Words: 7832
Reformulating News Media Bias: A New Theoretical and Methodological Approach By Peter Brinson Allegations of media bias are nothing new in the United States. Though conservatives have been the most vocal in recent years, liberals have also been known to argue that the news media systematically presents information in a way that privileges the opposition’s viewpoint. This debate has been carried out in the popular press as well, with each side struggling to provide the definitive proof that the media disadvantages their side in reporting the news.
In this paper, I argue that the question about whether or not the news media are biased is the wrong question to ask, not only because each side will always be able to construct an answer that suits them by ignoring evidence to the contrary, but also because producing an un-biased news media is simply not possible. Because there is no broadly agreed-upon definition of “bias,” debates about bias resemble arguments about some mythical beast that no one has ever seen.
What matters is not whether news media are biased in some abstract, essentialist sense, but how particular media outlets cover particular stories in more or less biased ways than their real-life counterparts. In this paper, I first survey the scholarly literature on news media bias in the sense commonly assumed in the claims of political partisans, the privileging of one political party/candidate/ideology over the other. Overwhelming evidence shows that mainstream news media taken together do not exhibit any systematic preference for either Democrats or Republicans.
Next, I argue that such null findings do not mean that media are “unbiased;” rather, there are many ways in which news content is shaped and framed in ways that make the news a distorted picture of reality. A bewildering array of structural factors, ideological factors, and individual actions at all stages of the media production process all contribute to the final product: the news as a subjective, social construction rather than a mirror image of reality. Given such overwhelming evidence of news distortion, I argue that the debate about whether or not news media are biased is a sterile one that yields an uninteresting answer: yes.
Instead, scholars need to shift their focus towards comparative research that asks which media outlets are more or less biased compared to others, and about what subjects are media outlets more or less likely to be biased. This pragmatic conception of media bias offers several important advantages over the essentialist conception that is the object of almost all studies of bias, and I show how those advantages can be taken into account methodologically by proposing a program for future research. Partisan Political Bias
While a host of popular works has claimed that news media are either biased against liberals or biased against conservatives, most works rely mostly on anecdotal evidence rather than any systematic study of the phenomenon. While substantively intriguing, a handful of isolated examples of news bias cannot justify the broad claim that news media are inherently biased. However, there have been a handful of rigorous studies that have claimed to produce concrete evidence of systematic partisan political bias, whether towards the right or the left. Before proceeding further, these works must be addressed.
Of the works arguing that news media are systematically biased towards the conservative viewpoint, no other has been as influential as Manufacturing Consent (Herman and Chomsky 1988). Employing detailed content analyses of news coverage of several important international events from the 1970s-80s that have ramifications for U. S. foreign policy, Herman and Chomsky advance a “propaganda model” that “suggests that the ‘societal purpose’ of the media is to inculcate and defend the economic, social, and political agenda of privileged groups that dominate the domestic society and the state” (298).
While this work is often taken as evidence of a conservative bias in the news media, the analysis is oriented towards the political system as a whole, regardless of which party is in power. As will be discussed later, Herman and Chomsky’s work does shed light on the problem of media bias, but not in a partisan way. On the other side, perhaps the most influential scholarly work taken as conclusive evidence of a liberal bias in U. S. news media is Patterson and Donsbach’s (1996) study of the ways that journalists’ political beliefs drive their decision-making processes in the newsroom.
In surveys given to journalists in the U. S. , Britain, Germany, Italy, and Sweden, Patterson and Donsbach find that in all five countries, journalists rate their political beliefs as more liberal than both the news organizations they work for and the population at large. Moreover, they attempt to demonstrate that journalists make decisions in the newsroom based on their political beliefs by giving journalists a list of fictional news scenarios and asking them to make judgments about the appropriateness of possible headlines and images that would accompany the story.
While they conclude that “partisanship can and does intrude on news decisions, even among journalists who are conscientiously committed to a code of strict neutrality” (466), they make three significant methodological errors that render this conclusion untenable. First, they assume that journalists’ responses to survey questions are an accurate indicator of how they actually make decisions in the newsroom, while in reality, the survey questions bear no resemblance to the processes of story construction and selection that actually occurs.
Second, the majority of their evidence for partisan political bias comes from Germany, Italy, and Sweden, countries with media systems that explicitly encourage partisan journalism. It is simply inappropriate to judge the U. S. media system by a different standard. Finally, their findings appear large by one measure, but quite insignificant by another measure. For the United States, individual-level correlations between journalists’ political beliefs and their selection decisions were . 09 and . 08 for each of the scales they used, a small correlation by any standard (463).
Thus, in the final analysis, the only evidence Patterson and Donsbach present for a liberally-biased news media are the political values of the journalists, which are, according to their own standards, more centrist than liberal. Despite their claims to the contrary, Patterson and Donsbach present no evidence that journalists shape the news according to a liberal political agenda. A similar pattern of conclusions not matching evidence is found in another work arguing that a liberal bias in news media exists.
In his new book, Press Bias and Politics, Jim Kuypers (2002) compares the text of six speeches made by political figures about race and homosexuality with the ways that the news media covered the events. He argues that the comparisons show that “only a narrow brand of liberal thought is supported by the press; all other positions are denigrated” (19). Kuypers’ comparisons of the original texts of speeches with the ways the speeches are covered in the news media do indeed show convincingly that the news media do not report the original meanings and intentions of the speeches; however, he presents no evidence that the resulting bias is liberal.
Any resemblance to liberal bias clearly results from his selection of examples to analyze, while the actual news media distortion appears to stem from non-partisan processes of framing,, “typification”, and simplification that journalists must employ in order to deal with the time and space constraints they face in creating the news (Tuchman 1978). While his empirical examples are excellent illustrations of these processes, his conclusions of liberal bias in the news media are based only on his political beliefs and not on any of the evidence he presents.
In support of these null findings, almost all systematic studies of news media content show little to no bias in the news media, either towards liberals or conservatives. Hofstetter’s (1976) study of the 1972 election campaigns demonstrates that, in aggregate, there is no evidence of systematic media bias against either of the Presidential candidates, their political parties, or the issues favored by one side or the other. This work, by measuring inequalities of news coverage in dozens of different ways, shows vividly why the debate about media bias continues.
For example, while Democratic challenger George McGovern received more press coverage than President Nixon, Nixon received more positive news coverage and less negative news coverage than McGovern (54). Depending on what the object of comparison is, how one measures coverage, how one measures bias, and what media outlet is studied, one can arrive at drastically different conclusions. The strength of Hofstetter’s work is that he considers all of these possibilities in order to make a judgment about news bias, considered in aggregate.
Similarly, a team of researchers who studied the 1992 elections also found that the news media treated both the Republican and Democratic candidates in a non-partisan way. In a methodologically sophisticated study, Dalton et al (1998) employ content analysis from a representative sample of U. S. newspapers and interviews with a sample of people living in the counties where these newspapers are located in order to determine to what extent newspapers presented evaluative content (“partisan cues”) of the candidates.
The authors find that there was no significant correlation between editorial stance and partisan cues in news content, and furthermore, that there was no significant correlation between the news content and readers’ perceptions of the newspaper’s political leanings. Subsequent analysis of these data demonstrated that people’s personal networks, membership in secondary organizations, and newspaper editorials were much greater sources of partisan messages than the news content itself (Beck, Dalton et al. 2002).
On the whole, the analysis of the 1992 election shows that newspapers consistently present both positive and negative partisan cues for each candidate to such an extent that newspaper readers cannot distinguish any real partisan bias. In short, the newspapers present a relatively balanced portrait of the campaign as they are supposed to. Taken together, research on elections shows irrefutable evidence of non-partisan news media coverage. D’Alessio and Allen (2000) conducted a meta-analysis of 59 studies of news media bias in 13 Presidential election campaigns spanning half a century and found no significant evidence of bias.
Even beyond election campaigns, the findings consistently show that non-partisan journalism is the norm. In two separate inquiries, David Niven demonstrates that the news media treat both Republicans and Democrats equally, whether they are governors who succeed in lowering the murder and unemployment rates in their states (Niven 1999) or members of Congress who switch party affiliation (Niven 2003). These two studies are significant because, unlike the election studies, each controls for “objective” variables of political behavior (switching party affiliation) or political success (lowering murder and unemployment rates).
Holding objective conditions constant, any observed differences in coverage would be plausibly due to news media bias; however, no such differences are found, thus adding further proof that news media do not exhibit a partisan bias. Explaining Non-partisan Political Reporting The inability of researchers to find strong evidence of media partisanship in news reports is significant, especially considering that such research has been conducted across a thirty-year time span, using a variety of research methods and studying a variety of different people, events, and media outlets.
What would make the case for media non-partisanship even stronger would be to explain the reasons why researchers consistently are unable to detect partisan media bias. Here, I advance three reasons for the null findings, having to do with the empirical world around us, the profession of journalism, and the ways audiences interpret media content. First, as Niven’s research above highlights, the objective world around us does not make it easy to discern media bias because reality does not treat Republicans and Democrats equally.
As illustrated by Kuklinski and Sigelman (1992) in what they call the “paradox of objectivity,” when one party becomes more influential or powerful than another, the news media will reflect it, given the norms and routines of newsgathering. Following Reagan’s 1980 victory in the Presidential election and a rapid increase in the number of Republican Senators, Republicans became more frequently mentioned in the news media.
Such an inequality does not represent a bias against Democrats, just the journalists’ reproduction of reality in which Democrats became relatively less powerful and less visible. In fact, one might argue that to attempt to present both sides equally regardless of their actual stature in reality would be a form of news bias. However, it is not an easy question to determine under what circumstances journalists merely try to reflect reality and under what circumstances journalists should go out of their way to call to attention aspects of reality that one might not notice otherwise.
Althaus (2003) finds that reporters do act as autonomous individuals seeking out oppositional viewpoints for the news, rather than merely parroting views they find from traditional sources, and while Althaus praises such a finding as evidence of an independent press unconstrained by the voices of power, Patterson (1997) worries that reporters are too freely allowed to let their own political agenda influence their work. Given such a conundrum, it is easy to see why scholars of media ias have had difficulty conceptualizing and detecting media bias. Additionally, we must note the difficulty in measuring the effects of media bias (on vote choice or attitudes, for example), since it would be difficult not only to isolate the independent effects of media from all other possible causes, but also because people tend to have well-formed attitudes to begin with (Bartels 1993). A second reason for the consistent null findings on media bias have to do with the norms and routines of journalists, especially the norm of objectivity.
Simply put, professional journalists are trained to keep their partisan political beliefs out of the newsroom to ensure that no bias towards either liberals or conservatives is conveyed to the general public. Schudson’s (1978) account of the rise of objectivity after World War I clarifies that the norm of objectivity was created as a journalistic technique to ensure that news could be trusted, that it would be different from government propaganda and corporate public relations campaigns.
In a sense, the norm of objectivity is a bedrock feature of American journalism (although it is by no means unchallenged). The norm of objectivity is embodied in journalists’ values???the belief in presenting as close to an objective picture of reality as possible and the belief that professionals should be unencumbered by political or economic pressures when presenting the news???and also in their actions???getting the “other side’s” point of view to balance out a news story.
The classic ethnographic studies of newsrooms by Tuchman (1978) and Gans (1979) remain the most thorough explications of how professional norms of behavior and the values of journalists influence the construction of the news, and Gans’ work in particular illustrates the strict division of labor that newsrooms try to uphold so as to prevent news executives and powerful political and economic interests from unduly influencing or censoring the news.
As a result of such professional norms, Mutz and Martin (2001) argue that people are exposed to more cross-cutting political viewpoints through news media than they are through their own social networks[1]. Although there is little disagreement that professional norms help to ensure that there is no partisan political bias in the content of the news, recent works have emphasized that such professional journalistic norms may inadvertently cause more problems of bias than they solve (Schudson 2003).
Deeper conceptions of media bias will be considered below; for the time being, it is enough to highlight the ironic finding of one team of researchers that claims of liberal media bias by conservatives actually were both cause and consequence of journalists’ practice of reporting on claims of media bias (Domke, Watts et al. 1999; Watts, Domke et al. 1999). In this way, ironically, professional norms helped to fan the flames of those alleging media bias. The final and most compelling explanation of the consistent finding that news media do not have a partisan bias (even though everyone thinks they do) is the “hostile media effect. Vallone et al (1985) first noted the phenomenon when they showed that pro-Israeli and pro-Arab partisans both viewed the same news program about the Middle East and evaluated them to be hostile to their respective viewpoints. Since this first study, the hostile media effect has been replicated or invoked by a variety of scholars to explain why people with opposite viewpoints perceive the same media content to be biased in favor of the opposing side (Dalton, Beck et al. 1998; Niven 1999; Gunther, Christen et al. 2001; D’Alessio 2003).
The hostile media effect has been found to depend on the issue (Gunther and Christen 2002) and on a person’s intergroup context (Duck, Terry et al. 1998), and it appears that the effect is limited to mass media (Gunther and Schmitt 2004). The most intriguing research on the hostile media effect, however, concerns the psychological mechanisms that are at work. Schmitt et al (2004) argue that what accounts for the different partisan evaluations is “selective categorization,” the different ways that partisans evaluate the same piece of information as either for or against their own position.
It is clear in the context of partisan bias how such a mechanism might work. If a neutral statement were given about an issue or candidate, partisans on either side would interpret the neutral statement as being too far towards the opposing viewpoint, relative to their own. Even though these authors reject alternative psychological explanations for the hostile media effect, it seems plausible that, depending on the issue, selective recall of media content and selective salience would make a difference in the viewer’s ultimate judgment about the bias of the media content.
Finally, Gunther et al (2001) place the hostile media effect in the context of other psychological processes that result in perceptions of public opinion. In particular, they show that the hostile media effect works in tandem with two other processes???the “extrapolation hypothesis” (people generalize the perceived bias from a single article to news media in general) and the “persuasive press inference” (the belief that media coverage has substantial influence on the opinions of others)???in order to create a perception of public opinion.
Additionally, they observe that the hostile media effect works in an opposite direction than the projection bias, in which people tend to overestimate the extent to which other people share their beliefs. Thus, the hostile media effect seems to be related to an individual’s concern that the mass media have greater power and reach than person-to-person communication (Gunther, Christen et al. 2001). This finding helps to explain why only mass media seem to be affected by the hostile media effect, while other forms of communication do not.
Other Forms of Media Bias Even though the evidence against partisan media bias is overwhelming, it would be unwise to conclude that the discussion of media bias can be simply written off. Indeed, an equally overwhelming body of research shows that the news media are biased in other ways, albeit in ways that defy simple description. Here, I want to discuss four sets of processes or factors that combine to produce news content that is not a mere reflection of reality.
Rather than fully explore any of these processes, however, the aim of this brief overview is to illustrate my contention that researchers should stop asking whether or not news media are biased. The answer is clearly “no” if one is talking about strict partisan bias (as has already been shown) and “yes” if one is talking about bias in a broader sense (as will be shown here). As this section demonstrates, news content is subject to such a bewildering array of contesting social, economic, political, psychological, professional, and institutional forces that the final news content cannot help but be significantly different from reality.
Rather than continue to build research projects around some essentialist notions of “bias” or “reality,” which can never be satisfactorily defined or operationalized and hence never clearly answered, the evidence presented here suggests that scholars begin to ask different kinds of questions about media bias. The first set of processes that shape news media content are political and economic interests that are implicated in the organization or the content of the news.
There is a rich tradition of research on the political-economy of media, which points to the power of political and economic interests to constrain the actions of journalists. In the United States, much of this research has emphasized that most mainstream news outlets are for-profit businesses, first and foremost, and that by definition, they must be oriented towards staying profitable and competitive (McChesney 1999; Bagdikian 2000; Croteau and Hoynes 2001).
All of these authors have pointed out how the need to sell an audience to advertisers can make media organizations act in ways that may not be in the public’s interest. Prior (2005) in particular has shown very vividly how the rapid expansion of news and entertainment options on cable television and the internet, perhaps the most noticeable manifestation of market forces at work in the media industry today, has made it less likely that audience members will choose to watch the news over some form of entertainment. Other scholars have emphasized the inter-related political and economic interests in shaping news content.
McChesney (1993) demonstrates how the interests of commercial broadcasters triumphed over media reformers in the debate over government regulation of the radio industry, thus resulting in an institutional structure of the radio industry that has privileged commercial broadcasting ever since. Herman and Chomsky (1988) emphasize the power of U. S. government sources and ideology in shaping news coverage of foreign policy, and they also point out that powerful individuals can give news organizations “flak,” or negative criticism or retaliation, if they produce news content that disagrees with the dominant view.
Such political and economic interests have been shown to intervene directly in the news production process on occasion by dictating what stories can and cannot be produced (Bagdikian 2000) or by dictating how issues are covered in the news (Gilens and Hertzman 2000). Considering the power of dominant political and economic interests, there is strong evidence for a news bias in favor of dominant groups in society. A second set of interrelated processes that shape news content have to do ith the institutional constraints on journalists and professional norms and behaviors of journalists. As discussed earlier, professional journalistic norms of objectivity help mitigate the problem of partisan media bias, but as Schudson (2003) points out, these same norms create more problems than they solve. Journalists face a wide array of constraints that they must overcome in order to produce the news, primary among them are a limited amount of time, limited amounts of space for news, and limited resources available to news organizations.
What’s worse, these constraints are getting tighter for journalists as resources and newsrooms shrink while the demand for news expands (Klinenberg 2005). In an effort to produce news as efficiently as possible, newsrooms have been organized into “beats,” journalists have developed routines and “typifications” for handling stories, journalists must make quick judgments about the “newsworthiness” of potential stories, and journalists have learned how to produce news stories as efficiently as possible (Tuchman 1978; Gans 1979; Kaniss 1991).
One way to ensure that all of this happens is to privilege powerful political and economic actors as sources, since whatever they do is almost always newsworthy, they are considered trustworthy sources, and they enable a journalist to quickly find the “other” viewpoint to balance the story (Gans 1979). This over-reliance on powerful sources tends to reinforce the status quo bias exerted by political and economic interests over news content.
In a study of domestic and foreign policy items in television news, Danielian and Page (1994) show that government actors comprise over half of the sources cited, and business groups comprised the largest segment of interest groups that were cited as sources. The views of ordinary citizens or citizen interest groups are much less common in the news. Additionally, journalists’ actions are shaped by a set of “enduring values” that are derived from general American culture, and these values are usually reflected in the final news content (Gans 1979).
Gans labels these values ethnocentrism, altruistic democracy, responsible capitalism, small-town pastoralism, individualism, moderatism, social order, and national leadership. Thus, the institutional demands of the news organization and the professional values of journalists all exert powerful forces over the production of news, and again, there is strong evidence for a bias towards dominant political and economic actors and toward reproduction of mainstream American values. In act, the institutional, professional, political, and economic exert such an influence over media organizations, and their political effects are so notable, that Cook (2005) has argued for media, in all their diversity, to be treated as a single, coherent political institution. Finally, scholars have identified processes of agenda-setting and framing at work in the news media. These two processes link the news production processes described above with the ways that the audience interprets the news and is affected by it.
Agenda-setting refers to the ability of news media to call attention to particular issues or to declare something newsworthy. Erbring et al (1980), for example, show that newspaper exposure and newspaper content can increase the salience of particular issues to audiences, depending on the audience member’s likelihood of already finding the issue to be salient. The process of agenda-setting is related to psychological issues of salience and priming, which demonstrate that people base their public opinions on facts and ideas that are most readily accessible in their minds (Krosnick and Kinder 1990; Zaller 1992).
Thus, news media have the power to bias public opinion by making certain issues, people, or beliefs more or less salient in people’s minds. Recent work has criticized media-centered agenda-setting based on both empirical and theoretical short-comings. Indeed, even the clearest cases of agenda-setting by the media ultimately must rely on other variables???the nature of the audience (Krosnick and Kinder 1990) or the agendas of political actors (Dalton, Beck et al. 1998)???to make a compelling case. Entman (1989, p. 49) has criticized the concept of agenda-setting as misleading on the grounds that “The only way to influence what people think is precisely to shape what they think about. ” Consequently, recent scholarly attention has been devoted to the process of framing. Gitlin (1980 (2003), p. 7) defines media frames as “persistent patterns of cognition, interpretation, and presentation, of selection, emphasis, and exclusion, by which symbol-handlers routinely organize discourse, whether verbal or visual. ” Entman (1993) suggests that frames simultaneously define problems, diagnose causes, make moral judgments, and suggest remedies.
The process of “framing” allows journalists to produce the news efficiently, but the resulting media frames necessarily distort reality in such a way that certain aspects of a story are emphasized while others are ignored (Kuypers 2002; Schudson 2003). For example, news usually emphasizes events and individuals rather than long-term conditions, it emphasizes the bad over the good, it emphasizes the “horse-race” in politics rather than issues, and it emphasizes “official” views over others (Schudson 2003). Thus, not only does framing shape news content, but framing also influences how audiences perceive the news.
Experiments conducted by Nelson et al (1997) about tolerance for Ku Klux Klan activity have shown that different frames applied to the same story successfully influence audience’s attitudes about the issue. Gamson and Modigliani (1989) also show how framing of controversial issues shapes public opinion. Ultimately, frames enable or constrain the desire or ability to act, whether individually or collectively, so the political ramifications of framing should not be taken lightly (Snow, Rochford et al. 1986; Snow and Benford 1988; Gamson 1992).
A Pragmatic Research Agenda of Media Bias As the above literature review has shown, the question of whether or not news media are biased has a simple answer: the news media are not biased in a simple partisan way, but the news media are biased in other ways, due to a wide array of political, economic, institutional, professional, and social-psychological factors. Continued debates about whether or not news media are biased are not only redundant, but they are counter-productive in the sense that they distract scholars’ attention from a more fruitful line of inquiry.
I argue that scholars should stop asking whether or not bias exists and instead ask how much bias exists in the news content of a particular media outlet and why the different (or similar) amounts of bias are observed. Studying media bias in pragmatic, real-world terms, based on the differences that one observes among media outlets offers numerous conceptual and methodological advantages. First, the only reason that scholarly debate about the existence of media bias still continues is that there is no real agreement on what constitutes media bias and what should be conceived of as something else.
There is no widely agreed upon definition of “bias,” so any study can be constructed around the desire to find (or not find) evidence of bias, however conceived. Furthermore, any study can be criticized for not taking into account particular forms or sources of bias. The solution, however, is not to attempt to construct a universal definition of bias. Such an endeavor would ultimately be a fruitless task that would privilege some points of view at the expense of others. “Bias” is not an objective entity whose essence can be clearly identified and delimited by abstract theorizing.
Instead of this essentialist notion of bias, I argue that we should conceive of bias as a variable, like mass or weight, that is in some ways defined by our measurement of it. Conceiving of bias in this way would close off the unproductive and politically-charged debate about whether or not the news media are biased and open up an inquiry that would help determine which media outlets or which issues are subject to more bias than others. Such information is immediately useful, not only to scholars but also to consumers, and it results in a definition of bias that is grounded in reality.
Second, conceiving of bias as a variable would introduce an important comparative dimension to our study of media bias, both theoretically and empirically. Studies of media bias should not be compared to fictional or idealistic standards, but to real-life, observable standards. It would be hard to imagine a media outlet that was completely unbiased or completely biased; such media outlets like have never existed and will never exist. It is more appropriate to compare media outlets to one another, to establish a baseline of what is possible or realistic and compare the relative extent to which media content is biased.
In short, bias should be studied relatively, not according to absolutely. Additionally, media bias is only as real as its observable effects; if there is no difference in media content across media outlets, then it is not helpful or convincing to argue that the media are biased. There must be a comparative standard, so that researchers can ask, “biased compared to what? ” Comparing media outlets to one another in terms of bias can help more accurately identify evidence of media bias.
If one media outlet behaves differently than all others, there is much stronger proof of bias than if all media outlets behave the same. Methodologically, this increases scholars’ ability to quantify and measure media bias and makes the results more reliable. Third, treating media bias as a variable outcome that is more or less present across different media outlets can help to more accurately identify the causes of bias. A claim that news media are biased must be able to be supported by identifying the mechanism that causes bias.
Using a comparative approach enables researchers to more easily identify the differences in media bias and the differences in media content and media organization that are responsible for creating the bias. Current research on media bias is hampered in that it is unable to conclusively distinguish among the dozens of possible sources of media bias described above; a comparative analysis that is sensitive to varying amounts of bias can more accurately predict which causal mechanisms are more and less responsible for shaping news content.
In this way, a shift towards a pragmatic research agenda and a variable conception of bias can extend our understanding of the relative importance of various causal mechanisms in shaping the final content of the news. Finally, this pragmatic conception of bias shifts the focus in the debate about media bias towards the real-life consequences of media bias and what can be done about them. If researchers find that one media outlet is significantly more biased than others, then steps can be taken by media owners, managers, and journalists to correct such a bias.
Additionally, consumers can make more informed decisions about which media outlets they want to use and can make more informed judgments about how to interpret the content that they consume. For researchers, they can use such results in order to generate more complete models of media production, media effects, audience reception, and public opinion. None of this can be accomplished, however, if researchers remain fixated on abstract, essentialist conceptions of bias and endless partisan debates about whether or not the mass media are biased.
Researchers should adopt a pragmatic, comparative research agenda and a conception of bias as a variable. Such measures would ultimately yield better theories of the media, more accurate results about claims of media bias, a fuller understanding of the causes of media bias, greater knowledge about how media organizations are different or similar, and enhanced ability of individuals and journalists to act to improve the quality of the news media. To conclude, I will briefly enumerate the essential components of such a research agenda, so that such goals might ultimately be able to be achieved.
Media Bias: A New Research Agenda In order to carry out research based on the pragmatic conception of media bias advanced above, several components would be necessary. Above all, the research must be based on a comparative design, and it should take into account all stages of the news production process, from the actual event being covered through the effects of the news coverage on the audience. To my knowledge, no research has attempted to include all of these stages in a single study, although Dalton and his colleagues (Dalton, Beck et al. 998) provide a good model from which we can base this proposal. The first requirement for constructing a thorough comparative study of media bias is a detailed or even “objective” account of the actual event as it happened. Unfortunately, most studies of media bias exclude this component in their research; because much of the studies have been done on election campaigns, it would be impossible to construct a neutral account. However, Kuypers (2002) demonstrates how such a task might be accomplished.
By focusing on the texts of speeches given by important political figures, Kuypers has a written account of exactly what was said, and it can be compared to the way the speech was reported in the press. Now, thanks to the omnipresence of video cameras, almost any speech, legislative hearing, or even social movement protest has some sort of official record that can be analyzed objectively by the researcher. By performing a content analysis on the text or transcript of some event, scholars can establish some “objective” baseline to which media bias can be compared.
This first step in the research is essential to making a convincing case for media bias, both because some possible sources of bias occur in the observing of the actual event, and because an analysis of media bias should be able to show how the biased account is different from what actually happened. The second component necessary for a thorough comparative study of media bias is content analysis of the news coverage of the event in question as it is covered across a number of different media outlets.
Almost all studies of media bias utilize analyses of multiple media outlets, although Dalton et al’s (1998) representative sampling of U. S. counties and their major newspaper(s) seems to be a model study design. In order to ensure that the comparisons minimize the additional sources of variance, the comparisons should occur within medium (only newspapers, only television, etc) and within organizational type (local or national, commercial or noncommercial). By comparing media organizations of the same type, researchers can state with greater confidence that x% the observed differences in news reporting is due to media bias.
Finally, the content analysis conducted of the news content should be significantly similar to the content analysis of the actual event to ensure that the same standards are applied to both the actual event and the news coverage of the event. By combining these first two steps, a tentative argument about media bias can be made, although it can be improved upon. The third component necessary for a thorough comparative study of media bias is some measure of audience attitudes about the event itself and about the news coverage.
A wide variety of methods could be used to gauge audience attitudes, such as focus groups or interviews, though surveys distributed to people who witnessed the news coverage of the event is perhaps the most efficient way to collect data on the audience. If the research focused on a specific event, an experimental design would be optimal for distributing the “audience” randomly between control and experimental conditions. However, on a larger event like an election campaign, surveys could be distributed by mail or conducted via telephone.
Among the topics that are essential for such a survey are the standard battery of demographic questions, questions about media use, questions about their attitudes about the relevant issues, and questions about their assessment of the news coverage of the event. The appropriate questions on the survey would of course depend on what issue was being studied and what media outlets were being studied. This step is crucial for the research design because it can test for differences in audience that might shape the news content and it can provide stronger evidence for a claim of media bias.
For example, if respondents were unable to correctly identify the bias in the media content and if the hostile media effect were observed, such findings could substantially weaken the claim of media bias. Conversely, if audience evaluations of one media outlet differed substantially from audience evaluations of other media outlets, we would have very strong evidence of media bias. Finally, although not an essential component, an additional measure of audience effects would improve the comparative study of media bias. For example, Dalton et al (1998) are able to compare findings of media coverage with the difference it made on voter choice.
Having such data would allow researchers to determine what difference media bias makes on the attitudes and behaviors of audience members. As another example, a recent poll about the war in Iraq showed that one’s knowledge about the war varied considerably depending on which television network respondents listed as their primary source of information (Kull 2003). This study was notable in the strength of evidence it presented that some combination of audience selection and media bias was having significant impacts on people’s knowledge and support for the war.
Since such information could also be obtained by survey or by interviews, it would cost little else to add such measures, and it can make a strong case for the importance of the presence or absence of media bias. By constructing a study of news media bias in this way, social scientists can fruitfully advance our knowledge of media bias, our understanding of the causes of media bias, and the real-life effects media bias has on political attitudes and political behavior.
Rather than remain mired in unproductive, partisan debates about whether or not the media in general are biased, a shift towards conceptualizing media bias as a variable to be compared across media outlets will advance scientific knowledge of the media, and it will enable concerned citizens to make more informed judgments about the media or even attempt to change the news media to make it more fit to serve democracy. References Althaus, S. L. (2003). “When News Norms Collide, Followthe Lead: New Evidence for Press Independence. ” Political Communication 20: 381-414.
Bagdikian, B. H. (2000). The Media Monopoly. Boston, Beacon Press. Bartels, L. M. (1993). “Messages Received: The Political Impact of Media Exposure. ” American Political Science Review 87(2): 267-285. Beck, P. A. , R. J. Dalton, et al. (2002). “The Social Calculus of Voting: Interpersonal, Media, and Organizational Influences on Presidential Choices. ” American Political Science Review 96(1): 57-73. Cook, T. E. (2005). Governing With the News: The News Media as a Political Institution. Chicago, University of Chicago Press. Croteau, D. and W. Hoynes (2001).
The Business of Media: Corporate Media and the Public Interest. Thousand Oaks, CA, Pine Forge Press. D’Alessio, D. (2003). “An Experimental Examination of Readers’ Perceptions of Media Bias. ” Journalism and Mass Communication Quarterly 80(2): 282-294. D’Alessio, D. and M. Allen (2000). “Media Bias in Presidential Elections: A Meta-Analysis. ” Journal of Communication 50(4): 133-156. Dalton, R. J. , P. A. Beck, et al. (1998). “Partisan Cues and the Media: Information Flows in the 1992 Presidential Election. ” American Political Science Review 92(1): 111-126. Dalton, R. J. , P. A. Beck, et al. 1998). “A Test of Media-Centered Agenda Setting: Newspaper Content and Public Interests in a Presidential Election. ” Political Communication 15: 463-481. Danielian, L. H. and B. I. Page (1994). “The Heavenly Chorus: Interest Groups on TV News. ” American Journal of Political Science 38(4): 1056-1078. Domke, D. , M. D. Watts, et al. (1999). “The Politics of Conservative Elites and the ‘Liberal Media’ Argument. ” Journal of Communication 49(4): 35-58. Duck, J. M. , D. J. Terry, et al. (1998). “Perceptions of a Media Campaign: The Role of Social Identity and the Changing Intergroup Context. Personality and Social Psychology Bulletin 24(1): 3-16. Entman, R. M. (1989). “How the Media Affect What People Think: An Information Processing Approach. ” Journal of Politics 51(2): 347-370. Entman, R. M. (1993). “Framing: Toward Clarification of a Fractured Paradigm. ” Journal of Communication 43(4): 51-58. Erbring, L. , E. N. Goldenberg, et al. (1980). “Front-Page News and Real World Cues: A New Look at Agenda-Setting by the Media. ” American Journal of Political Science 24(1): 16-49. Gamson, W. A. (1992). Talking politics. Cambridge [England] ; New York, NY, USA, Cambridge University Press. Gamson, W.
A. and A. Modigliani (1989). “Media Discourse and Public Opinion on Nuclear Power: A Constructionist Approach. ” American Journal of Sociology 95(1): 1-37. Gans, H. J. (1979). Deciding What’s News: A Study of CBS Evening News, NBC Nightly News, Newsweek, and Time. New York, Pantheon Books. Gilens, M. and C. Hertzman (2000). “Corporate Ownership and News Bias: Newspaper Coverage of the 1996 Telecommunications Act. ” Journal of Politics 62(2): 369-386. Gitlin, T. (1980 (2003)). The Whole World is Watching: Mass Media in the Making and Unmaking of the New Left. Berkeley, University of California Press.
Gunther, A. C. and C. T. Christen (2002). “Projection or Persuasive Press? Contrary Effects of Personal Opinion and Perceived News Coverage on Estimates of Public Opinion. ” Journal of Communication 52(1): 177-195. Gunther, A. C. , C. T. Christen, et al. (2001). “Congenial Public, Contrary Press, and Biased Estimates of the Climate of Opinion. ” Public Opinion Quarterly 65: 295-320. Gunther, A. C. and K. Schmitt (2004). “Mapping Boundaries of the Hostile Media Effect. ” Journal of Communication 54(1): 55-70. Herman, E. S. and N. Chomsky (1988). Manufacturing Consent: The Political Economy of the Mass Media.
New York, Pantheon Books. Hofstetter, C. R. (1976). Bias in the News: Network Television Coverage of the 1972 Election Campaign. Columbus, Ohio, Ohio State University Press. Kaniss, P. C. (1991). Making Local News. Chicago, University of Chicago Press. Klinenberg, E. (2005). “Convergence: News Production in a Digial Age. ” The Annals of the American Academy of Political and Social Science 597(1): 48-64. Krosnick, J. A. and D. R. Kinder (1990). “Altering the Foundations of Support for the President Through Priming. ” American Political Science Review 84: 497-513. Kuklinski, J. H. and L.
Sigelman (1992). “When Objectivity is Not Objective: Network Television News Coverage of U. S. Senators and the ‘Paradox of Objectivity’. ” Journal of Politics 54(3): 810-833. Kull, S. (2003). Misperceptions, The Media, and the Iraq War. College Park, MD, Program on International Policy Attitudes. Kuypers, J. A. (2002). Press Bias and Politics: How the Media Frame Controversial Issues. Westport, Connecticut, Praeger. McChesney, R. W. (1993). Telecommunications, Mass Media, and Democracy: The Battle for the Control of U. S. Broadcasting, 1928-1935. New York, Oxford University Press.
McChesney, R. W. (1999). Rich Media, Poor Democracy: Communication Politics in Dubious Times. Urbana, University of Illinois Press. Mutz, D. C. and P. S. Martin (2001). “Facilitating Communication Across LInes of Political Difference: The Role of Mass Media. ” American Political Science Review 95(1): 97-114. Nelson, T. E. , R. A. Clawson, et al. (1997). “Media Framing of a Civil Liberties Conflict and its Effects on Tolerance. ” American Political Science Review 91(3): 567-583. Niven, D. (1999). “Partisan Bias in the Media? A New Test. ” Social Science Quarterly 80(4): 847-857.
Niven, D. (2003). “Objective Evidence on Media Bias: Newspaper Coverage of Congressional Party Swtichers. ” Journalism and Mass Communication Quarterly 80(2): 311-326. Patterson, T. E. (1997). “The News Media: An Effective Political Actor? ” Political Communication 14: 445-455. Patterson, T. E. and W. Donsbach (1996). “News Decisions: Journalists as Partisan Actors. ” Political Communication 13: 455-468. Prior, M. (2005). “News vs. Entertainment: How Increasing Media Choice Widens Gaps in Political Knowledge and Turnout. ” American Journal of Political Science 49(3): 577-592.
Schmitt, K. M. , A. C. Gunther, et al. (2004). “Why Partisans See Mass Media as Biased. ” Communication Research 31(6): 623-641. Schudson, M. (1978). Discovering the News: A Social History of American Newspapers. New York, Basic Books. Schudson, M. (2003). The Sociology of News. New York, W. W. Norton & Co. Snow, D. A. and R. D. Benford (1988). “Ideology, Frame Resonance, and Participant Mobilization. ” International Social Movement Research 1: 197-217. Snow, D. A. , E. B. Rochford, Jr. , et al. (1986). “Frame Alignment Processes, Micromobilization, and Movement Participation. American Sociological Review 51(4): 464-481. Tuchman, G. (1978). Making News: A Study in the Construction of Reality. New York, Free Press. Vallone, R. P. , L. Ross, et al. (1985). “The Hostile Media Phenomenon: Biased Perception and Perceptions of Media Bias in Coverage of the Beirut Massacre. ” Journal of Personality and Social Psychology 49(3): 577-585. Watts, M. D. , D. Domke, et al. (1999). “Elite Cues and Media Bias in Presidential Campaigns. ” Communication Research 26(2): 144-175. Zaller, J. R. (1992). The Nature and Origins of Mass Opinion. New York, Cambridge University Press. ———————- [1] This finding is questionable, however, since what they are actually measuring is the extent to which people say they are exposed to oppositional viewpoints in the news, rather than the extent to which they actually are exposed to oppositional viewpoints. A single significant, non-negative coefficient in their analysis of perceived exposure to oppositional viewpoints through newspapers for Republicans/Conservatives (p. 108) suggests that what they are measuring is the belief that news media are biased against their own viewpoints.