Scholars often lament that the public - as individuals - are ignorant regarding politics, and of their own policy preferences. This was supported, in part, by Campbell et al (1964) who illustrate the policy preferences were not strongly correlated with vote choice. This was further supported by Converse (1962) who found that the opinions of individuals fluctuated rather wildly over time. As such, Converse argued "large portions of the electorate do not have meaningful beliefs even on issues that may have formed the basis of intensive controversy among elites for substantial periods of time" (1962). Converse (1970) furthered his argument, claiming the public had "nonattitudes," and simply gave a random response because they felt obligated to do so.
Page and Shapiro (1992) contend that the individual voter may be ignorant of political issues in general, that is, individual policy preferences may be "incorrect" from a rational perspective, but the aggregate collective policy preference of a majority of citizens is not only rational, but stable. In part, the authors contend that this occurs through the theory of large numbers and subsequently normally distributed samples.
Page and Shapiro (1992) recognize that citizens may not have a deep understanding of specific political issues, but they do possess a more general understanding of major issues. They cite the example that citizens may not know what the acronym SALT stands for but they do recognize that the US and USSR are engaged in arms reduction talks.
Citizens understanding of major issues are coupled with the argument that individuals have some fundamental needs and have "uncertain beliefs" regarding how public policies will affect those needs. As such, Page and Shapiro contend that individuals have a long-term, although vague, preference for policies which will best serve their fundamental interests. Because preferences are uncertain in the short-term, they are susceptible to various sources of information and may fluctuate. However, if we "average" an individual's preference over time, we can find their "true" preference. There will be a "central tendency" and responses will fluctuate around this point. As such, if the individual has a "true" preference, then in the aggregate, there will be a collective preference. What is of prime interest here is that with a large collective sample, things like misinformation, mood swings, sampling error, etc. will be balanced out. As such, widespread public opinion can be an accurate description of the real interest of the majority of citizens, even if their own personal policy stances may not be truly representative of their core self-interest.
Page and Shapiro also draw into question the idea that public opinion changes rapidly and is unpredictable. They contend that in general, public opinion is quite stable, and when it does change, it changes in predictable ways.
The authors contend that when policy opinion changes, it does so in response to "changes in information, and changes in reality" (1992, pg. 53). Page and Shapiro divide changes in reality into two categories (a) events that affect individuals directly, and (b) those which do not affect the individual directly, but are interpreted by the individual in regards to costs and benefits. Circumstances that affect people directly include events which change the social, economic, and political life of individuals. For example, Page and Shapiro cite increases in income may promote greater support for increased social programs, or a rise in industrialization may lead to increased support for labor activity. Page and Shapiro write, "Such trends can be viewed as exogenous influences upon public opinion - independent influences subject only minimally, if at all, to elite preference manipulation" (330).
However, when events do not have a direct affect on individuals, "Citizens' assessments of the significance of trends and events - indeed, their very awareness of them - often depend crucially upon information and (especially) interpretations provided by elites, largely through the mass media" (331). This suggests that the media and elite discourse has the potential to shape public opinion. As such, in order for individuals to form valid, rational policy preferences, citizens must have access to valid information. In part, this rationality stems from "collective deliberation."
Collective deliberation helps the public to wade through the tremendous amount of political information. Page and Shapiro contend that information is processed through a complex system with many actors. The system is composed of many "specialized elements," i.e. experts, journalists, policy analysis, etc. who are all part of a larger, complex system in which they communicate with one another. This leads to a collective analysis of relevant information regardless of policy (Zaller 1992 would contend that there is a control over this debate which shapes preferences). Again, this relates to the theory of large numbers in that many individuals analyzing information, inaccuracies and biases will cancel out. Page and Shapiro write, "The public as a whole responds sensibly to events, not only as a result of statistical aggregation of individual preferences, but also because social processes of collective reasoning often produce and communicate high-quality information and interpretations" (366). Again, Zaller (1992) would disagree.
