers, respectively. The idea that there are differences between local and not-local media accounts has some basis in the policy change literature. In examining an environmental policy debate in the Lake Tahoe basin, Sabatier and Jenkins-Smith (1993, p. 183) hypothesized that in-basin media outlets had divergent policy beliefs from the out-of-basin media outlets. In this case study, we empirically test for differences between local and national media accounts in the GYA. If the media is a conduit, then both local and national media accounts should reflect both Old West and New West Advocacy Coalitions core policy beliefs. On the other hand, if the local newspapers told their stories from the perspective of the Old West Advocacy Coalition, and the national newspaper articles primarily tell their stories from the 1 Nationalism theory supports the notion that decisions should be made at the federal level; compact theory supports decisions being made locally. 2 Conservation science is characterized by natural management, habitat and ecosystem protection, and biodiversity; technical fix centered science is characterized by management of the environment through technological innovation and the productive capacity of natural resources. 3 The biocentric view of the human nature relationship means that nature has an intrinsic value, on par with humans; the anthropocentric view of this relationship means that human concerns/wellbeing take precedence over concerns/wellbeing of nature.

8 perspective of the New West Advocacy Coalition, we could conclude that the media is likely part of these respective advocacy coalitions and a contributor in the policy process. There is abundant local and national media coverage of three GYA policy issues in which these competing advocacy coalitions are deeply engaged: snowmobile use/access to YNP, wolf reintroduction/management, and bison management. Given the volume of articles covering all three major policy issues, this study examines two of them: snowmobile use inside YNP and wolf reintroduction/management. Snowmobile access and wolf reintroduction in Yellowstone National Park The issue of snowmobile access to YNP is an on-going controversy with a history of law suits, court decisions, and administrative rulings that have resulted in decisions that were overturned, reversed, and reinstated. The National Park Service s winter use plan has gone through many iterations, ranging from no regulation of snowmobiling in YNP between 1963 and 1997 to total phase out of snowmobile access by 2003 to caps and cleaner machines in The reintroduction of wolves to Greater Yellowstone in 1995 ranks as one of the most studied environmental policy events of the last 20 years (e.g., Fischer 1995; Klein 2002; Nie 2003; Phillips and Smith 1996; Wilson 1997). The wolf was not protected under the Lacey Act of 1884 and by the 1930s was eradicated from YNP first by the US Army and later by the National Park Service. Between 1989 and 1994, the idea of reintroducing wolves to YNP went through administrative procedures and studies, concluding that wolves could be introduced but also managed (i.e., shot if they bother livestock outside YNP). In 1995, wolves were transported from Canada and released in YNP, and in 2004 they were delisted from the endangered species list. For both the snowmobile access and wolf reintroduction issues, there are local and national interests active in these policy debates to influence the policy outcome through funding efforts, lobbying, and spinning. Importantly, the media coverage of the contentious and continuing political activity means that there is a running record of local and national print media of these issues over the last two decades. Additionally, the contentious nature of these GYA policy debates in which no one advocacy coalition has a policy monopoly provides an excellent case to empirically test whether the media is a conduit or a contributor in the policy process. Methodology Drawing from media effects theories and narrative policy analysis to quantify policy beliefs and narrative framing strategies, this study examines print media coverage of two GYA policy issues between January 1985 and June A content analysis was performed on 175 articles found in local and national newspapers (see Table 1). Given that the central research question of this study focuses on whether print media accounts are policy stories, the sample included only non-editorial news articles written by the newspaper staff writers; excluded were Associated Press articles, editorials, and guest columns. Only articles addressing one of two policy issues in the Greater Yellowstone Area were used: (1) wolf reintroduction (93 documents) and (2) snowmobile access in YNP (82 documents). The population of national articles reflects the census of accounts published in the New York Times (herein NYT; 43 documents) and USA Today (herein USA; 31 documents) on

9 Table 1 Descriptive statistics of news outlets by issue New York Times USA Today West Yellowstone News Cody Enterprise Snowmobile access 20 (47%) 14 (45%) 31 (76%) 17 (28%) Wolf reintroduction 23 (53%) 17 (55%) 10 (24%) 43 (72%) Total 43 (100%) 31 (100%) 41 (100%) 60 (100%) Snowmobile access Wolf reintroduction Total National coverage 34 (41%) 40 (43%) 74 Local coverage 48 (59%) 53 (57%) 101 Total 82 (100%) 93 (100%) 175 Sources: New York Times, USA Today, West Yellowstone News, and Cody Enterprise, January 1985 June 2006 these two GYA issues. The national articles were gathered using LexisNexis, which reliably archives articles for both papers. Two YNP gateway community newspapers were chosen for the local media outlets: The West Yellowstone News (herein WYN) and The Cody Enterprise (herein CE). Articles from these newspapers were garnered primarily through archival research in the gateway community libraries and interlibrary loan of microfilm. Not surprisingly, the local articles coverage of the two GYA policy issues appeared at a much higher rate (271 documents) than the national coverage; thus, a stratified random sample (n = 101) was used to draw a sample of the local articles. The sample was stratified to reflect (1) the proportion of the total documents generated by each newspaper (40% were WYN and 60% were CE) and (2) the proportion of articles covering the snowmobiling and wolf reintroduction issues within each respective newspaper (WYN: 76% snowmobiling and 24% wolf; CE: 28% snowmobiling and 72% wolf). The variables used to empirically test for differences in policy beliefs in local and national newsprint media coverage are (1) federalism, (2) view of the human and nature relationship, and (3) type of science used. The variables used to test for differences in narrative framing strategies are: (1) New West Old West source cues, (2) New West Old West descriptors, (3) format of framing and (4) problem definition (see Table 2). Six of the seven variables are continuous, and one is nominal. Table 2 provides the equations that were developed for the continuous variables, with the idea that on each scale is a perfect Old West policy story and a means a perfect New West policy story, 4 representing the two competing advocacy coalitions in GYA. Policy beliefs Federalism Nationalism theory of federalism is the belief that the federal government should have decision making authority over local governments in policy decision making; compact theory of federalism is the inverse of nationalism theory, with a conviction that decision making authority should reside at local levels of government. In this study, federalism was content analyzed with the type of source cues used. Source cues are the anchor points of 4 The scale from to representing Old West to New West policy orientation does not apply to the variable Framing Format.

10 Table 2 Definitions of variables Variable Definition Policy beliefs Federalism Total # of national source cues total # of local source cues Total source cues National source cues: interest groups, elected officials/judges, agencies, individuals/businesses, and science that are outside of MT, ID, and WY Local source cues: all interest groups, elected officials/judges, agencies, individuals/businesses, and science that are from MT, ID, and WY Continuous variable: (all local) through (all national) Human/nature Total # of biocentric victims total # of anthropocentric victims relationship Total victims Anthropocentric victim: the victim is human-centered Biocentric victim: the victim is environment-centered Continuous variable: (anthropocentric) through (biocentric) Type of science Total # of conservation science total # of technology-based science Total scientific studies Conservation science: environmental issues are addressed with conservation or biology-based solutions Technology science: environmental issues are addressed with technological or human-based solutions Continuous variable: (technology) through (conservation) Narrative framing strategies New West/Old West Total # of New West source cues total # of Old West source cues source cues Total source cues New West source cues: entities which can be identified as pro-wolf reintroduction; anti-snowmobile access Old West source cues: entities which can be identified as anti-wolf reintroduction; pro-snowmobile access Continuous variable: (all Old West) through (all New West) New West/Old West Total # of New West descriptors total # of Old West descriptors descriptors Total descriptors New West descriptors: entities which can be identified as pro-wolf reintroduction; anti-snowmobile access Old West descriptors: entities which can be identified as anti-wolf reintroduction; pro-snowmobile access Continuous variable: (all Old West) through (all New West) Framing format Total # of thematic framing total # of episodic framing Total framing Thematic framing: the victim is general (e.g., the economy; wildlife) Episodic framing: the victim is specific (e.g., Joe s rentals; Boris the bison) Continuous variable: (all episodic) through (all thematic) Problem definition How does the media account define the problem? Nominal variable: economic, environmental, political, combination

11 the media article s information; the constellation of these source cues in an article is believed to reveal the policy value of federalism. Thus, national source cues are believed to be reflective of nationalism theory of federalism while local source cues are thought to be reflective a compact theory of federalism. These theories of federalism correspond to New West and Old West policy beliefs, respectively. Human nature relationship How a narrative presents the victim of the undesired policy outcome sheds light on the perspective of the relationship between humans and nature. If the victim of snowmobile access is wildlife, then the narrative engenders a biocentric view; conversely, if the victim of snowmobile access is a gateway community s local business, then the narrative has an anthropocentric view. This study content analyzes media accounts to determine how victims identified in the policy story are characterized biocentric or anthropocentric. The former corresponds to the New West policy beliefs, while the latter aligns with the Old West policy beliefs. Type of science Because science is generally thought of as the heralding of expertise, scientific results are used in policy debates to bolster a policy claim as the best one. However, there are often competing or contested scientific results. Thus, the citation of different types of science uncovers divergent policy beliefs. Science that is characterized by biodiversity and ecosystem protection (conservation science or conservation biology) is thought to harbor New West policy beliefs; conversely, science that embodies technological innovations that direct and manage the environment (technical science) is believed to be Old West in its policy beliefs. This study content analyzes the media accounts for what kind of science is cited. Narrative framing strategies New West Old West source cues The entity or individual a journalist cites as the purveyor of information is a source cue, which, in turn, frames the policy issue. Framing is the use of interpretive constructions that influence how people comprehend the issue. Source cues are categorized as New West (anti-snowmobile; pro-wolf) and Old West (pro-snowmobile and anti-wolf). This study content analyzes the extent to which the source cues used reflect New West or Old West sources. New West Old West descriptors Use of descriptors (i.e., symbols, adjectives, metaphors) can evoke powerful emotions and serve to contain or expand the issue in the policy system. Use of New West descriptors (pro-wolf; anti-snowmobile) and Old West descriptors (anti-wolf; pro-snowmobile) reveal a story s narrative framing strategy of a particular policy orientation. This study content analyzes the media accounts for use of these descriptors of the policy issue.

12 Framing format Episodic and thematic frames were determined by coding whether the victim in the article was specific (episodic) or general (thematic). While previous research has indicated that most media accounts use episodic framing to engage readers, this study tests for differences in framing format use between local and national media accounts. Contextual problem definition Given that framing strategies influence problem definitions, this nominal variable was open coded and further categorized into four broad categories by the researchers: economic, environmental, political, and combination. The content analysis was conducted by two experienced and trained coders using standard content analysis procedures (Johnson and Reynolds 2005, pp ). By developing a codebook (see Appendix A) and carefully reading each article, the data produced are much more rich and nuanced than a simple word count produced with content analysis software. We agree with Howland et al. (2006) who argue elegantly in praise of their human coders. Coding can be time consuming, but what our method lacks in efficiency, it makes up in thoroughness and transparency (p. 229). Additionally, the use of content analysis allows for a reliability test (see Appendix B) and permits longitudinal analysis. The reliability analysis was similar to that produced by Sabatier and Brasher (1993, p. 187). In this study, there is satisfactory inter-coder reliability, with the coders agreeing exactly 80% of the time or exact and within one value of each other 94% of the time. Hypotheses Based on our review of policy change theory, narrative policy analysis, and media effects theory, we propose seven hypotheses to test whether media accounts are, in fact, policy stories. If we find no differences in core policy beliefs and narrative framing strategies between local and national media then we can conclude that the media is simply a conduit for the two GYA Advocacy Coalitions and the media accounts reflect the policy divisiveness of GYA. Conversely, if the local and national media tell different stories, this suggests that the media are active contributors in the GYA advocacy coalitions. The first six hypotheses are tested using a t-test to compare difference of means on policy beliefs and supporting narrative framing strategies in the local and national print media coverage. Given nominal-level data, the seventh hypothesis is tested using Chi-Square analysis to determine if there is an association between local and national media coverage and problem definition. Policy beliefs null hypotheses H 01 There is no difference of means between local and national articles in federalism (use of locally-based and nationally-based source cues). H 02 There is no difference of means between local and national articles in their view of the relationship between humans and nature.

13 H 03 There is no difference of means between local and national articles in type of science cited. Narrative framing strategies null hypotheses H 04 There is no difference of means between local and national articles in use of Old West and New West source cues. H 05 There is no difference of means between local and national articles in use of Old West and New West descriptors. H 06 There is no difference of means between local and national articles in use of framing format (episodic and thematic). H 07 There is no association between local and national articles in use of problem definition (economic, environmental, political, or combination). Results We reject five out of the seven null hypotheses. Table 3 indicates the results for the difference of means tests; Table 4 indicates the results for the Chi-square test. Policy beliefs Federalism Most of the national articles (n = 73; 99%) and local articles (n = 100; 99%) used nationally-based or locally-based source cues in their respective accounts of the GYA policy issue and were thus coded for federalism. Overall, the national articles use national source cues (+.3999) and local articles tend to use local source cues (-.1000). There is a statistically significant difference in use of national and local source cues between the national media outlets and the local media outlets (t = 5.004, df = 170, p \.001, onetailed); we reject H 01. Thus, there are differences in the policy belief of federalism in these media accounts, with the national articles aligning with the New West Advocacy Coalition s nationalism theory of federalism and the local articles articulating more of the Old West Advocacy Coalition s compact theory of federalism. Human and nature relationship Who or what the article identifies as a victim sheds light on where the media narrative falls in terms of the relationship between humans and nature. A human-centered victim is more aligned with the Old West Advocacy Coalition, while environmental-centered victims are thought to be more allied with the New West Advocacy Coalition. Both the national and local articles tended to portray the victims of the policy issue as anthropocentric ( and , respectively). There is no statistical difference between national and local framing of the victims in these two policy issues. We fail to reject H 02.

14 Table 3 Policy beliefs, framing, and problem definition by newspaper type n Mean SD t-statistic df p Policy beliefs Federalism National 73 (99%) Local 100 (99%) (one-tailed) Human/nature relationship National 65 (88%) Local 47 (47%) ns Type of science National 17 (23%) Local 5 (5%) ns Framing New West/Old West source cues National 73 (99%) Local 100 (99%) (one-tailed) New West/Old West descriptors National 69 (93%) Local 75 (74%) (one-tailed) Framing format National 65 (88%) Local 47 (47%) (two-tailed) Source: New York Times, USA Today, West Yellowstone News, and Cody Enterprise Notes: Percentage in parentheses is the percent within national or local articles; ns = non-significant Table 4 Chi-square results for media problem definition by newspaper type Source: New York Times, USA Today, West Yellowstone News, and Cody Enterprise Numbers in parentheses are the number of articles National Local Total Economic 9% (7) 14% (14) 21 Environmental 14% (10) 13% (13) 23 Political 59% (44) 19% (19) 63 Combination 18% (13) 54% (55) 69 v 2 (df = 3) = , p \.001; Cramér s V =.449, p \.001 Type of science Whether an article highlights science that offers a conservation-related solution or a technological fix sheds light onto another aspect of policy beliefs in the media. The total number of references to science in the 175 documents was a mere 22 citations (12%). Of the science cited, both national and local articles tended to use conservation science

15 ( and , respectively). There is no statistical difference between types of science cited in these media accounts. We fail to reject H 03. Narrative framing strategies New West/Old West source cues The nationally and locally based source cues were also coded for New West or Old West policy orientation. In other words, source cues tended to espouse a policy leaning, pro- or anti-snowmobile access and pro- or anti-wolf reintroduction. In turn, these policy leanings were grouped into two categories: Old West (pro-snowmobiling and anti-wolf) and New West (anti-snowmobiling and pro-wolf). Overall, the national articles cite New West leaning source cues (+.1273), while the local articles tend to use slightly more Old West leaning source cues (-.0657). There is a statistically significant difference in use of New West and Old West source cues between the national media outlets and the local media outlets (t = 2.021, df = 170, p \.05, one-tailed); we reject H 04. The narrative strategy of framing issues through consistent use of source cues with a particular policy leaning does exist in these media accounts. The national articles use source cues affiliated with New West coalition values, while the local articles use slightly more source cues that represent Old West coalition values. New West/Old West descriptors A journalists use of descriptors in portraying the policy issue is another way to uncover narrative strategies of framing. Overall, the national articles tend to employ more New West descriptors (+.0303), while the local articles use more Old West descriptors (-.1995). There is a statistically significant difference in the use of New West and Old West descriptors of these GYA issues between the national media outlets and the local media outlets (t = 1.925, df = 131, p \.05, one-tailed); we reject H 05. Consistent with citation of New West and Old West source cues as a narrative strategy, national articles show a more New West policy orientation through its descriptors and local articles show a more Old West policy orientation through its descriptors. Framing format Whether an article uses an episodic or thematic framing format is reflective of a narrative strategy. The national articles tended to use thematic or general frames more than episodic (+.2779), but local articles robustly employed this thematic format (+.8241). While the national and local articles did not straddle the tipping point of zero, there is, nonetheless, a statistically significant difference between the two groups in terms of use of format in writing about the GYA policy issue (t =-4.519, df = 108, p \.001, two-tailed); we reject H 06. Contextual problem definition The coders were asked to interpret how the author generally defined the problem. This open-coded, subjective evaluation resulted in a low exact match between the coders (53%);

16 however, 97% of the time, the coders agreed or missed by one. The labels that the coders open coded were categorized into four broad categories of problem definition: economic, environmental, political, and some combination of these three. Table 4 indicates the results of a Chi-square test, testing if there is an association between problem definition and type of newspaper, national or local. There is a statistically significant association between problem definition in national and local news articles (v 2 (df = 3) = , p \.001), with an indication of a strong measure of association (Cramér s V =.449, p \.001). An economic problem definition is used by the national media 9% of the time (f o = 7; f e = 8.9), whereas this same definition is used in local accounts 14% of the time (f o = 14; f e = 12.1). An environmental problem definition is used by the national media 14% (f o = 10; f e = 9.7) of the time, whereas the local accounts use it at a rate of 13% (f o = 13; f e = 13.3). A political problem definition is used by the national media 59% of the time (f o = 44; f e = 26.6), whereas this same definition is used in local accounts 19% of the time (f o = 19; f e = 36.4). Finally, a combination of these problem definition is used by the national media 18% of the time (f o = 13; f e = 28.8), whereas a combination is used in local accounts 54% of the time (f o = 55; f e = 39.2). We reject H 07. Discussion Members of an advocacy coalition coalesce around a core set of policy beliefs (Sabatier and Jenkins-Smith 1993) that can be systematically documented through narrative analyses (McBeth et al. 2005). Contending coalitions in contentious policy battles also structure their narratives employing political strategies that are believed to affect their desired policy outcome policy change or status quo (McBeth et al. 2007). The purpose of this study is to empirically test whether media accounts are, really, policy stories with embedded policy beliefs and congruent narrative strategies to support their policy beliefs. In other words, is the media a conduit for policy actors or a contributor in the policy debate itself? The results indicate a more complex policy landscape, perhaps necessitating a move away from a dichotomous conceptualization about advocacy coalition membership as in (contributor) or out (conduit) to matters of degree (contributor in particular instances and conduit in others). To the extent that the media contribute policy beliefs in their own right, the media aligned with the GYA s New West and Old West Advocacy Coalitions on only one out of the three policy beliefs tested. Like their highly influential New West and Old West interest group counterparts (see McBeth et al. 2005), the national and local media construct policy stories through the policy belief of federalism. National papers predominantly used national source cues (+.3999) and local papers predominantly used local source cues (-.1000); this result is consistent with what we predicted. The use of national source cues signifies that the national papers view the problem as a national issue (nationalism theory of federalism), while the use of local source cues indicates that the local papers consider the issue as a local one (compact theory of federalism). Also concurrent with the results of the interest group study, federalism is the dominant policy belief exhibited in the newspaper articles in that it is used more than the remaining two policy beliefs. This convergence on the policy belief of federalism for interest groups and the media in the GYA empirically demonstrates (1) how both contribute to advocacy coalitions in the GYA