Abstract
A framing analysis of South Korea’s three major newspapers and two television networks reveals that coverage of presidential election polls was heavily biased toward the ruling party and its candidate. The Korean mainstream media stories more often focused on the race between candidates than on political issues of the campaign.
Public-opinion polls are one primary part of election coverage in contemporary mass media. They often take a considerable portion of campaign coverage, drawing enormous attention to candidates’ standing in the race and their tactics. 1
It is questionable, however, whether mass media play their expected roles in reporting public-opinion polls, especially during an election period. This doubt is closely related to the ways mass media deal with public polls. For example, political pundits have blamed mass media for using polls to deliver the “horse race” (who is ahead and who is behind) rather than the public’s preferences and interest. 2
During the 2012 presidential election in South Korea (hereafter Korea), civic organizations criticized mass media for biases in reporting public polls. 3 The organizations accused major media outlets of having intervened skillfully in the campaign with problematic reports on public polls. 4
This study approaches the media coverage on public polls by primarily using framing theory. Framing involves selecting a few aspects of a perceived reality and connecting them in a narrative that promotes a particular interpretation. 5
A multitude of studies have demonstrated that mass media tend to describe an election as a strategic game rather than as an opportunity for public debates on public issues. Nonetheless, whether the trend toward strategic game framing of election news is the same in a nation of different political and media systems with their unique characteristics remains an open question. This study examines the framing strategy by mainstream news outlets in Korea, where democracy is still a work in progress.
Literature Review
The present study makes two other contributions to framing literature. It incorporates the valence framing (positive, neutral and negative) into the strategic game framing approach. Thereby, this study attempts to show how media coverage of public polls takes on a political nature. In addition, the present study examines how different media types—television and newspapers—demonstrate different framing strategies. To these ends, the present study analyzes news stories collected from three major newspapers and two major television networks during the 2012 presidential campaign in Korea.
Those who study media framing of campaigns have distinguished between frames that highlight issues and those that accentuate the strategic elements of campaigns. Strategic news is defined as news that (1) centers on winning and losing, (2) favors the language of war, games and competition, (3) focuses on candidate style and perceptions and (4) emphasizes polling results and candidate standings. 6 Strategic framing often dominates the coverage of election campaigns and policy issues. In discussing the notion of strategic framing, Patterson stressed that election news tended to rely on “game schema.” 7 Game refers to strategic and electoral success, emphasizing candidates’ position in the electoral race. Patterson provides evidence of the historical increase in the use of the game or horserace frame in the press coverage of U.S. elections from 1960 until 1992. In a nutshell, the strategic game frame portrays elections as an ongoing contest or horserace between candidates and between parties and often is compared to typical sports coverage.
At the opposite side of the strategic game frame lies the issue frame. Patterson, for example, differentiated policy stories, which frame campaigns “within the context of policy and leadership problems and issues,” 8 from game stories, which frame campaigns “within the context of strategy and electoral success.” 9 An issue frame focuses on the debate regarding stances or positions between different parties or candidates. 10
Some contend that the dominance of the strategic game frame in the media eventually will cause harm to a healthy democracy. 11 Strategic-game framing of politics can displace a substantive discussion of political issues. 12 As a result, voters are left with partial or skewed understanding of the main issues and candidate stances on those issues. 13 An increased use of strategic game framing may cause the electorate to be less engaged in the political process, have negative feelings toward politics and be highly cynical about politicians and the political system. 14 Patterson has argued that what ordinary citizens obtain from mass media is information about the maneuvering by the candidates and the likely campaign outcome. 15 There are even indications that the dramatic growth of political cynicism since the 1960s may be attributable in part to the tenor of campaign coverage. 16 Cappella and Jamieson have argued that the end result of the strategic-game framing of campaigns is a “spiral of cynicism” that seduces candidates into concentrating their efforts on playing the game, trapping reporters into focusing more on the game and alienating the public from the political process. 17
Despite the usefulness of the “strategic game vs. issue” framework, there have been few efforts to consider it together with news slant and bias. Slant occurs when a news report emphasizes one side’s preferred frame in a political conflict while ignoring or denigrating another side. 18 One-sided framing emphasizes some elements and suppresses others in ways that encourage recipients to give weight to the evaluative attributes that aid the favored side’s interpretation. News bias is observed when slant persists and pervades the most influential media outlets. 19 Under this definition, biased content assists such entities as political parties or interest groups in consistently persuading people to accept interpretations that are helpful to the actor favored by some significant segment. When slant persists in this way, biased content helps beneficiaries obtain what they want from the government. In this sense, media bias can help distribute political power.
Although many scholars dismiss allegations of bias in the mainstream media, some scholars reported evidence of a deliberative choice for or against a candidate by the media. 20 This research refines the strategic game framing model by combining it with valence framing, thereby providing a theoretical foundation for systematic studies of news bias.
Most prior studies have not linked frames of poll coverage to news bias. By refining a conceptualization of news bias and empirically analyzing evidence on slanted news, this article attempts to fill the void. This study raises the following issues, such as how much news stories are slanted, what explains the bias of news and how news bias helps to distribute political power. A careful investigation of news bias may serve to illuminate the mass media’s role in the election process. This paper can build new insight into how the media become involved in operations of political power.
More than four decades after the Korean War in the 1950s, the Korean mass media have been robbed of a considerable portion of their autonomy and compelled to play the role of the mouthpiece of authoritarian regimes. For example, television broadcasting has been controlled by the government since its inception in 1961. The Korean government retains the power to appoint management and also exerts control over the financing of broadcast media organizations and their content. During the early 1980s, hundreds of “nonconforming” journalists were forced to leave their jobs and were not allowed to be employed for any media-related work. In other words, the Korean media system has been under heavy control by the government through licensing, appointment or sometimes censorship and has been expected to serve the state.
In Korea, 1987 was a momentous year. 21 Around the nation, the public protested in massive demonstrations for democracy. After the protests, the government’s control of media was somewhat loosened. However, although the changed political environment contributed to the expansion of media freedom, political authorities have still maintained direct and indirect control over broadcasting. The selection of presidents of the two major public stations—KBS and MBC—is proposed by the Ministry of Culture and appointed by the president of Korea; a board of directors is proposed by the president and appointed by the Ministry of Culture; and reviewers are proposed by the president of Korea and appointed by the Ministry of Finance.
Unlike the major TV stations, major newspapers of Korea gradually increased their political autonomy and market power. Despite the massive increase of media outlets after the 1987 protests, only three conservative daily newspapers—Chosun, Joongang and Donga Daily—thrived in the newspaper market mainly thanks to the dominant market power. Today, the three newspapers hold more than 50 percent of the Korean newspaper market. The high concentration of the Korean newspaper market by a few conservative presses can be attributed in many ways to the tactful collusion between major newspapers and authoritarian governments. In order to make it easy to control the press, the authoritarian regimes restricted the number of newspapers. The ruling power guaranteed the press monopolistic profits, such as authorization of a cartel and provision of tax and loan privileges. In return, the press established itself as a mouthpiece of the ruling bloc. 22 After 1987, the major newspapers became a self-declared political institution. 23 In almost every election, they intervened in the process of the campaign directly and indirectly, exerting a considerable influence on the final election results.
In brief, the mass media landscape of Korea can be summarized by three characteristics. First, even though there emerged several liberal media outlets after 1987, the mass media market of Korea has been and still is overwhelmingly dominated by major conservative media. Second, major TV networks are directly and indirectly controlled by the state. And finally, major newspapers voluntarily maintain a close relationship with the conservative ruling bloc.
In such a society, the pattern of news framing would be quite different from that found in advanced democratic societies. Examining the media coverage of an election in a country that has a considerably “unbalanced” media system compared with Western countries will make a contribution to the literature, by debunking a hidden structural link among the political establishment, mass media, and the electorate.
The background of this study is the 2012 presidential election in Korea. As in the previous presidential elections, this election was a contest between two ideological power groups. Park Geun-hye, the daughter of the former dictator Park Jung-hee, was the candidate from the ruling Saenuri Party; Moon Jae-in was the candidate of the opposing Democratic Party. The competition was intense with Park receiving overwhelming support from the conservative sector including major newspapers and TV networks; Moon received support from liberal civic groups including liberal media outlets. Park won with 51.6 percent of the vote.
Hypotheses and Research Questions
As noted above, the tendency to frame politics as a strategic game rather than focusing on the issues is widespread in democracies around the world. 24 This study expected a similar pattern from the newspapers and broadcasts of Korea during the 2012 election. Studies by Korean scholars reported that the mass media in this country tended to rely more on the strategic game frame than the issue frame in political reporting. 25 Since the 1987 presidential election, competition among the various candidates was extremely intense, and the press fanned this competition by releasing information about each candidate’s campaign strategy. 26 For example, the analysis of television news during the 1997 presidential election revealed that about 42 percent of the total political news dealt with the candidates’ campaign tactics. 27
Strategic-game framing will be more common than issue framing in the coverage of public polls by the major newspapers in Korea during the 2012 presidential election.
Strategic-game framing will be more common than issue framing in the coverage of public polls by the state-controlled TV networks in Korea during the 2012 presidential election.
Besides the strategic game vs. issue framework, this study pays attention to the valence framing of polling news. The valence of news coverage varies depending on political positions of each medium and its affiliation with political groups. Major Korean newspapers, which are mostly conservative, have maintained a close relationship with the conservative ruling group for more than four decades. 28 A similar prediction can be drawn from major TV networks whose CEOs are appointed by the government. KBS and MBC have often been blamed for reporting news biased toward the ruling government. 29 Thus, the study expects that the tone of polling coverage by mainstream newspapers will be oriented toward conservative candidates.
Reports of polls by the major Korean newspapers will show a more favorable attitude toward the conservative candidate from the ruling party than toward the liberal candidate from the opposing party.
Reports of polls by the state-controlled Korean TV networks will show a more favorable attitude toward the conservative candidate from the ruling party than toward the liberal candidate from the opposing party.
The present study also examined whether there is a difference in the coverage of polling news between major newspapers and major TV networks. As mentioned earlier, the major conservative Korean newspapers maintain a voluntary connection with the ruling bloc, while major TV networks are forced to run in accord with it because the ruling elites have the power to control the the networks’ governing people. Thus, this study examines which medium takes a more biased stance in the coverage of public polls during the presidential election.
In what ways does the tone of polling coverage differ between the major newspapers and major TV networks in Korea during the 2012 presidential election?
In addition, this study examines the extent to which Korean media provided information about public polls in publishing news stories about them. If the mass media fail to provide necessary and substantial information, readers are likely to be misled because they lack sources to use in their decision-making processes. Prior studies have reported that margin of error, response rates, wording of survey questions, sample sizes, interview methods and fieldwork dates are often not included in polling coverage. 30 Even though a number of studies repeatedly documented the lack of methodological information on public polls in the mass media of the West, such an approach has rarely been adopted in Korea. In addition, this study content-analyzed the number of commentators and election-relevant topics in the news articles during the election.
How many methodological data of polls, commentators and election topics are found from the news stories of the major newspapers and TV networks in Korea during the 2012 presidential election? [See Table 1]
Characteristics of Political and Media Systems of South Korea.
Source: Korea Press Foundation (2012).
Method
News stories for the analysis were obtained during September 19 through Dec. 19, 2012 (election day). This study examined three major newspapers that are national in scope: Chosun, Joongang and Donga. The three newspapers occupy 48.7 percent of the whole Korean newspaper market. 31 [See Table 2] Among the TV networks, KBS and MBC were chosen because they cover the national boundary and are under the control of the government. The two TV networks occupy 53.9 percent of the Korean TV market. 32
Distribution of Issue and Strategic Game Frame Articles.
A news article was included in the analysis if it covers polls about the presidential election and candidates. The analysis included only stories in which polls constituted the main story. From the two broadcasters, only the news stories reported in prime-time news programs were included. A total of 664 articles from the three newspapers, and 107 reports from the two networks were used in the final analysis.
The code sheet included a number of variables identified in previous research. Strategic game frame news was defined as a story focused on the tactics or strategies of political campaigning, horse race, battle for voters, the images of politicians, political power as a goal in and of itself or politicians as persons rather than as spokespersons for certain policies. 33 Issue frame news, on the other hand, was defined as a story focused on issues and issue positions, including real-life conditions and what had happened or what someone had said and done with respect to issues and issue positions. 34
In addition to two major meta-frames, this study coded three sub-frames subsumed in strategic-game meta-frames. Horserace frame news was defined as a story focusing on who was ahead and who was behind; political strategy frame news was defined as a story focusing on how the parties or candidates act during the election and why they do such things; and, finally, politicians as individuals frame news was defined as a news story focusing on politicians as individuals, with different attributes, characters and behaviors rather than as spokespersons for certain policies or ideologies. 35
The following set of methodological information was coded, based on the recommendations issued by the AAPOR 36 and the NCPP 37 : (1) identity of the conductor, (2) definition of population, (3) sample size, (4) interview method (phone, in-person, etc.), (5) interview dates, (6) sampling error, (7) response rate and (8) question wording. All of them were turned into dichotomous variables representing whether the information is present. A “total amount of methodological information’” variable was constructed by combining the eight items.
Valence of framing was coded according to four categories: positive, negative, neutral or mixed and irrelevant. To facilitate a more efficient analysis, the two latter categories were combined into one. To test for inter-coder reliability, seven percent of the news stories were randomly selected from the whole stories. The inter-coder reliability (Holsti) across the variables was from .89 to .96.
Findings
H1 predicted that strategic game framing would dominate over issue framing in the three newspapers. As expected, stories with the strategic game frame (573, 86.3 percent) overwhelmingly outnumbered those with the issue frame (26, 3.9 percent), x2(2, N = 599) = 121.18, p < .001. Thus, H1 was supported. As H2 expected, stories of the strategic game frame (71, 66.4 percent) in the two TV networks exceeded those of the issue frame (26, 24.3 percent), x2(2, N = 97) = 31.33, p < .001.
For H3 and H4, the study calculated the favorability score toward each candidate by subtracting the number of negative news stories from the number of positive stories. In the three newspapers the total number of articles that favored Park was 322, while the total number of articles that favored Moon was 59. [See Table 3] The same pattern was found from the two TV networks. The number of articles that favored Park was 43, while the number of articles that favored Moon was 9. Thus, H3 and H4 received support.
Valence of Framing in Polling News.
Note: The favorability score toward Park was calculated according to the following formula: [{(positive stories toward Park – negative stories toward Park)} + {(negative stories toward Moon – positive stories toward Moon)}]/the total number of news stories.
To answer RQ1, a favorability score toward Park was calculated according to the following formula: [{(positive stories toward Park – negative stories toward Park)} + {(negative stories toward Moon – positive stories toward Moon)}]/the total number of news stories. In the three newspapers, the favorability score toward Park was 43.8 [{(322 – 38)} + {(66 – 59)}]/664. In the two TV networks, the favorability score toward Park was 17.8 [{(43 – 21)} + {(6 – 9)}]/664. In other words, the three newspapers were favorable toward Park more than twice as many times as were the two TV networks.
In terms of sub-frames of the strategic game meta-frame, all three newspapers focused mostly on horserace reporting and candidates-as-individuals reporting as shown in Table 2. Some 257 (44.9 percent) of the total strategy meta-frame articles were found to include the sub-frame of candidates-as-individuals. Some 256 (44.7 percent) of the articles contained the horserace sub-frame. The political-strategy sub-frame was found in 10.5 percent of the strategic game frame stories. A similar pattern was found in the stories from the TV networks. The horserace sub-frame was found in 49.3 percent of the strategy of meta-frame articles. In addition, 33 articles (46.5 percent) included the candidates-as-individuals sub-frame.
The analyses revealed that the three newspapers were parsimonious in providing basic information about polls, which is deemed as necessary for readers to understand the significance of the polls. None of the newspapers included survey questions in their coverage. The response rate was only reported in 15.0 percent of the whole articles. Other pieces of necessary information were also missing in the four newspapers’ coverage. [See Table 4] The TV stations included more methodological information than did the newspapers (average information per article = 5.65). Except for the information about response rates and question wordings, the two networks provided enough information about poll conductors, population definitions, sample sizes, interview methods, interview dates and sampling errors.
Proportions of the Inclusion of Methodological Information.
Note: Information per article is the sum of the eight methodological information variables.
Discussion
Application of the strategic-game vs. issue framing tool to public-opinion polls during an election has been rare, especially in a society like Korea where democracy is still a work in progress. Also, little research took into account both strategic framing and valence framing in analyzing news frames. Under the presumption that the preponderance of “manipulative” framing of poll news can prevent voters from engaging in rational judgment on candidates and election issues, this study examined the coverage on public polls by mass media during the 2012 presidential election in Korea. It found several significant results, which can contribute to the literature of news framing.
First, Korean mainstream media depended excessively on the strategic game frame rather than the issue frame. This means that Korean mass media centered on portraying the presidential election as an image of battle between candidates or political parties, rather than encouraging voters to engage in constructive dialogue about important issues regarding candidates and parties. The over-dependency on the strategic game frame was observed in both major newspapers and state-controlled TV networks.
Second, Korean newspapers were reluctant to provide essential pieces of information about public polls, which are important for voters to assess the results and implications of the polls. Surprisingly, no newspapers or broadcasters included survey question wordings in their news coverage. Also, the response rate, sampling errors, interview methods and sample sizes were missing in most newspapers. Thus, Korean mass media failed to serve as a provider of necessary information to the electorate during the election.
Third, the Korean media revealed a partisan attitude in their coverage on public polls. The three conservative newspapers unilaterally favored Park, the ruling party candidate. Likewise, the two government-controlled TV networks showed a heavily biased attitude toward the ruling party and its candidate, paying less attention to the opposing party and its candidate. This finding suggests that the mainstream media failed to capture and mirror an accurate landscape of Korean public opinion. Studies show that more than half of Koreans can be considered politically middle-of-the-road. According to one study, 51.7 percent of Koreans identify themselves as moderates. 38 Conservative identifiers are 21.3 percent, while liberalism-oriented people are 27.0 percent.
Last and most important, newspapers showed a more partisan attitude toward the ruling bloc than did the state-controlled TV networks. This finding implies that, contrary to some scholars’ contentions, 39 biased news framing can arise more from a medium’s voluntary association with political power than from a legal or administrative intervention by the state. The study demonstrates that major newspapers in Korea are overtly biased and closely linked to specific political parties to the extent that they are paralleling the political system. 40 This phenomenon is quite different from what we see in most western countries where the press is hardly paralleling political parties. 41 As stated earlier, major Korean newspapers have maintained their market power regardless of the nature of the ruling bloc. Under the authoritarian regimes, they survived through a forced cooperation with dictatorial regimes. As democratic processes spread after the massive protests in 1987, the major newspapers voluntarily started forming a symbiotic relationship with the ruling bloc for their political and economic purposes. On the other hand, despite a drastic change in the political landscape, the Korean broadcasters have maintained a forced relationship with the political establishment because of their state-controlled governing system.
As expected, stories with the strategic game frame (573, 86.3 percent) overwhelmingly outnumbered those with the issue frame (26, 3.9 percent).
Whether public polls play their expected role in the democratic process is, to a considerable level, dependent upon the degree that poll results are communicated correctly and fairly to the public. Gallup and Rae 42 viewed polls as allowing election candidates to respond to the concerns of a majority of voters, ushering in greater democratic representation. The findings of the current study suggest that the way of covering polls by Korean mass media is depriving the public of its grip on the democratic process in two intertwined ways. One of them is the negligence of professionalism. The Korean media eroded professional conduct by omitting essential information about polls, such as response rates, population definitions and sample sizes. The coverage of public polls should start from transmitting undistorted information to the public, which is a key element in professional journalism.
Another way the Korean mass media isolated citizens from the election process is the over-dependency on partisanship. Polling enables the press to intervene in politics directly. 43 The media actively pursued their own political or ideological positions, departing unduly from news-gathering and reporting norms. 44 Korean mass media tended to misrepresent public polls as they wished, contributing to a serious misapprehension of the nature of public opinion and its ideal relationship to representative democratic processes. 45 Furthermore, the results of this research suggest that the media are often deeply allied with political groups. The findings confirm that “the mass media are often the unwitting handmaidens of the powerful.” 46
The political-disengagement effect by the mass media in polling-news coverage can be a more serious threat to a society in which the media system is ideologically and systematically lopsided than to a society having a balanced media system. Even though a number of alternative or liberal media appeared after the 1987 democratization, their scale and power are relatively small compared with those of the major conservative newspapers and government-controlled broadcasts that have dominated the public opinion market of Korea for decades. In such a situation, opposing voices are easily ignored or sometimes twisted by mainstream media that function as “managers of the status quo.” 47 Such a practice, in turn, distorts the actual nature of public opinion.
One thing that democracy requires from the mass media is to encourage citizens to engage in free and unfettered debates on public issues. 48 However, based on the findings, this study speculates that the Korean media coverage on public polls during the 2012 election rarely facilitated citizen engagement in vibrant discussion on election affairs or helped the Korean society to reach a consensus-based decision. Rather, the Korean mass media served as a conduit for their own political positions resonating with their respective political power group. If the information people receive from the mass media is tainted quantitatively, qualitatively and politically, they may not perform expected duties as democratic citizens.