Log in Page Discussion History Go to the site toolbox

501 Readings Week 10

From BluWiki

Mutz, Diana C. 1994. “Contextualizing Personal Experience: The Role of the Mass Media.” The Journal of Politics. Vol. 56 (3): 689 – 714.

What? Examines the role of the mass media in translating personal experiences into political preferences (689). Examines two competing hypotheses: 1. “high levels of media attention should facilitate self-interested political attitudes”. 2. “the effects of personal experience should be greatest among the inattentive and poorly informed” (690). “…I argue that mass media makes a unique contribution to the politicization of personal experience by exposing people to the similar experience of others…” by defining problems as national trends, by legitimizing the attribution of responsibility for such problems to national leadership, by sometimes making clear connections to political causes as sources of the problem, and by framing problems as either episodic or as societal trends (692 - 694).

How? Examines two potential methods by which media enables translation of personal experience into political preferences, one direct, the other indirect by way of a telephone survey and analysis of amount of media coverage (695, 698). In assessing these competing hypotheses, we need to be clear about what we are measuring. Many researchers are not clear whether they are examining differences in education, socioeconomic status, or media exposure (692). Analyzing media coverage enables some degree of separation of media influence and political interest or sophistication (698).

General Conclusion: While most argue that personal experience does not translate into political preferences because individuals tend to “morselize” such experiences, Mutz finds that mass media counters this tendency by “legitimizing” the translation of personal experience into political preference through stories that allow the creation of a story linking personal experience with larger national experiences or trends (689).

Argument: According to most research, people tend to morselize their experience – they tend to not see such experience as part of a larger pattern, because they tend to not attribute responsibility for their experiences to political causes (689). Personal concerns are typically juxtaposed with sociotropic concerns in explaining political preferences. Sociotropic concerns are posited to translate to political preferences more easily because information received from mass media emphasizes national concerns and clearly links these concerns with political causes (690).

One reason to believe that those who are more well-informed will be more likely to politicize personal experience is that because media presents information in aggregate form, exposure to media may enable the connections to be made between personal experience and larger national trends (690 - 691). The reason that we think that those who are poorly informed are more likely to rely on personal experience is that information processing theories suggest that those without alternative information will “default” to easily accessible personal experiences as bases for decisionmaking (691).

Sources of media differ in their ability to present information as part of the larger context. Television is least able, daily newspapers face constraints, but may be able to present information in context for certain stories, news magazines most able to provide context and connections to larger trends (694). For these reasons, types of media use need to be accounted for.

Findings: Party identification becomes more important to presidential approval with greater media consumption (701). Difference in media consumption, during periods of light coverage of unemployment, doesn’t seem to impact the translation of personal experience into political preferences. But, during heavy coverage of unemployment, those with more exposure to media are more likely to politicize their personal experience than those with relatively little media exposure (702). The types of personal experience politicized differ. Those who we would expect to be affected by experience of unemployment (“television-dependent and nonmedia groups”) tend to politicize actual experiences of unemployment while “print-intensive media groups” (those we would expect to be least subject to personal experience of unemployment) politicize personal concern about unemployment (705). Retrospective assessments also seem to be informed by media coverage: When media coverage of unemployment was highest, people were more likely to use personal experience in assessing national economic conditions (705-706). Because politicization of personal experience occurs across media user types during periods of heavy coverage, these findings suggest that heavy coverage can overcome level of attention to media and the tendency to morselize personal experience (707). That is, there is support for both direct and indirect influences of the media on the translation of personal experience into political preferences (708). But, we should be careful not to generalize these findings across issues, coverage of unemployment may differ from coverage of other issues (709).

Sears, David O., Richard R. Lau, Tom R. Tyler, Harris M. Allen, Jr. 1980. “Self-Interest vs. Symbolic Politics in Policy Attitudes and Presidential Voting.” The American Political Science Review. Vol. 74 (3): 670 – 684.

What? Test of two competing hypotheses about the source of political preferences: Self-interest versus symbolic politics (671).

Why? Seeks to understand the role played by issues in political judgments. Response to rational actor models of voting, examines the role of self-interest in voting decisions. Role of self-interest in determining political preferences has had relatively little empirical treatment; studies that have examined it have been methodologically flawed (671). Demographic variables have been used as proxies for policy impact but these variables also reflect socialization – using demographic characteristics doesn’t allow disentanglement of self-interest and symbolic politics (671 – 672).

How? To make the hypothesis falsifiable, they limit the definition of self-interest to the pursuit of short-term, material interests. Other goals may exert some sort of influence on preference, but less straightforward types of self-interest (long-term, altruistic concerns for example) are considered separately (671). Because there has only been weak support for the role of self-interest, this study examines special circumstances under which self-interest may play a larger role (672). Also examines the role of self-interest in the formation of ‘issue publics’ (672). Alternative to self-interest is termed “symbolic politics”. This is associated with partisan preferences and the like, political attitudes are seen as related to long-term values about the ‘good society’ and to pre-adulthood socialization. “In the world of ‘symbolic politics’, one’s political and personal lives exist largely isolated from one another” (671).

Findings: “Generally speaking, we found self-interest to have little effect on voters’ policy preferences, while symbolic attitudes had major effects” (673). The authors then examined the role of self-interest under “special conditions”. Specifically, they looked at the interaction between “political sophistication, private-regarding values, perceived government responsiveness, sense of political efficacy, and perceiving the issue as a very important problem” and self-interest. Generally, they found that these potential mediating variables had no ‘appreciable’ impact on the role of self-interest in determining political preferences (674). More specifically, they found that political sophistication strengthens the role of symbolic attitudes (674- 675). Further, they found that ideology does not influence self-interest (675). Additionally, perceived government responsibility appeared to have little impact on the role of self-interest. Symbolic political attitudes “seem clearly not to express the short-term personal impact of policy issues” (676). Variance contributed by demographic variables “overlapped” with that contributed by symbolic attitudes, so it seems likely that demographic variables reflect prior socialization and do not serve as a proper surrogate for self-interest (677). Similar findings emerged for voting: “So in voting behavior, as in the generation of policy attitudes, symbolic attitudes have the greatest impact [(here too, party identification is best predictor)], and short-term self-interest apparently has almost no effect” (678). Also found little support for the notion that self-interest plays a role in the formation of issue publics (679). Some studies have found self-interest to play a role in determining policy preferences, but these cases have involved studies with clear and concrete impacts on certain voters, so it may be that self-interest plays a role only rarely. This study calls into question the assumptions of the more ‘naïve’ versions of Rational Choice Theory. But, these findings also reflect a restricted definition of self-interest. Longitudinal studies are needed to better assess the role of self-interest in creating, maintaining, or changing symbolic political attitudes (680). Self-interest may operate through group membership, and this study provides some evidence for that, but more work is needed . Additionally, group identification may be more strongly related to symbolic politics (681). We need to amend current understandings of voters to incorporate and try to understand the role and origins of symbolic political attitudes (681 – 682).

Bibliographic information: Lodge, M., and M.R. Steenbergen with S. Brau (2001). “The Responsive Voter: Campaign Information and the Dynamics of Candidate Evaluation” Ch. 13 in Niemi, R.G. and H.F. Weisberg (eds) (2001). Controversies in Voting Behavior, 4th edition. Washington, DC: CQ Press.


This chapter explores how campaign information interfaces with decisions of voters. Two models of information recall are examined: traditional measures of recall, and on-line tally. It is found that on-line methods of evaluation are better suited for assessing information recall.

Supporting evidence:

Traditional models of information recall suggest that voters lack information. These traditional models assume that “the citizen’s failure to recall events is necessarily or even primarily a function of political inattentiveness, political ignorance, or (worse yet) irrationality” (240-241). Such assumptions are rejected. In turn, the traditional recall models lead to two paradoxes. First, “whereby voters (in the aggregate) oftentimes appear to be choosing the ‘right’ candidate and supporting the ‘correct’ issues stands but apparently without the conceptual or factual wherewithal to make such informed judgments” (241). Second, the disconnect between survey and experimental research. The former leads to “a strong, positive correlation between the mix of pros and cons in memory and the direction and strength of evaluation” (241). The latter lacks a direct “memory-to-judgment link” (241).

The authors suggest a bounded rationality approach where citizens are informed, “but are unable, for good reasons, to recollect accurately the considerations that entered into their evaluations” (241). Such good reasons may include the limitations of the human mind (242). Therefore, the fundamental issue is not recall beliefs or preferences, but how responsive “the citizens’ overall evaluation and vote choice are to the political information they considered throughout the campaign” (242).

Page 243 has an excellent figure which describes the on-line candidate evaluation.

The on-line evaluation model suggests a tally where a running tally is created of the candidate; this information is then stored in the long term memory of the voter (243). These long term impressions, then create a global affective assessment of the candidate. Thus, “at best, the citizen’s recollections will represent a biased sampling of the actual causal determinants of the candidate evaluation” (244). Yet, “at worst…the correlation between memory and judgments is the result of reversed causality, the causal arrow often going from evaluation to memory” (244).

The results of the experiment yield interesting results. First, salience of issues appeared to have little effect on the subjects. Thus, “issues were not recalled because of their salience” (251). Second, if it is assumed candidate evaluations are memory-based, it is seemingly possible to exclude the possibility that a vote choice is informed by campaign issues (252). Be as it may, recall about candidates was extremely limited, with “voting …based on no more than two issues about half the time” (252). Third, campaign message memory has an extremely short life span: It fades, and does so quickly (252).

Conclusions: Two conclusions are evident. First, voters are responsive to campaign information. That is to say, campaign information is effective at influencing how a candidate is perceived. However, the responsiveness is limited by the number and size of facts (with less retention available for ‘small’ facts), and the natural decay of memory. Thus, survey methods may not accurately reflect that voters are informed.

“Messages Received: The Political Impact of Media Exposure” Larry M. Bartels, American Political Science Review 87: 267-85

Previous analyses of the persuasive effects of the media have shown time and time again that the media has a minimal affect at best on its viewers. This may be because the media truly does have a minimal effect or this could be because of the methodological difficulties involved in measuring these effects. At the aggregate level, it is usually impossible to distinguish the effects of the media from the effects of the events themselves that are being reported on. At the individual level, differences in opinions between those that have been exposed to media coverage and those that have not could reflect preexisting differences in political attitudes among the groups rather then the media’s effect. Self-reporting has also proven to be problematic

Bartels tries to prove that media effects do have a strong effect in a presidential campaign setting, while also demonstrating why these effects are not more common than they are, and why previous analyses have not been able to find these effects. He “proposes a model of opinion formation that can help pinpoint the exact contribution which mass media makes to the individual’s cognition, feelings, and actions” (268).

Because many of the explanatory variables used are subject to measurement error, the biases created by introducing all variables as error free indicators of underlying opinions and behaviors will be substantial in magnitude and direction (268). Assumes a Bayes’ Rule that new information from the media is used to update political opinions.

Data comes form 1980 American National Election Study with three waves of opinion readings. Analysis incorporates party identification and demographics as exogenous influences on the nature of the message received (269).

Bartels argues the best way to deal with these biases from measurement error is to measure/estimate the magnitude of these errors. He argues that measuring “the same opinion or behavior at three or more time points provides leverage for distinguishing between change in underlying ‘true’ opinions and random measurement error” (269). In this study Bartels estimates variant with Wiley and Wiley model.

Finds that “measurement error is sufficient by itself to produce some upward revision in the apparent impact of media exposure.” (272), but also shows that measurement error overstates the degree of instability in preexisting opinions. Argues that because evidence demonstrates greater stability in preexisting opinions, the political messages required to persuade opinion must be new or distinctive.

Another way of saying this is that “the apparent effects of media exposure will often be modest in magnitude even when adjusted for the effects of measurement error – not because the media cannot be persuasive but because opinions at the beginning of a typical presidential campaign are already strongly held and because media messages during the course of the campaign, are in any case, only occasionally sharply inconsistent with preexisting opinions (275).

Looking at the media’s persuasive effect would be best measured by looking over long periods of time and by examining weak prior opinions, especially concerning “new” candidates and issues.

“Reexamining the ‘Minimal Effects’ Model in Recent Presidential Campaigns”

Steven E. Finkel

In this article, Finkel argues that, throughout time, “campaigns served mainly to activate existing political predispositions and make them electorally relevant” while possessing the potential to exert large electoral effects, but rarely doing so (1). He shows that attitudes such as party identification and presidential approval (measured before party conventions) can predict an overwhelming majority of votes. Both at the individual and aggregate levels, changes in party identification or political orientation during a campaign period generated little impact on overall vote choices. Campaigns, rather than changing existing opinions, serve mainly to “activate existing political predispositions and make them electorally relevant” (3). Finkel examines activation and conversion models with 1980 NES data. Finkel’s findings show that pre-campaign values of party identification and presidential approval ratings have strong effects on vote choice. However, the “change scores” for these variables are also statistically significant, which he claims are indicators that there is potential for conversion throughout the course of a campaign, even though his observed results showed limited examples of this. Furthermore, he finds that changes in “individual orientations for the most part went in accordance with their pre-campaign dispositions” (11). These findings are seen at both the individual and aggregate levels of voting in his analysis. He then continues by looking at the role of individual preferences and suggests that “the overall amount of conversion effects in a campaign may depend on the extent to which the electorate’s initial dispositions and initial preferences are in accordance with one another” (16). To sum up, his three main conclusions are: 1) a simple activation model which predicted individual votes (based on race, incumbent evaluations, and pre-campaing party ID) accounted for 80% of all votes in the 1980 presidential election, 2) political attitudes did change during the race but the result of these changes were not large enough to alter many individuals’ vote choices, and 3) extremely high rates of voter stability were seen with an analysis of pre-convention stated preferences. The implications of these conclusions suggest that political campaigns possess a potential to change the vote choices of individual voters, but that this effect has not been seen in recent elections (at the time of his writing). He goes on to say that the media may play a larger role in influencing campaigns than was thought at the time; most notably the media would be able to change the opinions of voters whose stated preferences and dispositions are “incongruent” at the beginning of a campaign. Finally, campaigns play many other roles besides converting voters. Campaigns can increase turnout, place items on the public agenda, offer a forum for coalition building, and articulate policy preferences of candidates.

“New Perspectives and Evidence on Political Communication and Campaign Effects“ Iyengar and Simon, Annual Review of Psychology 51:149-69


Politicians use the mass media as a tool to communicate important messages to the public and promote favorable political objectives. Media has invaded all government sectors especially during campaigns. Why do campaigns use the media as a resource if studies have proven time and time again that the media does not have more than a minimal effect of persuading viewers? Iyengar and Simon’s objective is to argue that conceptual and methodological roadblocks have obscured the ways in which the media influences voters.

Conceptual and Methodological Roadblocks

Previous research is limited to understanding the persuasive effects of the media but ignores other types of effects such as that on turnout or agenda setting. Too much reliance on survey methods, which have serious methodological issues. Assumption that self-reporting media exposure is most accurate measurement is problematic. These studies of self-reporting are more biased towards those that are more interested in politics or have better memories. Self reports do not deal with the reciprocal effects of exposure to campaign and partisan beliefs/attitudes

Bypassing the Roadblocks

In order to overcome these roadblocks we need to add “voter learning” (the acquisition of information about the candidates and issues) and “agenda control” (the use of campaign rhetoric to set the public’s political agenda).

Learning: Media exposure for campaigns more than ever is non-substantive and shallow in its coverage of the candidates and issues. More about the horserace and “juicy” news, rather than worrying about informing the public on important issues. Even with the shallow media exposure there is evidence from several studies that campaign exposure boosts citizens’ political information. Overview of studies on pg 155-6 “In sum, campaigns are information rich events. Contrary to the prevailing wisdom, the information they yield is multifaceted, encompassing the candidates chances of winning, their personal traits and mannerisms, and most important their policy and ideological bearings” (156).

Agenda Control: Voters are an arms distance away from politics and cannot be educated on all of the issues but are instead attuned to a few issues that are important at the moment. Because the media is a prime educator, politicians and candidates see a huge advantage in influencing the media and hence the citizenry in their favor. “Issues deemed significant by the electorate become the principal yardsticks for evaluating the candidates” (157). Candidates pursue issues for a comparative advantage Priming effects

New Theoretical Perspectives

The Resonance Model: Premise is that “campaign messages –whatever form they take – work their influence in concert with voter’s prevailing predispositions and sentiments…. Effects are contingent on the degree of fit between campaign messages and prevailing attitudes” (158). Strongest influence of electorate’s prior predispositions is party identification. Often creates a partisan “polarization” effect when activated by campaigns. Most influence of media goes to less intense Democrats and Republicans who are likely to move the most during campaigns. Can be seen in other prior dispositions such as group identification or political interest.

The Strategic Model: This model “specifies campaign effects as interactions but focuses instead on the competition between message strategies… recognizes the ability of the strategic interactions between the competing candidates and between candidates and the press to create different campaign contexts” (161). Messages conditioned by elite actors. Example is advertising tone of attack versus self-promotion (161). Another is how some news stations do ad watches on campaigns to fact check, constraining what the candidates can say in these ads (162).

Beyond Survey Methods: Incorporating experimental methods with better content analysis can allow for replication

“Anxiety, Enthusiasm, and the vote: the emotional underpinnings of learning and involvement during presidential campaigns”

Marcus and Mackuen

Marcus and Mackuen focus on two emotional indicators (anxiety and enthusiasm) in an analysis of political campaigns and voting behavior. The claim that anxiety “stimulates attention toward the campaign and political learning and discourages reliance on habitual cues for voting” while enthusiasm “influences candidate preferences and stimulates interest and involvement in the campaign” (672). They believe that voters use emotions as heuristics in forming political decisions during campaigns. The authors form an argument that suggests emotions are catalysts for political learning, and focus on the idea that “threat powerfully motivates citizens to learn about politics” (672). They claim that individual voter emotions aid in political reasoning and enhance the overall value of a democratic society. They provide a theoretical background, distinguish between the two emotions, and discuss the ways in which voting behavior is influenced by them. The authors distinguish between two emotional systems (one centered on anxiety and one centered on enthusiasm) that appear linked to “psychic orientation” (673). They then differentiate between the “behavioral inhibition system” and the “behavioral approach system”. The former continually matches incoming stimuli with current plans and expectations while the latter provides active, ongoing feedback of behavior and governs the “physical and mental resources necessary for success” (673). In other words, the appearance of threats causes individuals to prepare for action. Political threats, while not endangering physical states, threaten the symbolic worlds that consist of beliefs and values. Negative events have been shown to increase attention, which suggests that citizens begin to pay attention to politics when signals alert them to an imbalance or disruption in the status quo. Furthermore, the approaches suggest that variations in political enthusiasm should predict political involvement. Marcus and Mackuen then analyze their proposed system of dual emotional responses. The compare their system to an alternative view of “valence” responses and claim that their system reacts to the “psychic pressure” of electoral campaigns (674). They examine NES data from 1980 and a Missourian survey conducted on 1988 presidential campaign emotions. They find that enthusiasm and anxiety are indeed distinct emotional indicators (which defies the valence hypothesis) and use this finding to posit that emotions enhance the ability for an individual to interact with the surrounding environment. They claim that “only when emotions reliably react to changes in the informational environment can they encourage citizens to become engaged with their favorite candidate’s prospects or, more interestingly, interrupt citizens’ ordinary political activity and spur information processing” (675-76). With their data, the authors go on to say that anxiety is a dynamic feature of the political landscape, changing with each campaign, and that enthusiasm directly affects voting preference. Anxiety, instead, gives voters pause to base their decisions on candidates and information rather than voting with their gut instincts. The theory presented here suggests that enthusiasm will “directly affect the voting decision while anxiety’s role will be muted” (677). Anxiety will be the voice that coerces voters to gather new information and reevaluate their policy positions before an election. They conclude this portion of the paper by claiming that comparative enthusiasm “affects how closely people are willing to embrace either candidate” while anxiety “stimulates peoples’ attention and releases them from their standing decisions” (678). The authors continue by demonstrating that threatening situations provide learning examples for voters. They find that anxiety is associated with learning while enthusiasm is not. They conclude by claiming that enthusiasm increases campaign involvement while anxiety increases political learning. When voters become threatened or feel anxious about politics, they tend to open their eyes more in an effort to take in more information. However, when politicians inspire enthusiasm within the electorate, people immerse themselves in “the symbolic festival” (681). Finally, they conclude that emotionality both “empowers voters to confront their circumstances and react efficiently and appropriately” and enhances the ability of citizens to perform their democratic duties. In other words, the “nature of democratic government thus depends on how emotions get linked to political circumstances and how that link varies over time” (681).

Site Toolbox:

Personal tools
GNU Free Documentation License 1.2
This page was last modified on 5 April 2010, at 14:58.
Disclaimers - About BluWiki