Read The Nature and Origins of Mass Opinion by John R Zaller.

Zaller outlines and tests against opinion poll data what he calls the Receive-Accept-Sample (RAS) model based on four axioms:

Reception Axiom: The greater a person's level of cognitive engagement with an issue, the more likely he or she is to be exposed to and comprehend -- in a word, to receive -- political messages concerning that issue.

Resistance Axiom: People tend to resist arguments that are inconsistent with their political predispositions, but they do so only to the extent that they possess the textual information necessary to perceive a relationship between the message and their dispositions.

Accessibility Axiom: The more recent a consideration has been called to mind or thought about,the less time it takes to retrieve wthat considerations or related considerations from memory and bring them to the top of the head for use.

Response Axiom:Individuals answer survey questions by averaging across the considerations that are immediately salient or accessible to them. (p.58)

Zaller makes a range of deductions from these axioms which he tests against the available data. Much of the book is fairly heavy going if you are not au fait with regression analysis and similar statistical devices (terms such as nonmonotonic and collinearity get used a fair bit). Still, there is plenty of good stuff among the careful academic tedium.

Zaller's most striking and powerful point is that political opinion will vary strikingly according to both political awareness and predisposition. Particularly in the short-term, any opinion changes will be strongly concentrated among the moderately politically aware, because those of low awareness will receive little or no information to cause them to shift their predispositions while the highly politically aware will be most resistant to information and arguments against their predispositions because their political awareness is a product of high political engagement. He examines a wide range of cases to demonstrate the pervasiveness of this.

Zaller is writing to the tradition of trying to tease out the relationship between elite behaviour and mass opinion. This is unfortunate, because he ends up using elite and elites in a fuzzy way to encompass three different, albeit related, phenomena: expert judgement, public conduits of information, public advocates. This is much more a criticism of his use of terms rather than the actual analysis, as it is quite clear that he doesn't confuse these in his thinking.

If elite opinion (in the obove senses) is unified, then public opinion will tend to conform to it, because there is little counterveiling information. In particular, elite sources conforming to any particular set of (mainstream) public dispositions will be "on message". If elite opinion divides, public opinion is likely to shift accordingly. There is an obvious problem with this from recent Oz experience, as it is not hard to think of a range of issues (e.g. migration, land rights) where dominant elite opinion has not been particularly persuasive. But even though Zaller does establish the importance of shifts in expert opinion (for example, on race and homosexuality) he also emphasizes that the public need only to be able to recognize which elites share their predispositions and take cues from them (p.328). Thus, messages which explicitly go against basic, longstanding values (one set of rules for everyone) or involve accepting an insult (people like you are morally inadequate cretins who shouldn't have a say) aren't likely to be widely adopted.

Zaller makes it clear that the mere presentation of counterveiling opinion can have large effects. There is considerable social science research on this, well-reported in Cass Sunstein's Why Societes Need Dissent. The belief that, if one can drive dissenting views from the public arena, persuasion of the public will be much easier does get support from Zaller's work. (On the other hand, cognitive conformity also greatly increases the chance of substantial error.) Zaller emphasises the importance of cognitive diversity among experts in making their work useful and persuasive.

Zaller establishes well that public opinion surveys are often very poor ways of teasing out complex reactions to events and issues. That they are "effervesences" of more deeply held considerations and so subject to considerable change depending on which such considerations are salient when people give their responses (including made salient from the surveying process).

I was impressed with clear tie-ins with findings from other scholars. Zaller makes it quite clear the public opinion is typically about balancing considerations, that debates are within as much as between people. Something which comes out in such disparate works as Eamon Duffy's The Voices of Morebath and Alan Wolfe's One Nation After All.

An excellent and informative work.