Journal of Communication ISSN0021-9916 O R I G I N A L A RT I C L E
A Theory of Framing and Opinion Formation in Competitive Elite Environments
Dennis Chong&James N.Druckman
Department of Political Science,Northwestern University,Evanston,IL60208
Public opinion often depends on how elites choo to frame issues.For example,citizens’opinions about a Ku Klux Klan rally may depend on whether elites frame the event as
新年快乐字体设计a free-speech issue or a public safety issue.Past rearch has focud largely on documenting
the size of framing effects in uncontested ttings.By contrast,there has been little rearch on framing in competitive environments in which individuals receive multiple frames rep-renting alternative positions on an issue.We take an initial step toward understanding how frames work in competitive environments by integrating rearch on attitude structure and persuasion.Our theory of framing identifies the key individual and contextual param-eters that determine which of many competing frames will have an effect on public opinion. doi:10.1111/j.1460-2466.2006.00331.x
In fall1999,the Brooklyn Muum of Art opened an exhibit called‘‘Sensation.’’The works by young British artists included a lf-replica bust constructed from nine pints of the artist’s own frozen blood,a sculpture that incorporated a dead animal in formaldehyde,and,most notably,a painting of a black Madonna festooned with elephant dung and pornographic pictures.Not everyone regarded the works to be‘‘art.’’New York Mayor Rudolph Giuliani found the exhibit‘‘disgusting’’and denounced the Madonna portrayal as being‘‘anti-Catholic’’(Herszenhorn,1999, p.A1).So angered was Giuliani that he took steps to withdraw public funding for the muum(which totaled over$7million a year)and evict the muum from the city-owned building it occupied.Giuliani declared that government was entitled to regulate distribution of taxpayer dollars for cultural events(particularly tho houd in publicly owned buildings);he oppod using public money to support an exhibit that was so offensive to general tastes.
As Giuliani publicly disparaged the muum while entertaining a possible run for the U.S.Senate,reprentatives of the muum and general art community remained surprisingly mute.New York Times reporter,Michael Kimmelman,noted the con-quence of this asymmetry:‘‘There is always a moment when the terms of public debate are framed.For nearly a week.the Mayor had been allowed to frame the current issue almost by himlf,an extraordinary gift t
o a politician’’(1999,p.B5). Corresponding author:James Druckman;e-mail:druckman@northwestern.edu
Journal of Communication57(2007)99–118ª2007International Communication Association99
Framing and Opinion Formation D.Chong&J.N.Druckman The muum finally joined the fray when it filed a lawsuit defending its First Amendment right to free expression.Two opposing frames now competed to define the controversy:the artists’free-speech rights versus the right of government to control its public finances and tax dollars.David Strauss,a University of Chicago law professor,described the dispute as‘‘a black hole of First Amendment law.No one really knows how to think about it,including the Supreme Court Justices’’(Glaberson, 1999,p.B12).Nearly6months later,the two sides ttled out of court,with the Muum ostensibly prevailing as it retained its budget and building.
As the controversy over the Sensation exhibit illustrates,virtually all public debates involve competition between contending parties to establish the meaning and interpretation of issues.When citizens engage an issue—be it social curity, foreign aid,a hate-group rally,affirmative action,or the u of public funds for art—they must grapple with opposing frames that are intended by opinion leaders to influence public preferences.
Surprisingly,social scientists have little to say about which of many competing ,free speech,allocation of public funds,the right to overe tenants in public buildings)will shape public opinion.Sniderman and Theriault(2004,pp. 141–142)explain that‘‘framing studies.have neglected the fact that frames are themlves contestable.They have instead restricted attention to situations in which citizens are artificially questered,restricted to hearing only one way of thinking about a political issue’’(also e Entman,1993;Riker,1995;Wittman,1995).
Our goal in this paper is to further understanding of how opinion formation works in competitive mass communication(framing)environments.After situating our inquiry in the literature on framing and public opinion,we develop a typology of competitive contexts and show that most studies of framing effects have restricted their examination to one-sided communications.We then propo a new standard bad on a control group for asssing the magnitude of framing effects when studying competing frames.Next,we prent a theory that identifies the psycho-logical process and contextual factors that determine which frames will have the greatest impact on public opinion.We conclude by deriving testable proposi-tions from our theory and discussing the relevance of our theory for other types of media effects.
Framing rearch
Communication scholars and political scientists generally u the term‘‘frame’’in two ways(Druckman,2001c;Scheufele,1999).First,a frame in communication or a media frame refers to the words,images,phras,and prentation styles that a ,a politician,a media outlet)us when relaying information about an issue or event to an audience(Gamson&Modigliani,1987,1989).The chon frame reveals what the speaker es as relevant to the topic at hand.For example,in the Brooklyn Muum ca,the city’s counl focud on the prerogative of govern-ment to withdraw funding on the basis of artistic content:‘‘The city’s view of art is 100Journal of Communication57(2007)99–118ª2007International Communication Association
D.Chong&J.N.Druckman Framing and Opinion Formation that it’s totally inappropriate.becau of the nature of the art itlf.the actual uptting,violent,disgusting view of some of the paintings,they shouldn’t be supported by taxpayer money’’(Koromvokis,1999).
Second,a frame in thought or an individual frame refers to an individual’s cog-nitive understanding of a given ,Goffman,1974).Unlike frames in communication,which reflect a speaker’s emphasis,frames in thought refer to what an audience member believes to be the most salient aspect of an issue.An individual who felt the Brooklyn Muum had a First Amendment right to choo
its exhibi-tions would be in a‘‘free-speech frame of mind.’’
Scheufele(1999)synthesizes a quarter-century of framing rearch by identifying four process:(a)‘‘frame building,’’which focus on the dynamics of how speak-ers,such as media outlets,choo specific frames in communication;(b)‘‘frame tting,’’which concerns the influence of frames in communication on frames in thought,and the preci psychological process at work;(c)‘‘individual-level effects of frames,’’which refers to the impact of frames in thought on subquent behaviors or attitudes;and(d)‘‘journalists as audiences,’’which looks at how citizens’actions affect the initial frame-building process(also e D’Angelo,2002;Scheufele,2004).
The role of multiple competing frames in each of the process has gone largely unexplored.For example,how do journalists decide which of their audience’s frames to incorporate in the frame-building process?What is the influence of competing journalists and politicians on the choices?Although the are important questions, our focus here is on how the prence of multiple competing frames in mass com-munication—a defining element of most political contexts—affect the audience’s frames and,in turn,their ,we focus on frame tting and individual-level effects;the two process are often called an emphasis framing effect;e Druckman,2001c).1
Our focus on competitive framing effects also means that we do not explicitly address priming(issue salience)and agenda-tting effects.That said,to the best of our knowledge,work on the related effects also have paid little or no attention to competitive situations.We thus believe our general framework ultimately could be extended to studies of priming and agenda tting.
What determines the public’s preference for one frame over another?Past work provides little insight into this question.The typical framing-effect study is an experiment that employs a one-sided design in which individuals are randomly assigned to receive one of two or more alternative reprentations of an issue.For example,in studies of people’s willingness to allow hate groups such as the Ku Klux Klan(KKK)to conduct a rally,the issue is construed either as a matter of free speech or as a threat to public safety,and the relevant comparison is the difference of opinion between individuals in the two conditions(Druckman,2001b;Nelson, Clawson,&Oxley,1997;Sniderman&Theriault,2004).If this approach were ud to study the main competing frames in the Sensation controversy,individuals would receive either the city’s argument for regulating cultural content or the Muum’s First Amendment defen of its exhibit before being asked to choo sides.
骆驼祥子介绍Journal of Communication57(2007)99–118ª2007International Communication Association101
Framing and Opinion Formation D.Chong&J.N.Druckman Most one-sided studies have found that contrasting frames have a statistically significant impact when compared to one another.For instance,individuals expod to the free-speech frame are significantly more willing to allow Klansmen to stage a rally than individuals who receive the public safety frame.This rearch therefore suggests that if one side can establish the relevant terms of debate over an issue,it can successfully persuade individuals to support its position.
收视率排行榜
But politics is typically competitive,fought between parties or ideological fac-tions,and issues that are debated are framed in opposing terms.Individuals receive multiple frames with varying frequencies.In one of the few experiments that has tested the effect of simultaneous exposure to opposing frames,Sniderman and Theriault(2004)found that individuals favored the frame that was consistent with their values.On the hate-group issue,when individuals were expod to both the free-speech and public safety frames,tho who placed a higher priority on freedom than law and order were inclined to be tolerant,but tho who subscribed more strongly to law and order tended to be intolerant.Sniderman and Theriault con-cluded that framing might be less influential in politics than experimental studies have suggested becau competing frames may cancel each other and fail to move public opinion(however,e Brewer&Gross,2005).
Although competition likely improves the odds that citizens will resist framing strategies—as Sniderman and Theriault suggest—it does not guarantee that the opposing sides will be equal combatants or that audiences will receive equal and simultaneous exposure to equally persuasive alternative frames.Pan and Kosicki (2001,p.45)aptly state,‘‘Resources are not distributed equally.Actors strategically cultivate their resources and translate them into framing power.’’Campaigns that have greater resources to conduct public opinion rearch may be better able to identify the frames that appeal most to the public.Unequal resources may also permit one side to adverti its themes more frequently(and to a wider audience) and to enlist reprentatives and endorrs that can more credibly deliver its mes-sages to the public(Chong&Wolinsky-Nahmias,2003).
Therefore,the effectiveness of any framing strategy will depend on its design and implementation within a particular competitive environment.To asss whether it is possible for one side to gain a framing advantage over the other requires that we study framing effects under various competitive conditions.
A typology of rearch designs
The design of experimental rearch on framing can be abstractly conceptualized along two dimensions.One dimension reprents the relative quantity of competing communications received by individuals in the experiment.Studies typically exam-ine only one frame per side,but we also include cas in which each side puts forth veral different frames promoting its position.Therefore,the relative quantity of frames equals the total number of exposures to every frame communicated by each side,respectively.If we assume two competing parties,we can reduce all possible combinations of relative frequencies into three discrete categories(e Table1):(a) 102Journal of Communication57(2007)99–118ª2007International Communication Association
asymmetric one-sided studies,in which individuals only receive one frame (one or more times);(b)dual (or symmetric)studies in which individuals receive opposing frames in equal quantity;2and (c)asymmetric two-sided studies in which individuals receive opposing frames in unequal quantities.Asymmetric one-sided studies are therefore ‘‘noncompetitive’’becau individuals are expod to only one side of a controversy,whereas dual and asymmetric two-sided designs model different ‘‘competitive’’environments.
Within any particular noncompetitive or competitive rearch design,the frames ud to reprent contending positions will vary on a cond dimension defined by their relative (perceived)strengths .
We will later elaborate on what we mean by strength,but for now we looly define a frame’s ‘‘strength’’as increasing with the persuasiveness of a given frame.Weak frames are typically en as unpersuasive,whereas strong frames are more compelling.For example,presumably most people would e ‘‘public safety’’as a strong or persuasive frame for why a hate rally should not be allowed,whereas ‘‘preventing litter on the streets’’would be a weaker frame.Table 1Prior Framing Effect Experimental Studies
Asymmetric One-中国最发达的城市
Sided (exposure to
just one frame)
Competitive Situations Dual (exposure to both frames in equal quantities)Asymmetric Two-Sided (exposure to both frames in unequal quantities)
Strong
frames Conventional framing-effect studies that show that individuals’opinions are significantly affected by exposure to a frame (e.g.,Iyengar,1991;Kinder &Sanders,1990;Nelson,Clawson,&Oxley,1
997;Price et al.,1997).Sniderman and Theriault (2004),Brewer and Gross (2005).None Weak
frames
Studies that explore moderators of framing effects,such as source credibility,political knowledge,and prior beliefs (e.g.,Brewer,2001;Druckman,2001b;Gross,2000;Nelson,Oxley,&Clawson,1997).None None Strong
打斧头and weak
frames Not applicable None None
D.Chong &J.N.Druckman Framing and Opinion Formation Journal of Communication 57(2007)99–118ª2007International Communication Association 103
Framing and Opinion Formation D.Chong&J.N.Druckman Although strength lies on a continuum,in Table1,we simply distinguish ‘‘strong’’from‘‘weak’’frames.This is typically assd empirically by asking pretest participants to rate the persuasiveness of a message or frame by characterizing it as either strong or weak(Eagly&Chaiken,1993;Petty&Wegener,1998).The rows in Table1identify three combinations of strong and weak frames that can corres-pond to any given competitive or uncompeti
红葡萄酒的功效与作用tive environment.An experiment can employ strong frames exclusively,weak frames exclusively,or a mixture of strong and weak frames.
Taken together,variations in the relative quantities and strengths of frames combine to yield eight possible rearch conditions or competitive contexts(a ninth cell is not applicable).Table1indicates that almost all previous work investigates asymmetric one-sided designs using either strong or weak frames.The main excep-tion is Sniderman and Theriault’s(2004)balanced study using(apparently)strong frames(also e Brewer&Gross,2005).The most noteworthy feature of the table is thefive study designs of competitive situations that heretofore have not been imple-mented.The include dual studies in which opposing frames are received in equal quantity but are of unequal strength and asymmetric two-sided studies that expo individuals to opposing combinations of strong and weak frames in unequal quantities.
We find remarkable that the voluminous literature on framing effects has ignored perhaps the typical framing situation in which competing sides promote alternative interpretations of an issue.How do individuals respond to competitive frames of varying quantities and strengths?
饱览造句The literature suggests two possibilities.One hypothesis,focusing on the relative volume of competin
g messages,posits that whichever frame is loudest—that is, repeated most frequently—will have the greatest influence on an individual’s opin-ions,all el constant.Chong(1996,p.222)summarizes this perspective:‘‘Models of information transmission imply that the ideological faction that expends sufficient resources on propaganda and manipulation,and that nds sufficiently loud signals, can always prevail in defining the terms of debate.’’Similarly,Zaller(1992,p.311) states that citizens‘‘are blown about by whatever current of information manages to develop with the greatest intensity’’(also e Cappella&Jamieson,1997;Domke, Shah,&Wackman,1998;Iyengar,1991;Nabi,2003).In this view,the relative strength of the frame is not pertinent becau(as we will elaborate shortly)individ-uals are assumed not to evaluate strength consciously but simply to embrace the frame they hear most often and that most easily comes to mind.3
醋泡香蕉
An alternative hypothesis is that the strongest frame will exert the greatest influ-ence on individual opinion,regardless of repetition,all el constant.This prediction follows from work on strong and weak frames,which has generated a partial list of factors that contribute to a frame’s strength.This rearch suggests that a frame’s strength increas,for example,when it comes from a credible source(Druckman, 2001b),resonates with connsus values(Chong,2000),and does not contradict strongly held prior beliefs(Brewer,2001;Druckman&Nelson,2003;Haider-Markel &Joslyn,2001;Shah,Domke,&Wackman,1996).
104Journal of Communication57(2007)99–118ª2007International Communication Association
