HOW TO THINK ABOUT POLITICAL CANDIDATES: A TEXAS PRIMARY ELECTION STUDY

Transcription

1 Previous Paper Next Paper Contents HOW TO THINK ABOUT POLITICAL CANDIDATES: A TEXAS PRIMARY ELECTION STUDY SUNGTAE HA University of Texas at Austin (U.S.A.) Abstract This study explores the influence of news media on the voters attribute agendas (cognitive attribute agenda-setting), preferences (affective attribute agenda-setting), and criteria of candidate evaluation (attribute priming) in the 2002 Texas primary elections for governor. In the cognitive regard of media effects, the news media successfully set the voters attribute agenda for one of the three candidates. The tone of the media coverage of the campaigns also significantly influenced the voters affective attitudes toward the candidate for whom cognitive agenda-setting effects occurred. As an attribute priming effect, the voters further employed the salient attributes about the candidate s qualifications and character in the news media as the standards for candidate choice. For the other two candidates, however, the media was not successful in influencing the voters political perception and judgment. Political awareness appeared to be a precondition in which political attitudes are build up, and political attitudes again was a preliminary step for political behavior to occur. Strong attribute agendasetting and priming effects of news media occurred in a particular circumstance of high need for orientation and agenda accessibility.

2 HOW TO THINK ABOUT POLITICAL CANDIDATES: A TEXAS PRIMARY ELECTION STUDY Introduction For more than a century, the press has been a necessary condition for the operation of the political systems in our democratic society. The important role of news media lies in the fact that the press is a common carrier of the messages of political leaders. This is why the press has been at the center of modern political communication interacting with policymakers and the public (for overview, see Perloff, 1998). News media s role of informing the public about politics, however, has further implications. Numerous studies in the field of political communication have shown that the public relies on news media not only for the information about political agendas but for the ways of thinking about political issues and actors as well (see McCombs, M. E., Shaw, D. L., & Weaver, D., 1997; Scheufele, D. A., 2000). No matter what formats it takes, the mediated news has become the biggest and substantially the only accessible window of information about politics in modern democratic societies. News media provide the materials with which the electorate draws the picture of the political world outside, and frequently even the directions about how to draw it (see Lippmann, 1920). In short, without news media, both the voters and political actors cannot even imagine how to participate in and maintain the democratic process. In general, it is true that the higher the level of an election campaign, the greater amount of news media coverage and scrutiny on the candidates (Dunn, 1995). The role of news media, however, might be more instructive in local elections than the presidential elections. While in presidential elections news media allot excessive resources for the campaign coverage, candidates also devote astronomical figures of money to political advertising and the voters become more involved in political conversations with their friends, family, and colleagues, diluting the media effects. In local-level elections, voters relatively know so little about their candidates, and thus news about the candidates could play a crucial role of directing the evolution of campaign agendas in the public s minds (Geer & Kahn, 1993). In local settings, the voters may have loose schemata about their candidates, which should be tightly [re]constructed with new information about them to reach voting decisions. Because voters have less prior knowledge about the candidates running for state or local office and subsequent weaker preferences toward them, the supplies from news media could be an important trigger for voters political attention, preference, and even behavior. The 2002 Texas gubernatorial election is an excellent case to study the news media s role of informing voters and directing their decision-making: both the Republican and Democrat candidates are the first-time runners for the job of governor. This study, in the setting of the 2002 Texas primary election for governor, explores the influence of news coverage of the primary campaigns by The Austin American-Statesmen, the only major Austin area daily, on the voters attribute campaign agenda and their criteria of candidate selection. How is the salience of attribute agendas about the political candidates in the newspaper related to the importance of those attributes in the voters mind? And how does the emphasis of those attributes in the news media direct the voters decision-making? 330

3 To answer these questions, this study bases its analysis on two theoretical concepts of attribute agenda-setting and attribute priming. This paper analyzes those media effects on three different stages of communication effects the cognitive, affective, and conative dimensions to examine the full process of electoral communication. Three different dimensions of mass media effects Mass media effects are a common interest to the creators, mediators, and receivers of the messages because they are the direct participants in the communication process interacting with each other. During the last century, however, there have been controversies over the size of mass communication effects and researchers are still exploring the power of mass media under the newly evolving communication environment in the age of digital communication. From the bullet theory, through the limited-effects and moderate effects models, to the revived powerful-effects model (for overview, see Severin & Tankard, 2001), researchers have seen the same nature of communication effects from different perspectives. In a sense, all the proponents of the different effects models may be correct on their argued amounts of media effects although they tended to focus on only certain portions of the whole course of information processing by audience. To grasp the comprehensive effectiveness of mass communication, we need to look at all the dimensions of audience s information processing. Lavidge and Steiner (1961) organized six hierarchical steps of possible mass media effects to explain advertising effectiveness: Awareness, knowledge, liking, preference, conviction, and purchase. In this stair-step model, each step is a precondition for the next step to occur. They argue, however, that in some cases consumers move up from awareness to the final step of purchase almost simultaneously and, in other instances, they linger on some steps while passing through other steps in a moment. They also collapsed those six steps into three dimensions of mass communication effects: cognitive, affective, and conative. The cognitive dimension of the first two steps of awareness and knowledge is an intellectual or rational state in which audience is aware of information, ideas, and facts about an object. The affective dimension, composed of liking and preference, is an emotional or feeling state, which concerns audience s attitudes or preferences toward the object the messages are dealing with. The conative component, which includes the final two steps of conviction and purchase, is a state of decision-making to treat the object as a final goal. If applied to the context of electoral campaign communications, this behavioral state will include the voting decision. Attribute agenda-setting and attribute priming in political communication Although they developed the stair-step model to measure advertising effectiveness, virtually it can be a useful framework to explain the general mass communication effects. Particularly, this model fits to the field of electoral campaign communication in that political candidates begin with informing the public of their campaign agendas, try to change voters attitudes toward them, and eventually aim at attracting people s votes. The three different dimensions of media effects also can be applied to different theories in political communication depending on the steps they are dealing with. Traditional agenda-setting theory concerns only the cognitive dimension of communication effects. Attribute agenda-setting covers both 331

4 the cognitive and affective dimensions, and priming theory deals with the conative or behavioral dimension of mass media effects. Cognitive stage of media effects Agenda-setting is the realm of cognition, which is about the transmission of issue salience, not the determination of opinions pro and con about issues, persons, or objects (see McCombs, 1997). At the cognitive stage of political communication, news media provide voters with the information and facts about the political candidates and their campaigns. The cognitive effect of agenda-setting is the influence of the pattern of coverage and play of issues on the news agenda on the salience of issues on the public agenda. In short, the more prominently issues are addressed in the news media, the more prominent those issues become in the publicʹs mind. Affective stage of media effects Attribute or second-level agenda-setting, however, deals with both the cognitive and affective states. While traditional agenda-setting examines how media coverage influences what the public thinks about, attribute agenda-setting concerns how the public thinks about it (see Ghanem, 1997). An object is constructed with numerous attributes. News coverage defines the object by arranging its attributes of different salience, and when people think about the object, such as public issues or political candidates, they will respond to the news messages of the attributes of different salience. Consequently, attribute agenda-setting assumes that the attributes of an object emphasized in news will affect how the way the public thinks about that object. Second-level agenda-setting studies typically deals with the attributes of the cognitive and emotional dimensions of mass communication. In the setting of the 1995 Spanish elections, McCombs et al. (1997) compared voters descriptions of the candidate images and media coverage of the campaigns at the two dimensions of attributes: substantive and affective. They developed three substantive attributes of ideology/issue positions, qualifications/ experience, and personality, and three affective attributes of positive, negative, and neutral to examine the voters political information about and their attitudes toward the parliamentary and mayoral candidates. In general, newspapers exerted stronger agendasetting effects than TV news with more significance on the affective attributes. The correlation was.57 between voters substantive attributes for the parliamentary candidates and one of the two local newspapers, and.41 between the substantive attribute agendas of voters for the mayoral candidates and TV news. For the affective dimension, there were also strong correspondences of.66 and.88 between two local newspapers and voters feeling for the parliamentary candidates, and.59 for TV news and.44 for one local newspaper concerning the mayoral candidates. In another 1996 Spanish election study, McCombs, Lopez-Escobar, and Llamas (2000) also found the median correlation of.72 between the cognitive and affective attributes of 7 different mass media and voters for three candidates. This time, they used five different substantive attributes: ideology/issue positions, biographical details, perceived qualifications, integrity, and personality and image. On the other hand, discussion of the second-level agenda-setting entails the examination of another social science concept, framing, which also helps clarify the cognitive and affective effects of news media on the public s perception. Although a frame has been defined in many different ways according to scholarsʹ research purposes, media frames are commonly regarded as an important component of news coverage because of their influence on or interaction with the publicʹs perception. Some definitions of frames, like Entman s (1993), 332

5 Gitlin s (1980), and Tankard et al. s (1991), are very similar with the definition of attribute agenda-setting as both concepts emphasize selection and salience of frames or attributes as the essential elements of the mass communication effects. Actually, some studies have used the term of framing or frames to explain the similar phenomena of mass communication to what the concept of attribute agenda-setting can account for (Miller, Andsager, & Riechert, 1998; Jasperson et al., 1998; Pride, R. A., 1995; Iyengar & Simon, 1993). Thus, the two concepts are considered converging as The attributes of an objects are the set of perspectives or frames that journalists and the public employ to think about each object (Ghanem, 1997, p. 5). Some researchers, however, argue that framing and attribute agenda-setting are distinct concepts. Scheufele (1999) suggests that framing as a theory of mass communication effects should be conceptualized as a process model in which media frames are working routines both the journalists and audience use to identify and classify information. More specifically, Kim, Scheufele, and Shanahan (2002) explain that while attribute agenda-setting concerns the salience of specific attributes or perspectives of an object through their frequencies in media texts, framing deals with the emphasis of those attributes or frames by the terminological or semantic way they are presented in news media. They describe that agenda-setting is a matter of accessibility that Political issues that are most salient or accessible in a person s memory will most strongly influence perceptions of political actors and figures, whereas framing is about applicability that only if the interpretive cues provided by news media correspond with or activate pre-existing cognitive schema will there be a framing effect noticeable in terms of attitudes or subsequent judgments (p. 9). The fact, however, is that many framing studies have focused on the accessibility of particular issue attributes while only some on the applicability of certain frames. Particularly, it is noticeable that most framing studies using quantitative approaches hold the agendasetting perspective in common because it is hardly possible to measure the issue applicability using quantitative scales. Conative stage of media effects Priming, as an extension of agenda-setting (see Semetko et al., 1991; Iyengar & Kinder, 1987), is about the conative stage of mass communication effects, which deals with our behavior or desires toward an object (Lavidge & Steiner, 1961). As news media function as an agenda-setter raising the importance of particular issues among the public, by setting the agenda for an election campaign, they also determine the criteria by which political candidates are evaluated. Put another way, priming is the final threshold through the whole process of political communication in which voters try to treat political actors as their motivational goals. Iyengar, Peters, and Kinder (1982) introduced the interesting concept of priming that is the process in which news media call attention to some issues while ignoring others and thereby influences the standards by which the public judges political figures and issues. They found that the correlations between the overall ratings of the President Carter s general performance and the specific ratings of his performance on one of the three problem areas pollution, inflation, defense were much stronger for the subjects who saw the TV news programs emphasizing that issue (.63.88) than those exposed to the news neglecting it (.39 53). 333

6 Subsequent priming studies has followed this inceptive study about the news media effects on the public s political judgment, refining the priming research body under the various natural and experimental political circumstances, including the Gulf War and the Iran- Contra Connection (Iyengar & Kinder, 1987; Krosnick & Kinder, 1990; Sherman, Mackie, & Driscoll, 1990; Krosnick & Brannon, 1993; Iyengar & Simon, 1993; Mandelsohn, 1996; Willnat & Zhu, 1996; Pan & Kosicki, 1997). These previous studies, however, like the first-level agenda-setting studies, mainly focused on the priming effects at the issue level. The traditional priming studies deal with the influence of salient issues in news coverage on the weight assigned to those specific issues in voters political judgments. Testing priming effects at an attribute level may be a next, natural attempt. Recently, a priming study at the level of attribute agenda has been done in the context of a local development issue (Kim, Scheufele, & Shanahan, 2002). The researchers found that the most salient attributes about the issue in the local newspaper, which were the possible consequences of developing a local park into a commercial center, appeared to be the significant regressors to explain the issue evaluation (pro and con) among the heavy newspaper readers. Although this study examines a non-political issue, the results clearly showed that news media attend to specific attributes of an issue and thereby influence the standards of the issue evaluation of the audience. Applying the conative dimension of priming effects to electoral communication, we can expect that candidate attributes salient in news media will affect the weight ascribed to those attributes when voters decide for whom to vote. Attribute priming, as an extension of attribute agenda-setting, will explain the underlying psychological process of political communication, through which voters reach their voting decisions. Hypothesis Media effects are a series of cognitive, affective, and conative reactions to the media messages by human beings. Some messages just provide information, some change attitudes, and other direct behavioral desires about an object (Lavidge & Stein, 1961). Similarly, electoral communication is a process in which the voters attend to news media, update their attitudes, and finally reach their voting decisions about their political candidates. Through the brief literature review about political communication effects, this study expects that news media play an important role of setting the public agenda and priming its voting behavior. Here, three hypotheses related to media s agenda-setting and priming functions at the three different dimensions of communication effects cognitive, affective, and conative are tested. H1: The cognitive candidate attributes emphasized in the news media also will be salient in the voters cognitive attribute agendas about those candidates (cognitive attribute agenda-setting). H2: The affective candidate attributes emphasized in the news media also will be salient in the voters affective attribute agendas about those candidates (affective attribute agenda-setting) 334

7 H3: The candidate attribute agendas emphasized in the news media also will be the prominent criteria the voters employ when deciding for whom to vote (attribute priming). Data and analysis This study compared the candidate attributes in media coverage and among the voters in the 2002 Texas primary elections for governor. One methodological uniqueness of agendasetting research is that it typically combines content analysis of the media agenda with survey of the public agenda (Rogers, Dearing, & Bregman, 1993). This specific combination of methodologies provides an excellent perspective in understanding and defining mass media s role in a democratic society. More implications on the role of news media in our society follow in the discussion section of this paper. Media agenda This study analyzed news coverage of the primary campaigns in the single dominant local newspaper in Austin area, The Austin American-Statesmen, because the print media carry much more information than local television news programs about state-level campaigns, and voters generally get more in-depth information about their candidates from newspapers rather than television news (Goldenberg & Traugott, 1987; Westlye, 1991). All the stories about the primary campaigns, including news articles, columns, editorials, and other feature stories, that appeared in the Statesmen from Jan. 1 through Mar. 12, 2002, the election day, were content-analyzed for cognitive and affective attributes. A total of 67 items were found through the manual scanning of the Statesmen during the period in which the news media coverage of the primary elections had peaked. Coders identified a total of 1198 attributes about the two competing Democrats (Dan Morales, Tony Sanchez) and the single Republican (Rick Perry) candidates out of the 67 stories with 6 cognitive categories general political descriptions, specific issue positions, personal qualifications/character, biological information, campaign conduct, and support/endorsements. The candidate attributes were defined as any descriptions or assertions about the candidates if they can be a potential answer to the question of what you would tell your friend about the candidates. The coders further classified the attributes of personal qualifications/character, campaign conduct, and support/endorsements for their tone of positive, negative, and neutral as affective dimensions. From the affective viewpoint, the other attributes were coded as neutral ones. The assertions, which belong to candidates personal qualifications/character were again categorized as leadership, experience/competence, credibility/morality, caring people, and communication skills as the criteria for candidate evaluation. Public agenda Telephone surveys were conducted with 271 randomly selected adults, 18 years or older, in Austin, Texas from Feb. 19 to Mar. 12, Random selection involved a three-part procedure (Poindexter & McCombs, 2000). First, phone numbers were selected using systematic random sampling from the most recent metropolitan phone directory; second, a one was added to the last digit of the telephone number to ensure reaching new residents and unlisted numbers; finally, the gender of the respondent household head was randomly determined. Interviews were conducted by graduate and undergraduate communication students who were trained in telephone survey interviewing techniques and ethical research standards. 335

8 Two crucial questions were included in the survey questionnaire to detect attribute agenda setting and attribute priming effects of the news coverage respectively. For the voters attribute agendas about the candidates, the respondents were asked the question, Suppose that one of your friends has been away a long time and knows nothing about the candidates for Governor of Texas. What would you tell your friend about candidate X? To find out the voters attribute agendas that were employed to evaluate the candidates, surveyors also asked the respondents, Suppose that you vote in the upcoming March primary election for Texas Governor. What candidate qualification or personal character would you personally consider most important when deciding for whom to vote? More specifically, the answers to this qualifications/character question by the voters, who intended to vote for a certain candidate, were compared with the media attributes describing the personal qualification and character of that candidate. The survey coding of voters answers to the two questions basically used and followed the same categories and process that were applied to the media attribute agenda analysis. Coder reliability Eight journalism graduate students in a content analysis class analyzed the media attribute agendas and achieved the average agreement percentage of around 90. Findings The first hypothesis deals with the cognitive attribute agenda-setting effects. The Austin American-Statesmen treated the three candidates fairly enough, at least, in terms of the proportion of cognitive attribute agendas. The distribution of the cognitive attribute agendas for the three candidates in the news media shows similar patterns (Table 1). The calculation of correlation coefficients reveals substantial to strong relationships among them: the average correlation of.69. Voters attributes across the three candidates, however, have a somewhat different picture. Overall, the voters attribute agendas for each of the three candidates were not significantly matched up with each other. The mean correlation among the voters attribute agendas for the candidates was.27. The most frequently mentioned attributes in the Statesmen were specific issue positions, personal qualifications/character, and biological information. More than 71 percent out of 1198 assertions in the Statesmen go under the headings of these three attributes. Campaign conduct, General political descriptions, and support/endorsements follow them. One noticeable result is that the Statesmen paid less attention to the Republican candidate, Rick Perry. According to the number of assertions, he received only half as much coverage as the other Democrat candidates. Considering that he is the incumbent governor and has no competitors in the Republican primary election, however, this result is not a surprise. 336

9 Table 1. Cognitive Attribute Agendas about Candidates for Texas Governor in the News Media and Voters Candidate Morales Sanchez Perry Attributes ASS Voters ASS Voters ASS Voters General political descriptions Specific issue positions Personal qualifications and character Biological information Campaign conduct Support and endorsements Total 100% 100% 100% 100% 100% 100% Total number of attributes Note. The percentage figures are based on the total number of the cognitive attributes about each candidate in 67 news items from The Austin American-Statesmen (ASS) from Jan. 1 to Mar. 12, 2002, and the number of voters answers to the question about candidate descriptions. Table 2. Correlations of Cognitive Attribute Agendas about the Candidates for Texas Governor between the News Media and Voters Candidate Morales Sanchez Perry *-.09 Note. The Spearman correlation coefficients (two-tailed) are based on the rank orders among the six cognitive candidates attributes between the campaign coverage in The Austin American-Statesmen and the voters answers to the question about candidate descriptions. * p<.005. Calculations of the relationship between the cognitive attribute agendas for the candidates in the news media and voters reveal a mixed attribute agenda-setting effects at the cognitive stage (Table 2). While the Statesmen s coverage and voters descriptions about Sanchez, a Democrat candidate, almost perfectly matched up with each other (.94), the two agendas showed no relationship for Morales, the other Democrat candidate, and Perry, the single Republican candidate. 337

10 Table 3. Affective Attribute Agendas about the Candidates for Texas Governor in the News Media and Voters Candidate Morales Sanchez Perry Attributes AAS Voters ASS Voters ASS Voters Positive Negative Neutral Total 100% 100% 100% 100% 100% 100% Total number of attributes Note. The percentage figures are based on the total number of the affective attributes about each candidate in 67 news items from The Austin American-Statesmen (ASS) from Jan. 1 to Mar. 12, 2002, and the number of voters answers to the question about candidate descriptions. In general, on the other hand, the tone of the Statesmen was much more neutral compared with the public s voice (Table 3). The neutral tone of the media coverage accounts for about 48 percent of its 1208 attributes, while the positive and negative attributes explain each 21 and 31 percent of the total assertions. Voters were more positive about their candidate than negative and neutral: 42, 33, and 25 percents accounted for the positive, negative, and neutral voices of the public respectively. Table 4. Correlations of the Affective Attribute Agendas about the Candidates for Texas Governor between the News Media and Voters Candidate Morales Sanchez Perry -1.00* 1.00* -1.00* Note. The Spearman correlation coefficients (two-tailed) are based on the rank orders among the six affective candidates attributes between the campaign coverage in The Austin American-Statesmen and the voters answers to the question about candidate descriptions. * p<.01. The second hypothesis of this study expected that the tone of media coverage of the campaigns would influence voters emotional attitudes toward their candidates. Interestingly, the attribute agenda-setting effects at the affective stage illustrate a replica of the findings about the cognitive attribute agenda-setting (Table 4). Again, for Sanchez, the affective attributes of the voters were perfectly correlated with those in the Statesmen (1.0). The voters affective attributes, however, had reverse relationships with the tone of media representation for the other candidates. 338

11 Table 5. Media s Candidate Attribute Agendas about personal qualifications and Character and Voters Criteria for Candidate Evaluation Candidate Morales Sanchez Perry Personal qualifications and character AAS Voters ASS Voters ASS Voters (Attribute) Leadership Experience/Competence Credibility/Morality Caring People Communication Skills Total 100% 100% 100% 100% 100% 100% Total number of attributes Note. The percentage figures are based on the total number of the candidate attributes about personal qualifications and character in 67 news items from The Austin American-Statesmen (ASS) from Jan. 1 to Mar. 12, 2002, and the number of voters answers to the question about the criteria for candidate evaluation. To test the conative communication effects of attribute priming, this study compared the media s attribute agendas subsumed to the heading of candidate s personal qualifications/ character and the voters answers to the question asking about their criteria of candidate choice, which were also related to the same category (Table 5). This comparison makes it possible to map the underlying psychological process about how the news coverage of the campaigns influences the electorate s standards for candidate evaluation. In this primary election, the public regarded the attribute of credibility/morality as the most important criterion when they judge their candidates: about 64 percent of the total assertions of 61 were accounted for this attribute. The media, however, dealt with the attribute of experience/competence as the most salient one in its coverage of the candidates: the attribute took up about 55 percent of the whole assertions of 183. Table 6. Correlations of Media s Candidate Attribute Agendas about Personal Qualifications and Character with Voters Criteria for Candidate Evaluation Candidate Morales Sanchez Perry *.720 Note. The Spearman correlation coefficients (two-tailed) are based on the rank orders among the five candidate attributes about personal qualifications and characters between the campaign coverage in The Austin American- Statesmen and the voters answers to the question about the criteria for candidate evaluation. * p<.01. The correlation analysis showed an extended version of the attribute agenda-setting effects (Table 6). The assertions made about the candidates qualifications and character in the 339

12 Statesmen and the voters response to the open-ended question about their candidate evaluation criteria were significantly correlated with greatly high coefficient of.97 for Sanchez. For the other two candidates, the correlation coefficients were.72 but with no enough statistical significance. Through the correlation tests at the three different levels of communication effects, this study found not fully satisfactory but a consistent pattern of media effects and an impressive match-up of the media and the public attribute agendas for one of the three candidates. These results suggest that each of the three stages of communication effects is a precondition for the next step to occur. A person s behavioral desire would not occur without attitudes toward a thing and those attitudes would not be molded without any awareness about it. Consequently, when the Statesmen failed to set the voters cognitive attribute agendas for Morales and Perry, it subsequently failed to set their attitudes about them and influence their political judgment. For Sanchez, however, the newspaper told the public not only what to think about and how to think about it but what to do about it as well. Discussion This study examined the role of news media in the 2002 Texas primary election at the three different levels of mass media effects: cognitive, affective, and conative. At the cognitive level, the news media successfully set the voters attribute agenda for one of the three candidates for governor. The tone of the media coverage of the campaigns also significantly influenced the voters affective attitudes toward the candidate for whom cognitive agendasetting effects occurred. As an attribute priming effect, the voters further employed the salient attributes about the candidate s qualifications and character in the news media as their criteria for candidate evaluation when making voting decisions. For the other two candidates, however, the media was not successful in setting the voters attribute agenda and priming their standards for candidate choice. A couple of potential explanations can be suggested to justify these half-satisfactory but consistent findings. First, for attribute agenda setting and priming, there are certain cases or situations where those media effects would be evident and most pronounced. For example, the impact of media coverage and candidate ads for an unknown candidate, like Sanchez, could greatly contribute to the public s knowledge and understanding of such a relative political novice. Whereas in the cases of Morales and Perry, we find two candidates with a substantial history in office and on the campaign trail, as well as the media s own familiarity with both of these candidates. Second, the local newspaper was not fair in terms of the amount of news coverage. Perry who was the incumbent Governor and single candidate was the no-doubt electee in the Republican primary election. News media as well as voters, thus, might have lost their interest in the Republican primary election. According to the number of assertions, Perry received only about half of the news coverage that the Democratic candidates did. Accessibility to information is an important condition in which news media can function as an effective agenda setter. Low accessibility of information in news media and low political interest of the voters are the reasonable conditions in which the voters would not expose 340

13 themselves to the even fewer news messages, subsequently implying weaker agenda-setting and priming effects. Third, It seems to be natural that, in a contested primary, voters are more interested in the electable candidates than losing candidates, thus seeking to know more about the winning candidates. Jamieson (2000) found that the electorate s agenda matched more closely with the winning candidate s agenda than the losing candidate s agenda in 1996 presidential election. In fact, various media polls, which were carried out one month before the March 12 primary election, showed Sanchez surged ahead Morales by 8 to 10 percentage points (The Austin Chronicle, Feb. 15, 2002). Consequently, news media cannot exert strong influences across all the voters of different backgrounds in different electoral circumstances. Instead, some people may be more susceptible to news coverage, and others more resistant to the information supplies from news media. The notion of need for orientation provides some of the psychological explanation for individuals different susceptibility to media effects. Need for orientation deals with such motivational contingents as interest, relevance, and uncertainty about political issues (Weaver, 1977). Individuals with more interest and less certainty in particular issues will pay more attention to media messages to get the related information, becoming more susceptible to media effects. What do these effects say about the contribution of mass communication to democracy? Numerous agenda-setting and priming studies suggest that media effects can occur in any democratic society in which its political and media systems determine how the society is operated. In other words, mass communication effects cannot occur when the citizens do not believe in the media s independent role in the electoral process. The voters are more likely to rely on the independent news media as sources of news and political expression. In this sense, news media s influence on the voters is not necessarily undesirable. Democracy is a political system, which is operated by the dynamics and complex interplay of influences among politicians, the media and the public. Public opinion does not come from thin air, but is a product of the process of political communication among these three components of a democratic society. In this process, the news media should play a constructive role of bringing candidates and voters together as well as the traditional role of a watchdog. References Dunn, A. (1995). The Best Campaign Wins: Local Press Coverage of Nonpresidential Races in Campaigns and Elections American Style. eds. J. A. Thurber & C. J. Nelson. Boulder, Colorado: Westview Press. Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43 (3), Geer, J. G. & Kahn, K. F. (1993). Grabbing attention: An experimental investigation of headlines during campaigns. Political Communication, 10, Ghanem, S. (1997). Filling in the tapestry: The second level of agenda setting. Communication and democracy: Exploring the intellectual frontiers in agenda-setting theory, Eds. M. E. McCombs, D. L. Shaw, & D. Weaver. Mahwah, NJ: Lawrence Erlbaum Associates. 341

14 Gitlin, T. (1980). The whole world is watching: Mass media in the making and unmaking of the new left. Berkeley, CA: University of California Press. Goldenberg, E. & Traugott, M. (1987). Campaigning for congress. Washington, DC: Congressional Quarterly Press. Iyengar, S. & Kinder, D. R. (1987). News that matters: Television and American opinion. Chicago, IL: The University of Chicago Press. Iyengar, S., Peters, M. D., & Kinder, D. R. (1982). Experimental demonstrations of the notso-minimal consequences of television news programs. The American Political Science Review, 76 (4), Iyengar, S. & Simon, A. (1993). News coverage of the Gulf crisis and public opinion: A study of the Gulf crisis and public opinion. Communication Research, 20, Jamieson, K. J. (2000). Everything you think you know about politics...and why youʹre wrong. New York, NY: Basic Books. Jasperson, A. E., Shah, D. V., Watts, M., Faber, R. J., & Fan, D. P. (1998). Framing and the public agenda: Media effects on the importance of the federal budget deficit. Political Communication, 15, Krosnick, J. A. & Kinder, D. R. (1990). Altering the foundations of support for the president through priming. American Political Science Review, 84 (2), Krosnick, J. A. & Brannon, L. A. (1993). The impact of war on the ingredients of presidential evaluations: Multidimensional effects on political involvement. American Political Science Review, 87 (, Lavidge, R. & Steiner, G. A. (1961). A model for predictive measurements of advertising effectiveness. Journal of Marketing, 25 (6), Lippmann, W. (1922). Public opinion. New York: Free Press. Just, M., Crigler, A., & Buhr, T. (1999). Voice, substance, and cynicism in presidential campaign media. Political Communication, 16, McCombs, M. (1997). Building consensus: The news media s agenda-setting roles. Political Communication, 14, McCombs, M., Liamas, J. P., Lopez-Escobar, E., & Rey, F. (1997). Candidate image in Spanish elections: Second-level agenda-setting effects,ʺ Journalism and Mass Communication Quarterly, 74 (4), McCombs, M., Lopez-Escobar, E., & Llamas, J. P. (2000). Setting the agenda of attributes in the 1996 Spanish general election. Journal of Communication, 50 (2), McCombs, M. E., Shaw, D. L., & Weaver, D. (1997). Communication and democracy: Exploring the intellectual frontiers in agenda-setting theory. Mahwah, NJ: Lawrence Erlbaum Associates. Mendelsohn, M. (1996). The media and interpersonal communications: The priming of issues, leaders, and party identification. The Journal of Politics, 58 (1), Merritt, D. Buzz. (1998). Public Journalism and Public Life: Why telling the news is not enough. Mahwah, NJ: Lawrence Erlbaum Associates. Miller, M. M., Andsager, J. L., Riechert, B. P. (1998). Framing the candidates in presidential primaries: Issues and images in press releases and news coverage. Journalism and Mass Communication Quarterly, 75 (2), Patterson, T. E. (1994), Out of order, New York, NY: Vintage. Perloff, R. M. (1998). Political communication: Politics, press, and public in America. Mahwah, NJ: Lawrence Erlbaum Associates. Poindexter, P. M. & McCombs, M. E. (2000). Research in mass communication: A practical guide. Boston, NY: Bedford/St. Martin s. 342

15 Pride, R. A. (1995). How activists and media frame social problems: Critical events versus performance trends for schools. Political Communication, 12, Rogers, E. M., Dearing, J. W., & Bregman, D. (1993). The anatomy of agenda-setting research. Journal of Communication, 43 (2), Scheufele, D. A. (1999). Framing as a theory of media effects. Journal of Communication, 49 (4), Scheufele, D. A. (2000). Agenda-setting, priming, and framing revisited: Another look at cognitive effects of political communication. Mass Communication & Society, 3, ) Semetko, H., Blumler, J., Gurevitch, M., & Weaver, D. (1991). The formation of campaign agendas: A comparative analysis of party and media roles in recent American and British elections. Hillsdale, NJ: Lawrence Erlbaum Associates. Severin, W. J. & Tankard, J. W., Jr. (2001). Communication theories: Origins, methods, and uses in the mass media, 5 th edition. New York, NY: Longman. Sherman, S. J., Mackie, D. M., & Driscoll, D. M. (1990). Priming and the differential use of dimensions in evaluation. Personality and Social Psychology Bulletin, 16 (3), Tankard, J. W. Jr., Hendrickson, L., Silberman, J., Bliss, K., & Ghanem, S. (1991). Media frames: Approaches to conceptualization and measurement, Paper presented to the Association for Education in Journalism and Mass Communication, Boston, NY.The Austin Chronicle, Feb. 15, 2002 Vol. 21, No. 24. Weaver, D. H. (1977). Political issues ad voter need for orientation. The emergence of American political issues: The agenda setting function of the press. Eds., D. L. Shaw & M. E. McCombs. St. Paul, MN: West Publishing. Westlye, M. C. (1991). Senate elections and campaign intensity. Baltimore: Johns Hopkins Press. Willnat, L. & Zhu, J. (1996). Newspaper coverage and public opinion in Hong Kong: A timeseries analysis of media priming. Political Communication, 13, Previous Paper Next Paper Contents 343