Issue Emergence,Evolution of Controversy,and Implications for Competitive Framing: The Case of the HPV Vaccine

Abstract

Although scholarship on competitive framing acknowledges that framing is a dynamic process in which the early stages may matter most, very little research has focused on the dynamics of issue emergence. In this article, we draw on several literatures to develop theories for how controversy related to new issues will emerge and expand in news coverage. Through a comprehensive content analysis of 101 local newspapers across the fifty U.S. states, we explore the dynamic and evolving process wherein a new issue—the HPV vaccine—emerged into public discourse and a legislative debate over school requirements for vaccination began. We find that coverage of controversy is a function of proximity, driven primarily by events within a state, although external events also influence local coverage. We also find that the legislative discussion in the media did not necessarily start out as controversial, but as the issue evolved, we observe a large increase in the proliferation of both actors taking positions and the types of arguments made to influence debate. The findings yield important insight into issue emergence with implications for how future research might test competing frames to better understand how the presentation of controversy in the mass media affects public opinion.

Keywords

news production news reporting political conflict media framing health

Introduction

Scholars have long acknowledged the central role mass media play in “framing” issues by emphasizing certain features of a debate (Entman 1993). In recent years, because of the recognition of the importance of competition in most public policy debates, research incorporating different issue frames—alternative interpretive packages—within the same experimental treatment has proliferated (see especially Chong and Druckman 2007a; Sniderman and Theriault 2004). Rather than demonstrating the influence of single, or one-sided, messages on public opinion, scholars examine whether and how multiple messages in competition can shape public attitudes and beliefs. Although most scholarship on framing in general and competitive framing in particular agree that “framing is best conceptualized as a process that evolves over time” (Chong and Druckman 2007b: 108) and that the early stages of a framing debate are likely to matter most because of the malleability of opinion (Chong and Druckman 2007b: 188), we still have a very limited understanding of how, when, and from where competing messages emerge.

When new issues emerge on the public policy agenda, opposing groups of practitioners, professional societies, industry representatives, advocates, and consumers often compete to define the problem and influence the ensuing debate, through lobbying, advertising, and media attention. This process may have a particularly pronounced influence on the public for the emergence of new medical technologies, which can sometimes find their way into practice before there is general elite agreement or professional guidelines regarding their scientific, medical, or political appropriateness (see Rettig et al. 2007 for an illustration of how this happened in breast cancer treatment). Media coverage of the groups competing for attention thus serves as one important setting for public deliberation, and coverage early on in the process of issue construction has the potential to shape elite and public opinion, utilization of a new technology, and/or even eventually the course of public policy actions (Nisbet et al. 2003).

The emergence and evolution of competing actors and arguments in these instances of new issues deserves greater attention by media and political communication scholars. A key question that has been rarely addressed in prior literature is germane to many, if not most, issues: In the early stages of discussion over complex, multifaceted issues, how do media present the emerging messages, interests, and stakeholders prior to the crystallization of professional opinion, and how does coverage of developing controversy evolve and change over time?

In this article, using a large-scale content analysis of local media, we assess how media coverage of the human papillomavirus (HPV) vaccine and HPV vaccine school requirements emerges and develops over time. The recent introduction of and public policy controversy over the vaccine provides us with a useful case with which to examine the evolution of controversy, as it represents a genuinely new, multifaceted issue on which experts’ and politicians’ arguments and positions were in flux. The resulting analysis makes several important theoretical and empirical contributions to the existing literature across multiple streams of scholarship. In particular, contributing to the framing literature, we argue and provide evidence demonstrating that competing frames are indeed dynamic and should be conceptualized as such, but they also may follow a predictable pattern as controversy develops from repackaged arguments early in a debate to more nuanced or complex argumentation later. Second, we contribute to the literature on news production, showing that predictable factors such as controversy and proximity of controversy to a target audience drive the extent to which an issue will maintain media attention. Finally, we contribute to theories of policy making and elite and public opinion in at least two ways. We document the vagaries of expert opinion, shifting, dramatically at times, in response to political events in new issue domains. And, while the study was not designed to examine the direct influence of media on policy outcomes, our findings suggest that as media controversy increases over the issue-emergence process, the policy options available to actors on both sides of a given conflict may narrow.

Issue Emergence, Problem Definition, and Integrating Disparate Literatures

It is widely acknowledged that elite communication, media, and public opinion play a central role in the policy-making process. In particular, scholars argue that the transmission of elite and expert opinion through the media along with the presentation of particular interpretations of policy problems (i.e., frames) can shape both public opinion and outcomes (Brody and Shapiro 1989; Gamson and Modigliani 1989; Iyengar 1991; Joslyn and Haider-Markel 2006; Zaller 1992). A brief review of three related but somewhat disparate literatures is important for informing our expectations of how controversy over new issues will appear in news: the literature on media framing, the literature on agendas and issue construction, and the literature on news routines and production.

Competitive Media Framing

There are two types of relevant framing studies: one examines the development of frames through content analysis and the other (and perhaps more prominent one in political science and communication, see Chong and Druckman 2007b: 109) examines the effect of frames on citizen attitudes. Both types of research sometimes acknowledge that framing is a dynamic process, in which definitions of problems change over time and even “traditional” issues can be reframed and/or transformed into “new” issues (Chong 2006). Notably, using content analysis, Miller and Riechert (2001) have sought to identify the general phases of framing debates. They describe four different periods of stakeholder competition (what they call a “framing cycle”): emergence, then conflict (over definition of the problem), then resonance (where one side becomes ascendant and the other adjusts its messaging), and finally, equilibrium/resolution (where one side dominates, followed by enactment of public policy). Meanwhile, Chong and Druckman have evaluated the dynamics of competitive framing over multiple time points (Chong and Druckman 2007b, 2007c).

Yet research focusing on dynamic processes represents the exception, rather than the rule. Even if scholars acknowledge longitudinal processes, most rarely explicitly address this dynamism by exploring the effects of frames over time or the evolution of issue frames in discourse (Kinder 2007). The lack of research in this area can be explained by certain methodological conventions. First, researchers who conduct laboratory-based experiments (the majority of research on framing) tend to craft the messages they assume to be potentially influential, rather than to conduct observational research on the origin and evolution of competition in the real world (Kinder 2007). Second, studies on framing (both observational and experimental) tend to identify frames inductively, by culling preexisting academic, popular, legal, or stakeholder communications (Arnold et al. 1998; Brewer 2001; Chong and Druckman 2007c; Terkildsen et al. 1998). In other words, despite the recent, more explicit focus on dynamics in framing, existing research still tends to start with the assumption that there are clearly defined, static, competing sides/alternatives, rather than explore how competing or alternative “interpretive packages” gain attention or are developed in response to media attention. While these analytic techniques are useful in identifying and tracing issue packages post hoc or testing the effect of prominent existing frames, such methods ignore frame emergence and frame building, which can be a much messier and more chaotic process. This process warrants attention in its own right, particularly for new issues wherein stakeholders are just beginning to develop strategies to achieve dominance over the issue definition.

Thus, we examine the initial phases of a framing cycle—from issue emergence to conflict over problem definition. We contend that many issues do not necessarily start with obvious competing stakeholders or clearly defined alternatives; new issues are not inherently controversial, but become so through a social and political process. Furthermore, the beginning phases of a framing cycle may be even more important than later phases, because policy makers or other interest groups find it difficult to change perspectives once an issue has been characterized early in a media debate (Linsky 1986; Schön and Rein 1994), and initial frames may be particularly influential on malleable opinions (Chong and Druckman 2007b). Thus, the beginning stages of a framing cycle can shape both the range of acceptable options pursued by policy makers as well as public support for government action. The lack of a theoretical understanding of how competing frames emerge and develop over the course of public debate is a serious hindrance to our understanding of how the media contribute to the process through which some policy options become viable while others lose out. To gain this theoretical insight, we move beyond the traditional framing literature and incorporate the intersection of the public agendas and news production literatures.

Public Agendas

Research on public agendas identifies the important role of policy entrepreneurs in defining public issues. Issues in society only come to be viewed as problems for government to address through a strategic contest wherein actors compete to define the causes and solutions of a problem as well as which groups and stakeholders are relevant (Baumgartner and Jones 1993; Stone 2002). In the early stages of problem definition, entrepreneurs (from government, interest groups, or other expert organizations) invest time and resources to create and then refine their arguments regarding the policy issue at stake, combining and applying ideas from other issue domains to the subject at hand in their search for messages that will resonate with the values and beliefs of their constituencies (Kingdon 1997). Thus, even for an ostensibly “new” issue (like the HPV vaccine), entrepreneurs will choose to define their positions using other, well-established arguments that are familiar to the public.

The literature on social movements similarly envisions an active competitive process of framing, where advocates construct problems by repackaging previously available discourse in ways that will mobilize constituents (Benford and Snow 2000; Brown and Zavestoski 2004). Groups aiming to mobilize public support for a new issue will be more effective if they draw from preexisting cultural symbols and strong, established, networks of participants (Jenkins 1983). As an issue evolves through this strategic process from a neutral condition to a problem for government to address, more entrepreneurs will seek to become involved and compete to have their message heard. In response, these entrepreneurs will develop “counterframes” (Benford and Snow 2000). Thus, over time, we would expect to observe learning and evolution across stakeholders as groups expand the available repertoire of arguments and respond to countervailing forces. The types of stakeholders involved vary by issue domain, with health-related social movements tending to rely on medical actors for their expertise and credibility even while activist organizations seek to challenge their authority and alliances with existing state and industry power structures (Brown and Zavestoski 2004).

News Production and Media as an Interdependent Actor and Formal Institution

The literature on news routines and news production remind us that the news is first and foremost a business seeking to attract consumers (Hamilton 2004; McManus 1995) but also an independent actor exerting its own influence on the shape of public discourse (Terkildsen et al. 1998). To do so, the media use a range of tactics, one of the most prominent being the discussion of public events through a lens of conflict and controversy (Graber 2008; Patterson 1994). In addition, the news media are a political institution with well-established norms and routines that lead to some predictability in terms of which events and sources are likely to be covered, as we discuss below (Cook 1998; Graber 2008).

All issues are not equal in terms of their ability to attract media attention, and newsrooms have discretion over what they select. Graber (2008) argues that there are five salient criteria journalists consider in selecting stories: (1) events should have a strong impact on the audience; (2) violence, conflict, disaster, or scandal sells; (3) familiar storylines are better; (4) proximity matters (i.e., stories that are “close to home” draw more interest); and (5) stories should be timely and novel. Of these criteria, Graber finds that conflict, proximity, and timeliness are most important in determining what gets covered rather than left on the cutting room floor. All else equal, the extent that an issue is controversial (or can be covered as such), close to home, and happening now, we should expect to find more coverage in comparison to issues that are uncontested, distant, or old news. Given the news media’s focus on medical breakthroughs and new technologies (in contrast to, for instance, the social structural factors associated with population health), we expect that attention to a new vaccine will be high (Gasher et al. 2007; Nelkin 1987; Seale 2003).

Journalists and newsrooms clearly make choices about not only what to cover but how to cover it. Most work documents media dependence on elite sources, usually government actors and other partisan officials (Bennett 1990), but other issue-related relevant experts, such as scientists or the medical community, are also important (Zaller 1992). Given a range of possible perspectives from sources, when differences among elites or experts exist or are expected to emerge, such differences will draw journalistic attention (Terkildsen et al. 1998). Because of the journalistic norms of “objectivity” and balance, resulting coverage is likely to produce magnification of any given disagreement (Bennett 1990), with journalists trying to balance competing perspectives (even in the face of more imbalance among sources in reality) while also selecting credible (e.g., expert) sources on the subject at hand. Source selection by journalists likely varies with the stages of a framing cycle. In the early stages of issue emergence, journalists are less likely to have a choice of sources, and therefore may have to rely more on those who are available, whereas in the latter stages when more actors have entered the fray, journalists may exercise more discretion over whom to select.

Generally, then, our expectations for the emergence of competition from the production literature are as follows. News is fundamentally event driven, and although events other than public discussions or entrepreneurial action can propel an issue to the forefront of media attention, ultimately news coverage of a potential issue will only be sustained when followed by policy activity. New issues will not necessarily start out as controversial, but because of media norms surrounding controversy, we expect conflict to be a prominent feature of coverage. Controversial coverage should increase over the course of the initial phases of debate, as both journalists and stakeholders learn about the multifaceted nature in which an issue could be discussed. Coverage of controversy should also be related to the phase of the debate in a given environment (e.g., controversial coverage should increase when policies are under active consideration compared to when they are not) and to the proximity of the debate for a given news audience (e.g., local policies will garner more coverage than policies under consideration in other states).

Justification for the HPV Vaccine as a Case

The HPV vaccine, wildly heralded as a medical breakthrough, came under scrutiny just three short months after FDA approval when lawmakers from Michigan introduced legislation proposing compulsory vaccination for all girls entering sixth grade. Many other states followed suit, and to date, twenty-four states and the District of Columbia have introduced proposals to require HPV vaccination for school entry, with two passing (Virginia and DC). The policy issue became particularly contentious—and remains so even in 2011—when Texas Governor Rick Perry issued an executive order in February 2007 to require the vaccine, an action that was later overturned by the legislature. In addition, a total of forty-one states have considered legislation aimed at funding or providing education about the vaccine, nineteen of which were successful.¹ Despite the fact that legislative proposals requiring vaccination for school entry were labeled “mandates,” most included generous opt-out clauses, which would have allowed parents to obtain exemptions even without a medical, philosophical, or religious justification.

Although the ensuing, high-profile debate was largely characterized as a standoff between vaccine and women’s health advocates, arguing for prevention of cervical cancer, and social conservatives, arguing that HPV vaccination may promote promiscuity (Roll 2007), media attention in each state suggests this may have been simply the initial hook for journalists prior to the entry of other actors. In fact, the condensed timeline between FDA approval and legislative discussion made for an especially volatile, highly visible debate in which even public health and medical experts were not always in agreement over the best course of action, and some sought to change their positions midstream (Haas et al. 2009).

In sum, several features of the HPV mandate controversy make it particularly well suited as a case to contribute to our understanding of competition in issue emergence under public scrutiny. Whereas most issues (e.g., abortion) have been around for a sufficiently long time that experts and the public alike are aware of the standard arguments used on either side of the debate and opinion and arguments at both levels tend to be relatively stable, the case of HPV vaccine mandates represents a genuinely “new” issue on which even political elite and medical experts’ opinions were dynamic. Unlike other newly emergent issues on the public agenda such as obesity (Oliver and Lee 2005), however, the discourse surrounding HPV vaccine mandates was inherently politicized, making the presence of political actors a salient feature of the public debate.

Research Aims and Expectations

Broadly speaking, we seek to understand the emergence of controversy in mass media. When and how does a new issue emerge, and are there predictable patterns to the way in which it will develop in the media? How will coverage vary across different issue environments? Our theoretical contribution to the competitive framing literature lies in the argument that the issue emergence and argument-building stage of a framing cycle is dynamic but systematically explainable by factors well known in the policy agenda and news production literatures, as described below.

First, knowing that controversy is not predetermined and news is fundamentally event driven, we expect to find more coverage and more coverage of controversy when states are actively considering HPV vaccine mandates for school entry because of the news production norms of proximity along with journalist proximity to the scene. More specifically, we expect more coverage of controversy in states that have initiated legislative discussions compared to those who have not, and we also expect coverage of controversy to be more prevalent within any given state after the state has formally introduced a bill requiring the vaccine than prior to doing so. However, to the extent that an event is sentinel (like the heated conflict in Texas), we should also expect to find mentions of said event in other states.

We expect that media discourse will start out citing only a few actors or arguments and rely on arguments previously resonant in other domains; coverage may even be one sided. But because legislative discussion will spur activity of policy entrepreneurs seeking to shape the debate, leading to more arguments and more sources for media to draw on, we should expect to see a growth in actors referenced and arguments used on both sides of an issue as the framing cycle moves from issue emergence to problem definition. The news production literature leads us to the same expectation of proliferation of controversy over time, given norms of balance and drama. In addition, at the time of issue emergence, we expect that journalists will rely on the major elites relevant to the debate, the medical experts and partisan elites considering vaccine legislation. However, as journalists and other stakeholders start to think of a new issue as controversial, more actors from other expert domains may enter the fray to influence the direction of the debate.

Data and Methods

We examined media coverage regarding the HPV vaccine in all fifty states from January 1, 2006, through December 31, 2007. This time frame was chosen so as to include six months prior to licensure of the vaccine by the FDA to eighteen months after vaccine approval. A total of 101 newspapers were sampled, including two local newspapers per state when available, and two national papers (USA Today and The New York Times).² Newspapers were sampled using the following selection criteria. We required newspapers to be available through either the LexisNexis or NewsBank online databases for the full time period of the study. The first paper selected was the largest circulating paper from the capital city of that state available through either database. The second paper was either the highest circulating paper available in the state or the second-highest circulating paper if the capital city paper was already the highest circulating paper in the state.³

Newspaper articles within the 101 newspapers were identified using a validated key word search (quotation marks included): HPV or “HPV vaccine” or (Merck and Gardasil) or (Merck and HPV) or (Merck and “cervical cancer”) or “vaccine mandate.” Preliminary content analysis aided the development of a coding instrument inclusive of actors and arguments that frequently appeared in newspaper coverage. Through a series of iterative adjustments and decision rule modifications, we arrived at a valid codebook, and each coder analyzed a subsample of articles for intercoder reliability (ICR) before beginning initial coding, and a subsample of 183 articles were double coded as a posttraining ICR test. We explicitly chose to code messages at the argument-level, a level more discrete than frames because, in the early stages, messages may not yet be packaged together in a systematic way, since communication strategies among elites are likely in flux. Our coding instrument is consistent with other work that seeks to detail the myriad of arguments used in the HPV vaccine debate (Vamos et al. 2008).

We coded articles for three general types of information that would convey controversy to the public, one explicit and two implicit: (1) explicit mentions of controversial state actions, (2) the distribution of arguments for or against government mandates, and (3) the positions actors took regarding the potential action. First, we captured mentions of state action with respect to mandating the HPV vaccine and explicit references to controversy over such action within and outside of the state that published the paper. We then identified arguments used for and against government mandates requiring HPV vaccination for school entry. Unless explicitly stated otherwise, throughout this article, when we refer to “mandate,” “school mandate,” or “requirements for school entry,” we mean governmental (legislative or executive) action that requires school-aged girls to receive the HPV vaccine. Arguments for and against mandates fell into one of four general categories: (1) political (e.g., government, science, or parents should have the authority to guide vaccination policy), (2) practical (e.g., vaccine cost/cost-effectiveness, access to vaccination, and educational benefits), (3) medical (e.g., cervical cancer or HPV prevalence, vaccine safety, side effects), and (4) moral (e.g., effect on promiscuity, abstinence as an alternative, moral value of saving lives). Finally, we coded for the presence of key stakeholders (politicians, advocates, and healthcare actors) along with each actor’s stated position with respect to government requirements for school entry (e.g., in favor, against, mixed). All variables except actor position were coded dichotomously, one for feature present and zero for absent. Intercoder agreement on all variables was consistently higher than 85 percent, and Kappa scores exceeded 0.70 except in a few cases (as noted below) driven primarily by the low incidence of the variable in the sample of double-coded articles: mentions of government action (agreement range = 90.8, 98.3, Kappa range = 0.8, 0.9), mentions of controversy (agreement range = 86.3, 91.3, Kappa range = 0.7, 0.7), and arguments (agreement range = 85, 90, Kappa range = 0.6, 0.7).⁴

The validated key word search identified a total of 2,181 articles for content coding once duplicates were excluded. Coders excluded 522 articles by hand after determining that the article did not mention the HPV vaccine (even if it did mention HPV or cervical cancer), was a preview to an upcoming article (e.g., “HPV vaccine discussion heats up, see page A15”), or was a calendar event (e.g., “Gardasil town hall discussion, Tuesday, 7pm”). Therefore, a total of 1,659 articles were included in this analysis. The data analysis consisted of calculating frequencies of the key content variables, assessing differences in messages conveyed and actors cited across time (defined in terms of sentinel issue events) and across state environmental contexts (defined as a simple measure of whether a state introduced a legislative mandate or not). Although we have a census of observations for our sample of newspapers, we treat each story as a random draw of possible ways in which a story could have been covered and therefore use statistical inference to assess differences over time and contexts.

Results

Figure 1 displays the volume of articles appearing in the 101 papers across the full two-year time period, with vertical lines indicating different phases of the legislative debate over school requirements. The prelegislative period, prior to the introduction of the first bill to require the vaccine for school entry in Michigan includes the spike in coverage during June 2006, when Gardasil® was approved by the FDA. The second big spike in coverage occurs in response to the heated debate after Texas became the first state to mandate the vaccine through Perry’s executive order that was eventually overturned by the legislature following the firestorm of public disapproval. There were a total of 368 articles that appeared in the prelegislative phase (January 1, 2006, to September 10, 2006—an average of 44 stories per month), 244 that appeared in the pre-Texas phase (September 11, 2006, to February 1, 2007—an average of 51 stories per month), and 1,047 articles that appeared in the post-Texas phase (February 2, 2007, to December 31, 2007—an average of 95 stories per month).

Figure 1.

Volume of coverage over time

We test the growth of controversy in two ways. First we examine the number of actors expressing explicit pro or con positions regarding state requirements related to the HPV vaccine. We find that the number of actor position statements referenced grew from 115 in the prelegislative policy period (an average of 0.3 actor statements per article) to 235 in the pre-Texas period to 1,000 in the post-Texas phase (both with an average of 1 actor statement per article). Aggregating to the level of the day, rather than the article level, we see a similar growth of actor statements: 0.5 actor statements each day in the prelegislative phase, to 1.6 per day in the pre-Texas phase, and 3.0 actor position statements per day of coverage in the post-Texas phase.

Second, we investigate the percentage of articles mentioning controversy over state action (including mentions of controversy in states other than the one where the newspaper comes from) in the early legislative phase (pre-Texas) compared to the post-Texas phase. We find that the presentation of the issue as controversial increases from 20 percent of all articles to 38 percent of all articles, a statistically significant increase.

Next we examine the influence of proximity, namely to what degree does media emergence of controversy depend on the state-specific policy environment? Table 1 details two comparisons: (1) between the twenty-four states that introduced school requirements for HPV vaccination compared to those that did not and (2) whether legislative introduction was associated with more controversial coverage within the state of introduction. The former question asks whether states that introduced school requirements for HPV vaccination had distinctive media environments (as measured through mentions of controversy), while the latter asks whether the specific state environment is associated with the nature of media coverage. For each comparison, we examine the results overall first and then exclude the newspapers from Texas as a robustness check, given that the Texas case is a clear outlier in volume and intensity of political controversy. As Table 1 shows, we find strong evidence of the influence of proximity on the emergence of media controversy even after excluding Texas. Readers of newspapers from states that introduced HPV vaccine school requirement legislation were significantly more likely to encounter coverage of controversy surrounding the legislation in their newspapers than those who did not. Moreover, within a given state, the introduction of mandate legislation is associated with a substantial increase in the emergence of media presentation of controversy within that state, lending support to the expectation that proximity matters for the way issues emerge in media coverage and that policy action will help sustain media attention because of event-focused reporting.

Table 1.

Mentions of Controversy Comparisons

Comparing across States	No Mandate Introduced	Mandate Introduced
All HPV-related articles	646	853
% mentioning controversy	18.6	34.1***
Excluding TX newspapers	646	713
%	18.6	26.8***
Comparing within States That Introduced Mandates	Before Introduction	After Introduction
All HPV-related articles	292	561
% mentioning controversy	11.6	45.8***
Excluding TX newspapers	282	431
%	11.0	37.1***

Note: National newspaper articles excluded (because of lack of state attribution).

***

p < 0.001 (one-directional test for increase; difference statistically significant compared to prior)

As a final check on proximity, we examine coverage of internal versus external policy discussions. More than two thirds (69 percent) of controversy mentions were about the state in which the paper was published, but we also find evidence that seminal events matter, as 30 percent of all mentions of controversy in stories published after Perry’s executive order refer to events in Texas.

As an issue evolves, we expect to see growth in the number and type of actors weighing in on the debate along with an increase in opposing viewpoints. Figure 2 displays a count of the number and the range of actors presented in newspaper coverage as having explicitly pro or con positions regarding school mandates, ranging from political (state governors and legislators) and ideological actors (conservative and other advocacy groups) to medical actors (associations, individual practitioners and public health professionals), religious groups, and other nonelite actors (i.e., parents, parents’ groups, and editorial columns). Panel A displays the array and frequency of actor positions for the prelegislative phase of debate; Panel B encompasses the pre-Texas phase between Michigan’s introduction of legislation and the Texas executive order; and Panel C shows the period following the Texas order. As displayed by the figure, there was both a proliferation of actors referenced along with a growth in opposing viewpoints within and between types of actors. In particular, HPV vaccine legislative discussions went from primarily referencing speculation of conservative groups’ likely opposition to government requirements for school entry to a more inclusive discussion, featuring political actors (including many Republicans) and medical actors as being supportive of HPV school mandates in the middle phase of the debate. In contrast, in the post-Texas phase, there was a stark increase in both (1) volume of position statements and the number and type of actor and actor groups cited in coverage and (2) a sharp division among most actor groups with respect to their positions on whether or not the HPV vaccine should be mandated for school entry.

Figure 2.

Number of articles containing actor positions on vaccine mandates for school entry by (A) prelegislative discussion and (B) post-Michigan, pre-Texas legislative discussion, NVIC=National Vaccine Information Center

Indicative of issue emergence and evolution in the early stages, however, in addition to growth in number of actors, there is also evidence of shifts in opinion within actor types. Democratic legislators were largely in favor of school-entry requirements overall; however, their support waned over the course of the debate from 92 percent in favor in the pre-Texas period to 78 percent post-Texas. Although conservatives were widely portrayed as opposing school-entry requirements, Republican legislators were more divided, and their (collective) opinions were not static. Prior to the Texas executive order, 64 percent of Republicans featured were in favor of school requirements, but in the post-Texas phase, it was just the opposite, with 60 percent opposing the policy. This suggests learning based on the public backlash to Republican Governor Perry’s executive order to require the vaccine.

Finally, although one expects to observe division between partisans and ideologues, as Panel C especially indicates, even medical experts were starkly divided with respect to the vaccine mandates. Although medical associations featured in the news were slightly more likely to oppose school-entry requirements, both individual medical practitioners and academic/public health professionals were evenly divided. And similar to political actors, medical actor positions were dynamic throughout the course of the debate. Although medical actor positions were not as prominent in the pre-Texas period, those that did appear were largely in favor of school-entry requirements, while a much stronger majority opposed such requirements in the post-Texas period. For example, 82 percent of articles containing medical practitioners’ positions on school-entry requirements in the pre-Texas phase stated that practitioners were in favor of the requirements, compared to only 35 percent of articles in the post-Texas period. As with the pattern for Republican legislators, it seems that the complexity of the issue may have caused some medical actors to change their views on mandates and/or to speak up in opposition to political action, especially once major medical journal op-eds warned of the risks on public confidence of controversial and hasty vaccine mandates (Gostin and DeAngelis 2007).

The final way we characterize emergence of controversy is to examine the types of arguments used in favor and in opposition to school mandates in general and by each phase of the debate, portrayed in Table 2. Despite a fair amount of coverage of HPV vaccines in the prelegislative period, most of this coverage (not surprisingly) contained few arguments for or against school mandates. The few stories that mentioned arguments regarding vaccination requirements for school entry centered primarily on the likelihood of conservative or moral opposition surfacing, likely due in part to the familiarity of those messages. During the pre-Texas phase, when legislatures in several states began in earnest to discuss the possibility of requirements for school entry, we find evidence of an increasingly multifaceted debate featuring arguments from all four general categories both for and against the legislation. The most prominent messages in favor of legislation surrounded medical benefits of the HPV vaccine (e.g., reducing the incidence of cervical cancer and saving lives), practical implications (e.g., increasing access or availability of the vaccine), and political arguments (e.g., HPV vaccine mandates include generous opt-out clauses), with 12, 9, and 6 percent of articles mentioning each type, respectively. Messages against school-entry requirements tended to focus on moral (e.g., effects on promiscuity) and political (e.g., parents should have the decision-making authority) dimensions, with 19 and 20 percent, respectively. Although messages against school mandates tended to be slightly more prevalent in the pre-Texas phase, for the most part, both sides of the debate were well represented, with the practical and medical arguments tending to favor government action and the political and moral arguments more prominent for the opposition.

Table 2.

Percentage of Articles Mentioning Each Type of Argument, by Time Period and by Views on Mandates

		Time Period
Type of Argument	Views on Mandates	Prelegislation (Pre-Michigan)	Post-Michigan, Pre-Texas	Post-Texas
Politics	In favor	1.1	6.2**	5.3
	Opposed	9.0	19.7***	22.4
Practical	In favor	0.5	8.6***	5.1
	Opposed	0.8	3.3**	7.2***
Medical	In favor	1.4	11.9***	9.6
	Opposed	0.3	8.2***	15.5***
Moral	In favor	0.8	3.7**	2.4
	Opposed	8.2	19.3***	20.9
Total no. of articles		367	245	1047

***

p < 0.001, **p < 0.05, *p < 0.1 (one-directional test for increase; difference statistically significant compared to prior).

In contrast, the post-Texas period was characterized by prevailing opposition messages across all five types of arguments. Although both arguments for and against school-entry requirements increased following the heated Texas controversy, opposition messages overwhelmed messages favoring government mandates by a margin of 3 to 1. Arguments in favor of mandates tended to rely primarily on the medical benefits of the vaccine, with occasional references to political argument. Meanwhile, the opposition continued to invoke both political and moral arguments but also increasingly relied on medical arguments (such as concern over vaccine safety) in particular, though their appeals to practical concerns also continued to increase, and both the medical and practical argument increase in the post-Texas phase was statistically significant. Because of the great increase in the volume of articles between the pre- and post-Texas periods, it is also important to remember that not only did the overall percentage of arguments increase but readers of these local papers were being inundated with many more such arguments predominately on the side of the opposition.

Discussion

This large-scale news media content analysis provides important theoretical insights to the literatures on competitive framing, news production, and elite opinion during issue emergence, each of which will be discussed at length below. While others have addressed the scientific and medical implications of news media coverage of the HPV vaccine (Forster et al. 2010; Abdelmutti and Hoffman-Goetz 2009; Calloway et al. 2006; Habel et al. 2009) and the effects of framing medical attributes of the vaccine on the public (Kelly et al. 2009; Bigman et al. 2010)—no previous studies, to our knowledge, have focused on the vaccine’s political aspects. In addition, no previous work has specifically addressed the longitudinal and dynamic context—the emergence and evolution of news coverage—that is required to understand the dynamics of competitive framing (Kinder 2007).

Turning to issue emergence, we find, as our theories predict, that controversy is indeed a function of proximity, associated more closely with events within a state than outside it (in our case, active legislative consideration of government action), although seminal external events (i.e., Texas) can also correspond with local coverage patterns. As expected, we find that the legislative discussion in the media did not necessarily start out as controversial, although journalists first picked up on the controversial potential of the debate by focusing their initial attention on conservative commentators as sources rather than medical actors. As the HPV vaccine school mandate issue evolved and state legislative discussions began, we observe a large increase in the explicit mention of controversy as well as a proliferation of both actors and arguments. We also find an increase in the multidimensional nature of the debate, both in terms of actors and arguments involved as elites and journalists learn and exploit the different (e.g., political, practical, medical or moral) ways the issue could be discussed. In addition, we show that actor positions changed over time. Actors (such as Republican legislators) may have learned from the tenor of public opinion (particularly after the Texas policy debate), or alternatively, they may have interpreted increasingly controversial messages proliferated by other media as representative of public opinion and therefore adjusted their views accordingly (Baumgartner and Jones 1993).

Literature on policy agendas suggests that entrepreneurs in emergent issue domains would first rely on familiar arguments, those that resonate with key constituents (Kingdon 1997). Moreover, the literature on news routines predicts a popular focus in the media on issues that are familiar to most Americans (Graber 2008). Not surprisingly, then, we observe that the arguments that were most popular earlier in the cycle of issue emergence were those addressing moral arguments and political arguments, both of which might be common to previous political and/or medical issues. New biotechnologies are often framed in the media as having moral concerns, such as was the case with human embryonic stem cells or cloning (Nisbet et al. 2003). Thus, when the HPV vaccine emerged, journalists relied on a familiar expert source—conservatives—to provide the familiar moral counter-argument, even though the specific moral message (namely, the notion that the HPV vaccine would encourage young girls to become promiscuous) was new. Moreover, the political arguments that dominated the early phase of debate, especially questions of government authority over individual rights, are classic issues that have characterized American public health debates for decades (Gostin 2008). Later in the issue emergence cycle, however, argument types expanded to include other messages not obviously recycled or repackaged (Benford and Snow 2000) from previous issues. This is an important insight because it suggests that competitive frames will be different depending on the stage of the framing cycle.

While this analysis has focused on the earliest stages of issue emergence and frame development, it is important to acknowledge that opposing messages did not universally win out. Although hindsight tells us that legislative mandate introductions largely failed (with the exception of DC and Virginia),⁵ entrepreneurs were more successful in pushing alternative legislation to fund or educate the public about the vaccine. One potential implication of the findings is that as messages became more contested and controversial in tone, the policy options available to advocates or entrepreneurs became more limited (shifting from legislative mandates to, for instance, education). Therefore, the HPV vaccine case provides an illustrative example of how public issue emergence processes might limit potential policy options down the road. Future work will examine this hypothesis by assessing the association between media content and policy success and failures across states and over time.

These findings have important implications for practitioners and researchers. Although scholars of public opinion expect inconsistencies in mass opinion, they tend to characterize elite opinions as fairly stable and constrained (Jennings 1992). It is clear, however, that publicized messages from elites—especially those emanating from the medical community—regarding school-entry requirements for HPV vaccination were far from consistent throughout the debate. The effects of such instability are not fully known, but those who examine the effects of publicized controversy point to decreases in public support for institutions (Hibbing and Theiss-Morse 1995) and for changes to the status quo (Samuelson and Zeckhauser 1988) and increases in cynicism toward government (Cappella and Jamieson 1997). Accordingly, the case of HPV vaccine school mandates represents a cautionary tale about the consequences of rushing legislation on a new issue prior to crystallization of expert opinion (see especially the concerns in Gostin and DeAngelis 2007 and Schwartz et al. 2007 but also Gollust et al. 2010 for the first empirical test of these concerns). Further, the case is suggestive of the important role that framing may have had in shaping the public response to the HPV vaccine. The legislative proposals were all framed as mandates, in spite of their generous opt-outs, and may thus have spurred the political and moral arguments that emerged from the opposition first. While our data do not speak to this, alternative framing of the policy by vaccine advocates early in issue emergence (such as emphasizing the choice to opt-out inherent in the way the policies were set up) may have changed the course of the discourse.

Finally, our content analysis reveals two potential sources of controversy’s influence upon public opinion that future research ought to explore: the messenger and the message. Elites first wielded moral arguments opposing HPV vaccine school requirements and over time moved toward medical arguments for the opposition, such as that the vaccine is too new or may result in side effects. It is yet unknown whether these different types (i.e., moral versus medical) of arguments differ in their persuasive impact, or whether it is the messenger of the arguments (i.e., conservatives versus medical actors) that matters more. As the tide seemed to shift against the HPV vaccine requirements, was it because medical actors’ opposition was increasingly highlighted in the media, or because medical arguments became more persuasive than moral ones? Careful experimental designs that unpack the differential effects of source cues and message effects will be required for a fuller understanding of the effects of mass-mediated controversy on the public. Nevertheless, this study represents an important step forward in understanding the emergence of controversy and competitive framing.

Footnotes

Acknowledgements

We thank the Robert Wood Johnson Foundation Scholars in Health Policy Research Program and the Robert Wood Johnson Foundation Health and Society Scholars Program. The authors thank Alan Cohen, Tara Watson, Leslie Hinkson, Chris Adolph, Hans Noel, Patricia Strach and the participants of the RWJF Health Policy Aspen conference for helpful feedback. In addition, we thank Dr. Waisbord and the anonymous reviews for their suggestions.

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

The authors received financial support for this research from the Robert Wood Johnson Foundation Scholars in Health Policy Research Program.

Notes

References

Abdelmutti

Nazek

Hoffman-Goetz

Laurie

. 2009. “Risk Messages about HPV, Cervical Cancer, and the HPV Vaccine Gardasil: A Content Analysis of Canadian and U.S. National Newspaper Articles.” Women and Health 49(5):422-40.

Arnold

R. Douglas

Graetz

Michael J.

Munnell

Alicia Hadock

, eds. 1998. Framing the Social Security Debate. Washington, DC: Brookings Institute.

Baumgartner

Frank

Jones

Brian

. 1993. Agendas and Instability in American Politics. Chicago: University of Chicago Press.

Benford

Robert D.

Snow

David A.

2000. “Framing Processes and Social Movements: An Overview and Assessment.” Annual Review of Sociology 26:611-39.

Bennett

W. Lance

. 1990. “Toward a Theory of Press-State Relations in the United States.” Journal of Communication 40(Spring):103-25.

Bigman

C. A.

Cappella

J. N.

Hornik

R. C.

2010. “Effective or Ineffective: Attribute Framing and the Human Papillomavirus (HPV) Vaccine.” Patient Education and Counseling 81(Suppl):S70-76.

Brewer

Paul R.

2001. “Value Words and Lizard Brains: Do Citizens Deliberate about Appeals to Their Core Values?” Political Psychology 22:45-64.

Brody

Richard A.

Shapiro

Catherine R.

1989. “Policy Failure and Public Support: The Iran-Contra Affair and Public Assessment of President Reagan.” Political Behavior 11(4):353-69.

Brown

Phil

Zavestoski

Stephen

. 2004. “Social Movements in Health: An Introduction.” Sociology of Health and Illness 26(2):679-94.

10.

Calloway

Jorgensen

C. M.

Saraiya

Tsui

2006. “A Content Analysis of News Coverage of the HPV Vaccine by U.S. Newspapers, January 2002-June 2005.” Journal of Women’s Health 15:803-9.

11.

Cappella

Joseph N.

Jamieson

Kathleen Hall

. 1997. Spiral of Cynicism: The Press and the Public Good. Oxford: Oxford University Press.

12.

Chong

Dennis

. 2006. “Free Speech and Multiculturalism in and out of the Academy.” Political Psychology 27:29-54.

13.

Chong

Dennis

Druckman

James N.

2007a. “A Theory of Framing and Opinion Formation in Competitive Elite Environments.” Journal of Communication 57(1):99-118.

14.

Chong

Dennis

Druckman

James N.

2007b. “Framing Theory.” Annual Review of Political Science 10:103-26.

15.

Chong

Dennis

Druckman

James N.

2007c. “Framing Public Opinion in Competitive Democracies.” American Political Science Review 101(4):637-55.

16.

Cook

Thomas E.

1998. Governing with the News: The New Media as a Political Institution. Chicago: University of Chicago Press.

17.

Entman

Robert M.

1993. “Toward Clarification of a Fractured Paradigm.” Journal of Communication 43:41-53.

18.

Forster

Alice S.

Wardle

Jane

Stephenson

Judith

Waller

. 2010. “Passport to Promiscuity or Lifesaver: Press Coverage of HPV Vaccination and Risky Sexual Behavior.” Journal of Health Communication 25(3):205–17.

19.

Gamson

William A.

Andre

Modigliani

. 1989. “Media Discourse and Public Opinion on Nuclear Power: A Constructionist Approach.” American Journal of Sociology 95(1):1–37.

20.

Gasher

Mike

Hayes

Michael

Hackett

Robert

Gutstein

Donald

Ross

Ian

Dunn

James

. 2007. “Spreading the News: Social Determinants of Health Reportage in Canadian Daily Newspapers.” Canadian Journal of Communication 32:557–74.

21.

Graber

Doris A.

2008. Mass Media and American Politics. 8th Edition. Washington, DC: CQ Press.

22.

Gollust

Sarah E.

Dempsey

Amanda F.

Lantz

Paula M.

Ubel

Peter A.

Fowler

Erika Franklin

. 2010. “Controversy Undermines Support for State Mandates on the Human Papillomavirus Vaccine.” Health Affairs 29(11):2041-46.

23.

Gostin

Lawrence O.

2008. Public Health Law: Power, Duty, Restraint. Berkeley: University of California Press.

24.

Gostin

Lawrence O.

DeAngelis

Catherine D.

2007. “Mandatory HPV Vaccination: Public Health vs. Private Wealth.” Journal of the American Medical Association 297(17): 1921-23.

25.

Haas

Marion

Ashton

Toni

Blum

Kerstin

Christiansen

Terkel

Conis

Elena

Crivellil

Luca

Lim

Meng Kin

Lisac

Melanie

MacAdam

Margaret

Schlette

Sophia

. 2009. “Drugs, Sex, Money, and Power: An HPV Vaccine Case Study.” Health Policy 92(2-3):288-95.

26.

Habel

Mellissa A.

Liddon

Nicole

Stryker

Jo E.

2009. “The HPV Vaccine: A Content Analysis of Online News Stories.” Journal of Women’s Health 18:401-7.

27.

Hamilton

James

. 2004. All the News That’s Fit to Sell: How the Market Transforms Information into News. Princeton, NJ: Princeton University Press.

28.

Hibbing

John R.

Theiss-Morse

Elizabeth

. 1995. Congress as Public Enemy: Public Attitudes toward American Political Institutions. Cambridge: Cambridge University Press.

29.

Iyengar

Shanto

. 1991. Is Anyone Responsible. Chicago: University of Chicago Press.

30.

Jenkins

J. Craig

. 1983. “Resource Mobilization Theory and the Study of Social Movements.” Annual Review of Sociology 9:527-53.

31.

Jennings

M. Kent

. 1992. “Ideological Thinking among Mass Publics and Political Elites.” Public Opinion Quarterly 56:419-41.

32.

Joslyn

Mark R.

Haider-Markel

Donald P.

2006. “Should We Really ‘Kill’ the Messenger? Framing Physician-Assisted Suicide and the Role of Messengers.” Political Communication 23(1):85-103.

33.

Kelly

B. J.

Leader

A. E.

Mittermaier

D. J.

Hornik

R. C.

Cappella

J. N.

2009. “The HPV Vaccine and the Media: How Has the Topic Been Covered and What Are the Effects on Knowledge about the Virus and Cervical Cancer?” Patient Education and Counseling 77(2):308-13.

34.

Kinder

Donald R.

2007. “Curmudgeonly Advice.” Journal of Communication 57(1):155-62.

35.

Kingdon

John

. 1997. Agendas, Alternatives, and Public Policies. 2nd Edition. Pearson Education.

36.

Linsky

Martin

. 1986. Impact: How the Press Affects Federal Policymaking. New York: Norton.

37.

McManus

John

. 1995. “A Market-Based Model of News Production.” Communication Theory 5(4):301-38.

38.

Miller

M. Mark

Riechert

Bonnie P.

2001. “The Spiral of Opportunity and Frame Resonance: Mapping the Issue Cycle in News and Public Discourse.” In Framing Public life: Perspectives on Media and Our Understanding of the Social World, ed. Reese

Stephen D.

Gandy

Oscar H.

Grant

August

. Mahwah, NJ: Lawrence Erlbaum.

39.

Nelkin

Dorothy

. 1987. Selling Science: How the Press Covers Science and Technology. New York: Freeman.

40.

Nisbet

Matthew C.

Brossard

Dominique

Kroepsch

Adrianne

. 2003. “Framing Science: The Stem Cell Controversy in an Age of Press/Politics.” Harvard International Journal of Press/Politics 8(2):36-70.

41.

Oliver

J. Eric

Lee

Taeku

. 2005. “Public Opinion and the Politics of Obesity in America.” Journal of Health Politics, Policy and Law 30(5):923-54.

42.

Patterson

Thomas E.

1994. Out of Order. New York: Vintage Books.

43.

Rettig

Richard A.

Jacobson

Peter D.

Farquhar

Cynthia

Aubry

Wade M.

2007. False Hope: Bone Marrow Transplantation for Breast Cancer. Oxford: Oxford University Press.

44.

Roll

Carrie A.

2007. “The Human Papillomavirus Vaccine: Should It Be Mandatory or Voluntary?” Journal Health Care Law and Policy 10:421-30.

45.

Samuelson

William

Zeckhauser

Richard

. 1988. “Status Quo Bias in Decision Making.” Journal of Risk and Uncertainty 1:7-59.

46.

Schön

Donald A.

Rein

Martin

. 1994. Frame Reflection: Toward the Resolution of Intractable Policy Controversies. New York: Basic Books.

47.

Schwartz

J. L.

Caplan

A. L.

Sugarman

2007. “Lessons from the Failure of Human Papillomavirus Vaccine State Requirements.” Clinical Pharmacology and Therapeutics 82(6):760-63.

48.

Seale

Clive

. 2003. “Health and Media: An Overview.” Sociology of Health and Illness 25(6):513-31.

49.

Sniderman

Paul M.

Theriault

Sean M.

2004. “The Structure of Political Argument and the Logic of Issue Framing.” In Studies in Public Opinion, ed. Saris

W. E.

Sniderman

Paul M.

Princeton, NJ: Princeton University Press.

50.

Stone

Deborah A.

2002. Policy Paradox: The Art of Political Decision Making. New York: Norton.

51.

Terkildsen

Nadya

Schnell

Frauke I.

Ling

Cristina

. 1998. “Interest Groups, the Media and Policy Debate Formation: An Analysis of Message Structure, Rhetoric and Source Cues.” Political Communication 15(1):45-61.

52.

Vamos

Cheryl A.

McDermott

Robert J.

Daley

Ellen M.

2008. “The HPV Vaccine: Framing the Arguments FOR and AGAINST Mandatory Vaccination of All Middle School Girls.” Journal of School Health 78(6):302-9.

53.

Zaller

John R.

1992. The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press.