Microsociological Aspects of Public Opinion Research Paper

1. Theoretical Foundations

The theoretical literature on public opinion addresses diverse phenomena that have puzzled social thinkers. They include its resistance to changes in social structures and policy outcomes; its capacity for immense movement if ever this resistance is overcome; its sensitivity to the ordering of social shocks; and its imperfect predictability.

Most variants of the theory draw on a class of constructs known as ‘threshold,’ ‘critical mass,’ ‘bandwagon,’ or ‘cascade’ models. Originally designed to explain epidemics, these were subsequently applied most influentially by Schelling (1978) and Granovetter (1978) to various social phenomena that exhibit periods of stability punctuated by sudden change, for example, stock market bubbles, bank runs, cultural fashions, corporate conservatism, and academic fads. The common feature of these phenomena is that they are shaped by interdependencies among individual decisions.

Public opinion harbors two distinct interdependencies. Wherever individuals lack reliable information of their own, they look to others for clues about the reality they are seeking to grasp. Such ‘free-riding’ on publicly available knowledge makes people’s genuine wants interdependent. Also, in articulating preferences and conveying knowledge, individuals frequently tailor their choices to what appears socially acceptable. This conformism, which is motivated by the wish to earn acceptance and respect, creates interdependencies among the wants that individuals convey publicly.

2. The Public Opinion Of A Poorly Informed Group

The essential contributions of the theory may be sketched through a model drawn from Kuran (1995), which contains additional details. Certain initial assumptions will be relaxed further on.

2.1 Thresholds And Expressive Equilibria

Consider a 10-person society that must decide a certain issue, say, the government’s crime-fighting budget. Two alternatives present themselves: low (L) and high (H). Each individual will support whichever option appears socially optimal. Everyone will want this budget increased if crime seems to be rising but decreased if crime appears to be falling.

By assumption, no one has perfectly reliable private information. Consequently, every individual’s judgment will depend partly, if not largely, on the apparent judgments of others. This does not mean that individual beliefs must converge. Because of differences in private information, any given common signal about the distribution of preferences may make some people favor H and others L.

For any individual, the minimum share that must consider H optimal for him or her to concur constitutes what is called that individual’s responsiveness threshold. This threshold is a number between 0 and 100. Listing all 10 thresholds in ascending order yields society’s threshold sequence (T). Here is an example:

T: (0, 10, 30, 30, 40, 50, 50, 60, 70, 80)

These individuals may be labeled a, b, …, j, with a considered female, b male, and so on, in alternating fashion.

Given the heterogeneity reflected in this sequence, the perceived division between the supporters of H and those of L can influence dramatically the realization of public opinion. For a demonstration, suppose that initially only 10 percent of society is thought to favor H. Whatever the source of this perception, a and b will favor H, because their thresholds lie at or below 10; and the remaining eight members of the group will favor L, because their thresholds lie above 10. The resulting public opinion will thus lean heavily against H, which will enjoy the support of only 20 percent of the 10 individuals. The outcome will not impel anyone to switch positions, so it represents an expressive equilibrium. The knowledge that only two of the 10 individuals favor H will make exactly that share do so.

This particular expressive equilibrium is not the only possible self-sustaining outcome. If initially society was believed to be divided evenly between the two options, the first seven members of the sequence (all except those with thresholds above 50) would have favored H. Observing that 70 percent favor H, the two individuals with thresholds of 60 and 70 would then have joined the majority, raising this support to 90 percent. Finally, completing the learning process, this second adjustment would have impelled the lone holdout to join in. As with the above case of a divided public opinion, the resulting unanimity would have been self-reproducing.

2.2 Properties Of Public Opinion

This pair of illustrations highlights several characteristics. Public opinion is path dependent, in that changing its historical starting point might alter the outcome. It is also sensitive to small events, which are intrinsically insignificant happenings that an analyst seeking a restrospective explanation would probably overlook. Whereas an initial perception of 20 percent support for H would be self-confirming, a perception of 30 percent would lead to a consensus in favor of H. The difference in these two outcomes is vastly disproportionate to the difference in their starting conditions, whose determinants would not necessarily register on a historian’s viewscreen. Finally, public opinion may be unrepresentative of the available private information. Where the outcome is a consensus, it conceals the existence of substantial private information favorable to the unanimously rejected option.

2.3 Informational Cascades

The iterative process through which a self-reproducing public opinion takes shape generally is known as an informational cascade (Bikchandani et al. 1992). Under certain conditions, an informational cascade will enhance social efficiency. For example, where the first people to articulate their views are those with the most accurate information, the consequent interdependent learning will help to dampen private ignorance. However, such a fortuitous outcome cannot be guaranteed. Instead of educating the poorly informed, the informational cascade might serve to un-educate the most knowledgeable. This finding points to yet another characteristic of public opinion. It is sensitive to the order in which people express themselves.

Insofar as the evolution of public opinion influences policy decisions, it may weaken the political system’s effectiveness in addressing social concerns. For example, it might induce the government to pour resources into solving a problem widely considered serious only because the pertinent social learning was spearheaded by the ignorant. Yet, the resulting allocation would not be undemocratic. Whatever the reason why public opinion has come to favor the allocation, and however wasteful an outsider might consider the instituted policy, the members of society would only be exercising the freedom to make their own mistakes. Moreover, the government would simply be responding to the popular will.

3. Reputational Determinants Of Public Opinion

Though analytically useful, the above account provides an undersocialized view of human nature. In practice, people do not express their potentially controversial views mechanically, without considering how their social standings might be affected. As social beings who wish to be liked and accepted, they may refrain from expressing beliefs or preferences that are likely to offend powerful groups. Some evidence lies in Asch’s (1957) line experiment, in which subjects contradicted the evidence of their own senses to avoid isolation and criticism, and in Noelle-Neumann’s (1984) pre-election surveys, in which the publicly acknowledged party preferences of respondents, though not necessarily their private sentiments, varied according to their perceptions of which party was running ahead. Additional evidence comes from Scott’s ( 1985) ethnographic research, which finds that peasants may profess loyalty to their political regime, yet also work to undermine it in settings where they can do so with impunity.

The activists who direct society’s pressure groups seek to exploit the reputational concerns that such studies highlight. They do so partly by treating people differently according to their public expressions and partly by providing incentives for criticizing, belittling, perhaps even ostracizing, overt opponents of the groups they are promoting. Insofar as such tactics are successful, individuals tailor their public positions to the prevailing social demands, and public opinion is thereby affected.

3.1 Private And Public Preferences

Depending on the nature of the information distorted to obtain social rewards and escape social punishment, such misrepresentation is called either preference falsification or knowledge falsification. On any given issue, preference falsification occurs when an individual’s publicly expressed preference, or simply public preference, differs from the corresponding privately held preference, or private preference. Similarly, knowledge falsification entails a discrepancy between the communicator’s private knowledge and public knowledge.

Whereas a person’s private preference always reflects the relevant private knowledge, this person’s public preference and public knowledge need not be mutually consistent. In practice, however, the two choices tend to be tightly coordinated in order to make the former credible. Someone who pretends to favor one particular option, but then proceeds to give reasons for another option’s superiority, will hardly come across as sincere; the person’s intended audience will sense that they are concealing something.

3.2 Reputational Cascades

For further insights into how individual expressions both mold public opinion and are molded by it, reconsider the threshold sequence of Section 2.1, adding the assumption that if the 10 individuals were polled anonymously, H would be favored five-to-three, with two abstentions. Specifically, the five members with thresholds of up to 40 would support H, the two with thresholds of 50 would be indifferent, and the three with the highest thresholds would support L. These preferences expressed under the cover of anonymity are private preferences. Thresholds depend partly on private preferences, because submitting to others’ demands causes psychological discomfort, albeit of a magnitude subject to variation across contexts and individuals. A second determinant of thresholds is sensitivity to social pressures. Under this section’s more realistic scenario, then, a threshold reflects not only the nature and depth of an individual’s private preference but also that individual’s preparedness to pay the price of joining an underprivileged expressive minority. Put differently, a threshold captures an individual’s tradeoff between the benefits of truthful self-expression and those of reputationally advantageous preference falsification. If person c’s threshold is 30, the sum of their expected satisfaction from all sources, including truthful self-expression and both favorable and unfavorable reputational consequences, is maximized by expressing a preference for H whenever the share of its supporters appears to be at least 30. Otherwise, their expected satisfaction is maximized by publicly favoring L.

This threshold sequence produces, as already noted, two stable equilibria that represent highly different realizations of public opinion. Under the current assumptions, the selection among these equilibria occurs through a reputational cascade. In the course of a reputational cascade, people conform to the apparently popular choice not because they see wisdom in numbers but merely to maintain a favorable reputation. In so doing they make it increasingly difficult for others to remain nonconformist.

3.3 The Possibility Of Inefficiency

Like the social learning mechanism described earlier, this social adaptation mechanism driven by reputational concerns makes public opinion path dependent, unrepresentative of the available private information, and sensitive to both small events and the order of individual choices. Additionally, it may make public opinion unrepresentative of private opinion. In the illustration of Section 3.2, neither of the possible outcomes matches what an anonymous opinion survey would identify as the ‘popular will.’ Whereas three members privately favor it, H could achieve either unanimous support or the support of just a two-person minority. Either way, social policies designed to accommodate public opinion would be responding to signals distorted by preference falsification. As such, these policies could be inefficient.

3.4 Availability Cascades

There are contexts in which public opinion is shaped by either purely informational or purely reputational cascades. As a rule, however, the outlined informational and reputational mechanisms feed on each other, or at least interact. The composite mechanism whereby public opinion evolves through both social learning and conformism aimed at securing social acceptance is known as an availability cascade (Kuran and Sunstein 1999).

4. Feedback From Public Discourse To Private Opinion

Reputational cascades treat private preferences as fixed. For their part, informational and availability cascades recognize the plasticity of private preferences, but only during the transition to a self-sustaining public opinion. In reality, this plasticity is not limited to periods of disequilibrium. Whether public opinion is at rest or in flux, the private knowledge that undergirds private preferences is ordinarily shaped and reshaped by public discourse, which consists of the suppositions, observations, and arguments that are communicated publicly (Coleman 1990).

Preference falsification affects this private learning through two channels. As both informational and availability cascades posit, the state of public opinion, whatever its inconsistencies with private opinion, molds individual interpretations of reality. In addition, the knowledge falsification that accompanies preference falsification sows ignorance and confusion by corrupting information in the public domain through the substitution of facts that people know to be false for ones they consider true. Over time, knowledge falsification may even make once-common objections to a publicly popular option disappear from society’s collective memory.

Under diverse conditions, population renewal will facilitate such an outcome. If public discourse excludes criticisms of the fashionable option, its drawbacks will tend to be forgotten as society’s composition changes through births and deaths, or simply entries and exits. This is because new cohorts will not be exposed to the unfiltered private knowledge of previous cohorts but only to the reconstructed knowledge that was considered safe to communicate. The previous cohorts may have disliked the option yet refrained from challenging it. The younger cohorts will preserve it not only because they are afraid to object but also, perhaps mainly, because the impoverished public discourse of the past has blunted their capacity to criticize the inherited social order and imagine better alternatives. Insofar as past preference falsification removes individual inclinations to want something different, current preference falsification will cease to be a source of public opinion’s stability. People will support established arrangements genuinely rather than to make a good impression (Kuran 1995).

A complementary mechanism leading to intellectual impoverishment and the absence of dissent is driven by individual decisions to choose ‘exit’ over ‘voice.’ When people withdraw from a troubled society to avoid the costs of speaking up, the society’s ability to institute reforms diminishes (Hirschman 1970).

The foregoing dissent-dampening mechanisms have universal relevance. Customs, norms, laws, and other institutions that keep societies poor, organizations ineffective, and inequalities unchallenged often owe their stability to the paucity of overt demands for reform. By the same token, other institutions that promote growth, efficiency, and equality are held in place through pressures that diminish both the desire and the willingness to work toward their overthrow. As Tocqueville (1945) observes, these mechanisms even contribute to the stability of traits that give nations and cultures their unique identities.

5. The Imperfect Predictability Of Public Opinion

If public discourse were the only determinant of private knowledge, once a public consensus had taken shape, it would become frozen. In fact, private knowledge has other determinants, including the material consequences of prevailing policies. By sowing doubts about established structures, these other factors can induce changes in private preferences and, ultimately, also in public preferences. However, the unraveling of an established consensus need not track the escalation of private opposition to the status quo. As long as the prevailing social pressures keep preference falsification optimal for all potential dissenters, private opposition will intensify without any change in public opposition. Yet, as the undercurrent of opposition swells, a shock that affects the right people’s thresholds in the right manner will stimulate switches in certain public preferences. Even without further shocks, these switches will then encourage individuals with slightly higher thresholds to join the growing chorus of dissent. Once put in motion, then, open opposition can feed on itself, with each addition generating further additions until much of society publicly identifies with the reformist camp.

Because preference falsification had concealed the opposition brewing under the surface, this revolutionary transformation will not have been anticipated. Even so, the massive change in public opinion may easily be explained with the benefit of hindsight. The revolution’s success will have lowered the personal risk of acknowledging past acts of preference falsification. It will also have created incentives for people who had supported the status quo ante sincerely to pretend that they had simply been waiting for a safe opportunity to speak out for change. Examples of largely unanticipated revolutions include the French Revolution of 1789, the Iranian Revolution of 1978, and the East European Revolutions of 1989. Each of these earth-shaking transformations entailed a massive shift in public opinion that stunned its beneficiaries, victims, and observers.

6. Methodology

A distinction of the choice-based, microsociological analysis of public opinion is that it admits circular relationships generated by feedback effects. For one thing, it links the social pressures that determine the extent to which private and public opinion diverge to the very collective process that shapes public opinion. For another, it recognizes that public opinion shapes the private preferences that are among its own determinants. By virtue of these characteristics, the microsociological approach differs methodologically from social scientific traditions in which either culture or the motives of individuals are fixed. Another salient distinction of this approach lies in its multidimensional conception of individual motives. In recognizing that individuals make tradeoffs between their expressive and reputational needs, it stands out from analytical traditions, including most variants of the public choice school, that treat individuals as invariably sincere (Downs 1957). At the opposite extreme, it stands out from self-presentation theories that view individuals as one-dimensional agents interested only in making a good impression (Goffman 1973).

Nothing in the arguments surveyed here presuppose that certain cultures promote truthfulness and others mendacity. Because people everywhere express themselves with an eye toward balancing diverse objectives, the identified social processes may be observed in any society—economically developed or advanced, democratic or undemocratic, large or small. Where public opinion shows variations across societies—as it often does markedly—the variations do not stem from differences in the underlying basic mechanisms. Rather, they reflect historical contingencies that generated persistent differences in institutions, social pressures, and private knowledge.

Bibliography:

1. Asch S E 1957 Studies of independence and conformity: I. A minority of one against a unanimous majority. Psychological Monographs 70: 416
2. Bikchandani S, Hirshleifer D, Welch I 1992 A theory of fads, fashion, custom, and cultural change as informational cascades. Journal of Political Economy 100: 992–1026
3. Coleman J S 1990 Foundations of Social Theory. Belknap Press, Harvard University Press, Cambridge, MA
4. Downs A 1957 An Economic Theory of Democracy. Harper, New York
5. Goffman E 1973 The Presentation of Self in Everyday Life. Overlook Press, Woodstock, NY
6. Granovetter M 1978 Threshold models of collective behavior. Am. J. Soc. 83: 1420–43
7. Hirschman A O 1970 Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Harvard University Press, Cambridge, MA
8. Kuran T 1995 Private Truths, Public Lies: The Social Consequences of Preference Falsification. Harvard University Press, Cambridge, MA
9. Kuran T, Sunstein C R 1999 Availability cascades and risk regulation. Stanford Law Re . 51: 683–768
10. Noelle-Neumann E 1984 The Spiral of Silence: Public Opinion— Our Social Skin. University of Chicago Press, Chicago
11. Schelling T C 1978 Micromotives and Macro-Behavior. Norton, New York
12. Scott J C 1985 Weapons of the Weak: Everyday Forms of Peasant Resistance. Yale University Press, New Haven, CT
13. de Tocqueville A 1945 Democracy in America. Knopf, New York