Frank Zaller, The Nature and Origins of Mass Opinion (1992)

Main Argument:

Partisan stereotypes have considerable influence in political information processing, suggesting that political parties play an important role in voter’s decision-making (heuristics). Zaller challenges the idea that voters only have one true preference; instead he presents a model where individuals have conflicting views on specific issues and the “winning” view at any given time is determined by what considerations are at the top of your head.


Key Definitions:

Consideration: Any reason that might induce an individual to decide a political issue one way or another.

Political awareness: An individuals reception and comprehension of communications from the political environment. According to Zaller, political awareness is best measured by simple tests of neutral factual information since factual information is critical for intellectual engagement with politics.

Political predispositions: Stable, individual-level traits that regulate the acceptance or non-acceptance of the political communication the person receives. Predispositions are the critical intervening variable between the communications people encounter in the mass media, on the one side, and their statements of political preference, on the other (since they determine the accept part of the RAS model).

Values: General and enduring standards that hold a more central position than attitudes in individuals belief systems.


Method: mostly theory-building. To test his RAS model, Zaller relies primarily upon NES survey data. Specifically, he applies his theory to the dynamics of public opinion on a broad range of subjects, including domestic and foreign policy, trust in government, racial equality, the Vietnam War, and presidential approval.



== Notes ==



Chapter 2: Information, predispositions and opinion


–  A foundation for understanding the relationship between information and predispositions, forming public opinion

–  People have a limited capacity to absorb political information; hence they take cues from political elites in order to internalize political events/phenomenon

–  Three key themes: political awareness, political predispositions and the resulting marriage of the two


Political awareness: “the extent to which an individual pays attention to politics and understands what s/he has encountered” [21]

–  It is measured in terms of political expertise, cognitive complexity, political involvement, attentiveness, sophistication and political acuity

–  Zaller concludes that awareness is associated with exposure to information as well as the ability to react critically

–  Information is disseminated through elite actors such as parties or the media, creating a depiction of reality that is simple enough for ordinary people to understand

–  While the highly politically aware are the most likely to follow elites, Zaller states that it is the moderately aware that are most susceptible to influence as they do not possess critical skills to resist biased information [19]


Political predispositions: “stable, individual level traits that regulate the acceptance on non-acceptance of the political communications received by an individual” [22]

–  Predispositions are long-term variables not influenced by elites and the critical intervening variable between political communications and political preferences

–  Impact of predispositions depends on whether people possess the contextual information needed to translate their values into support for policies or candidates [25]


Measurement of opinion however, is constrained by various factors: instability, measurement error, response effects and question wording effects (see pages 30-34 for definitions)

–  Instability: individuals lacking strong opinions, but acting as if they do in order to fulfill the interview [31]

–  Measurement error:  the difficultly encountered in mapping pre-existing opinion into the vague language of survey questions [31]

–  Response effects: error that is produced owing to changes in question context, order and trivial alterations in questions [32]

–  Question wording effects: the possibility of an issue being made salient to the respondent strictly due to its mention or positioning in the survey [33]


Problems with Survey Research and Mass Opinion:

–  Mass opinion surveying is further hindered by the error that is built into the question-answer process

–  This refers to the process by which individuals construct opinion reports in response to the particular stimulus in front of them [35]

–  In other words, an issue is made more salient to an individual because of recent experiences, news broadcasts or even the questionnaire itself

–  Zaller states that it is a general misconception that citizens, although generally poorly informed, take pause to learn about issues that are important to them

–  What appears to occur, rather, is that information cues individuals to form opinions rooted in their level of awareness and previous predispositions


The author concludes that people are exposed to a constant stream of information but are not critical about what they internalize (36).  This can be directly compared with the Lodge et al. model of an individual’s “on-line tally”.


Chapter 3: How citizens acquire information and convert it into public opinion


\”Receive-Accept-Sample\” model: your stated opinions reflect considerations that you have received (heard or read about), accepted (if they are consistent with prior beliefs), and sampled from (based on whats salient at the time)

–  This outlines the process in which people receive new information, decide whether to accept it and sample at the moment of answering questions

–  According to the RAS model, individuals will support or oppose policy based on the immediate positive and negative considerations available in the person’s mind when answering survey questions

–  Zaller refers to this as “top of head” information, which is stated to be more powerful than “true attitudes” towards a subject [50]


–  Two types of political messages can be derived from elite discourse: persuasive and cueing messages

–  Persuasive messages provide the reader with a reason to take a point of view; such messages appeal directly to non-rational emotion

–  A cueing message is contextual information that encourages the citizen to form ideological implications  These messages enable individuals to perceive relationships between persuasive messages and therefore, allow them to respond critically to such information [42]


Four Axioms of Zaller’s Model: [42-50]


  1. Reception axiom: The greater a person’s level of cognitive engagement with an issue, the more likely s/he will be able to comprehend political messages
  2. Resistance axiom: People tend to resist arguments that are inconsistent with their political predispositions, but are limited by access to contextual information and political savvy in doing so
  3. This gives no allowance for citizens to reason or deliberate about politics
  4. People react mechanically on the basis of external cues
  5. This assumption differentiates between high and low involvement citizens
  6. Low involvement use superficial cues such as source credibility for accepting/rejecting messages
  7. Accessibility axiom: The more recently information or opinion has been used, the closer it will be to “top of the head”, therefore it takes less time for the individual to retrieve their opinions on that topic
  8. Response axiom: Individuals answer questions by averaging across the considerations that are immediately salient or accessible to them


Chapter 5: Making it up as you go along


–  Using his RAS Model to explain motivations behind public opinion, Zaller further explores limitations to mass opinion surveys

–  With his RAS assumption that people do not really know their opinions, but instead are heavily influenced by what is at the top of their minds, Zaller states that survey answers are susceptible to exogenous influences

–  Some of these influences may be question ordering, characteristics of the interviewer, reference groups and priming effects


Priming Effects: predispositions are stated to function as “gatekeepers”, acting as reference points for opinion formation

–  Priming effects can be either long or short-term, accomplished through elite discourse over a series of months or by one particularly poignant news story

–  Salience and framing: influence opinions as does question framing, which can engage different “gatekeeper” attitudes [83]

–  This is further complicated by the multitude of attitudes an individual may hold towards a particular issue; the result of which is different priming effects having different results depending on the time and issue salience


The “stop and think” method:

–  Extra thought on an issue may encourage people to further search their memory for their opinions and thereby raise considerations in formulating their response

–  Quantitative data however, failed to find a correlation between the “stop and think” method and ideological consistency

–  This proves to be a limitation of the RAS model as it did not account for why inducing people to utilize a larger base of considerations should undermine reliability




  1. Zaller establishes that “Ambivalence Deduction” indicates that individuals harbour varied feelings towards aspects of most issues
  2. Individuals base their survey responses on the considerations that are most salient to them; this is referred to as the “Response Axiom”
  3. When an individual makes aggregate judgments there is no need to reconcile contradictory reactions to events and issues  [92]


Chapter 9: Two Sided Information Flows


–  Examines the effect of the RAS model on mass opinion when conflicting information is offered

–  Public attitudes on issues may alter in response to changes and intensities of competing streams of political communication as filtered through the reception-acceptance process [190]


Case Study: Reactions to the Vietnam War

Zaller outlines three simplifying assumptions in the RAS model that account for malleability in intensities with respect to two-sided communication

–  First, it is not possible to measure the reception of individual speeches, news stories or individual communications

–  Second, no consideration remains active for longer than two years unless it is reinforced by current arguments

–  Finally, people respond to survey questions off the top of their head as opposed to internalizing all of the available information [191]

The expected result of these assumptions under the RAS model is that rates of acceptance of incoming messages ought to decline as a function of political awareness and ideological distance [192]



–  The relationship between ideological evaluations and issue opinion may flow in either direction

–  Effects of values and awareness on political attitudes depend on elite cues for activation; and values and awareness had significant relationship with resistance biased communications, but only after elite commentary activated public predispositions towards support or opposition [202]


Implications for the strength of two-sided communication effects:

–  Countervalent resistance occurs among highly aware individuals

–  Two-sided messages can take different forms at different times, depending on the intensities of the opposing messages and the prior distribution of opinion [207]

–  Mass belief systems alter over time in response to a complex stimulus [208]


To conclude, Zaller finds that public attitudes are susceptible to competing intensities of political communication.  This competition forms a relationship between elite actions and ideological responses; when elites agree, the public’s response is non-ideological and when they disagree along partisan lines, response becomes ideological.  Therefore, the degree to which the masses polarize along ideological lines reflects the relative intensity of the opposing information flows [210]


Place in the Literature: Zaller can be seen as a more nuanced version of Converse (1964) and Iyengar and Kinder (1987). See Zaller and Feldman (1992) for a more concise statement of essentially the same argument. But see also Zaller (1998), in which Zaller argues that the argument in this book attributed too much influence to elites