The Problem of the Status Quo Approach to Public Opinion
What is public opinion? Seems an easy question, does it not? The immediate response, whether one asked a member of the public or political scientist, would be that public opinion is simply how the entire population of a polity feel about a given issue. If the issue is, say, gun control, then public opinion is the aggregation of every single for and against opinion that exists in the entire country. So, if we could ask every person “how do you feel about gun control?,” public opinion would be the aggregation of their individual responses. That is the de facto conceptualization, and the measurement strategy for such a conceptualization is obvious. You just need to ask people about issues, record their responses, and aggregate them. Asking the entire polity is obviously difficult, so you do the next best thing—you get a random, representative sample and get their responses, and then use that to infer to the entire population.
We see evidence of this measurement strategy (and the conceptualization behind it, which is almost always unstated) all around us within both political science and the political world in general. In the United States, the General Social Survey and American National Election Studies both provide examples of public opinion in this format, and this has become the de facto practice internationally. A survey tool I often use in my own work, the Arab Barometer, is a prime example. Regarded as the gold standard for public opinion work in the MENA region, the Arab Barometer uses this same conceptualization and measurement strategy detailed above. I receive constant email updates about “What Iraqis think” about a given topic, and I can presumably be confident in knowing that I have cornered the market on public opinion in Iraq because they have followed a statistically-sound sampling strategy. I can therefore infer that their results represent public opinion in Iraq.
This approach, which has become ubiquitous, is certainly impressive in its methodological utility, but presents a view on what public opinion is which is rather divorced from scholarship on public opinion from an historical standpoint. I believe many of these neglected works on public opinion present a far richer conceptualization of public opinion than this more recent Gallupean tradition. Beyond cutting us off from previous literature, why exactly is the status quo conceptualization of public opinion bad? Well, there are two other reasons: explicitness and representation.
Explicitness
What I mean by the problem of explicitness is that this status quo conceptualization of public opinion is almost always unstated and has to be assumed from the measurement strategy employed. Public opinion, in this case, is usually an applied term—public opinion about something, some issue. But what is public opinion in the first place? What does it consist of? That more fundamental definition is never put forward. And that’s a problem because it prevents us from reaching a true ‘systematized concept’ of public opinion, to use Adcock and Collier’s (2001) term; a systematized concept being a specific formulation of a concept used by a given scholar (or preferably) a group of scholars (or entire discipline) which involves an explicit definition. As Adcock and Collier posit, without a systematized concept to work from, it is difficult to make conceptual improvements, let alone find relevant indicators for that concept. To put it more baldly, I think a lack of explicitness in the conceptual definition of public opinion is hindering the accumulation of knowledge that science relies upon to advance in a given field. If everybody shrugs off the definitional part of public opinion, how can we know we agree about the fundamentals of what public opinion is?
Representation
The status quo conceptualization (from what we can infer given the fact that few make the conceptualization explicit) is also weak because it is unrepresentative. It is not an accurate representation of the abstract concept of public opinion; it does not faithfully tap into its general nature and properties. As I sketched out earlier, it relies on the idea that public opinion is simply the aggregation of every opinion that exists in the mind of every citizen in the polity. But that is a very limited, unrepresentative conceptualization. It ignores the role of elites and of opinion leaders. It ignores the fact that some opinions clearly matter more to the general public than others, that all opinions are not created equal. I suspect most people would agree that the voice of, say, Barack Obama carries more weight in the public sphere on political issues than a random citizen, and yet the status quo approach treats them the same. That may be normatively appealing from an egalitarian standpoint, but it is conceptually unrepresentative; it does not resemble reality (such as it is for such an abstract construction). As Herbert Blumer puts it (1948, pp. 543-546): public opinion depends upon the society in which it is embedded, and that society has an organization; it is not just an aggregation of disparate individuals. As part of that organization are institutions where public opinion can be formed and spread, as well as elites who lie at strategic points of public opinion formation. Collecting opinion only by survey totally ignores that structure, those institutions, and the general clash of groups, elites, and the media.
My Conceptualization of Public Opinion
Procedural
I will lean most heavily on arguments taken from Bryce (1910) as a starting point. His work helps us understand that public opinion is best understood as a process, always in flux, and difficult to pin down by nature of its constant interaction with itself. Bryce himself articulated three general stages (pp. 251-252). In the first, individuals are exposed to, or experience, some political thought or event. In the second, which develops in concert with other people, attitudes and reactions begin to develop from that first event or thought. And in the third, more reactions occur from those initial attitudes or stances taken. And so on down the line—from that point on, actors within the polity are reacting to what they perceive others believe about an event. Lazarsfeld (1957, p.47) also articulates something similar with his description of the ‘feedback effect’. Here, however, it is the general citizenry who are reacting to the behaviour of elites. The example Lazarsfeld gives is of public opinion about a law being more favourable just by virtue of its passage and recognition by Congress. But we could imagine a Congress thinking specifically about what the general public will believe about a law too; there is a symbiosis. Relying only on survey data cuts us off from seeing this process or procedure very easily since it captures only a moment in time (although this is mitigated somewhat by longitudinal survey data).
Structural
Bryce also introduces us to the idea that, as part of this interaction, the structure of one’s networks and society matter. If public opinion is made up out of the interaction between people, that does not mean that people are always interacting on a level playing field. Indeed, as Bryce suggests (1891, p.258) the makeup of one’s peer group, the opinions of leaders, and the opinions of the press are all going to pull public opinion in one way or another. There are those who create public opinion and the rest of us who mostly react to it, but even reaction is important. Bryce describes a cycle of “action and reaction” where makers of opinion both react to, and try to anticipate how they think the general public will react (1891, p.258).
Majoritarian
The majoritarian definitional category as defined by Susan Herbst (1993, p.45) could be added here to enrich Bryce’s structural description above. That is, that when it comes to public opinion, the majority opinion matters more than the minority opinion. This is intuitive in a democracy, because we would expect political action to be more likely to follow the will of the majority. But it also fits in with the “action and reaction” description Bryce gives because actors are often reacting to what they perceive the will of the majority of people to be, precisely because they believe they believe that the majority will dictate political action by virtue of their numbers in a democracy.
Norms & Consensus
Although Herbst (1993, p.45) discusses this as an additional category, her description of consensual public opinion could also fit in here as a sort of super-majoritarian principle. When enough people in a society have come to some sort of decision (whether explicit but more likely implicit) about something, a social norm forms that determines political behaviour. Flouting that social norm brings the risk of social opprobrium. The work of Gamson (in Crigler, 1996) is illustrative here. In capturing the conversations of ordinary Americans, Gamson found that “every group on every issue in our study showed some awareness that there was a public discourse around them" (p.115). And more than just awareness of what that discourse was, Gamson found that participants in his study often referred to a shared knowledge of what ‘everyone knows’ (p.120). This is a common phenomenon in everyday life. You do not need to conduct a poll about the permissibility of racial slurs, it’s easily seen in the norms of society by its absence (and the uproar when one is used). ‘Everyone knows’ that using such language is outside the bounds of acceptable social behaviour.
Media
As a final valuable contribution, Bryce highlights the role of the press in forming and changing public opinion (1891, p.258). In his day, those were newspapers, and Bryce noted how newspapers’ control of the agenda of what political events to report helped distill public opinion. He then makes the curious claim, however, that newspapers mainly distribute the news rather than creating views, but I would argue that agenda control implies that they are significantly affecting (if not necessarily creating) public opinion. If the media does not report that a tree falls in the forest (the stage one event), there can be no interaction as Bryce describes in stage two and three. What enters the sphere of public dialogue is crucial, and the media has traditionally been a key part in that.
The importance of the media in public opinion formation has been shown empirically by Gamson (1996). By analyzing conversations citizens have about politics, Gamson found that participants used the media to spotlight facts, discuss public figures, and frequently adopted popular catchphrases (p.119). This is apart from the agenda setting function and priming that I have mentioned previously and for which a whole literature exists. To use the most famous example, we could look to Iyengar, Peters, and Kinder (1982). Their experimental treatment was edited news stories that emphasized either inadequacies about American defense, pollution, or inflation. Participants came to Yale to watch the television stories in small groups, and were told they were watching the previous night’s news. So half received the news with added segments and half received the news in its original context. Those exposed to the treatment news conditions thought those issues were more important, thus supporting their hypothesis about the importance of agenda setting. They find less conclusive, although still confirmatory, evidence about priming. When reacting to the agenda setting of the experimental treatment, participants were primed to evaluate President Carter mostly on defense. That became the most important criteria for judgment. So, there is certainly compelling evidence that the media is an important pillar of creating and dispersing public opinion.
Of course, the media environment today is far different from when Gamson was writing in the 1990s. The internet, and especially social media, has fragmented the traditional news media almost beyond recognition. The same agenda setting power and priming power exists, however, albeit in a more diluted form.
Summing Up
Based on the five dimensions discussed, I would define public opinion as the product of the interaction of political actors, including ordinary citizens, in a polity. That interaction occurs within the hierarchical information flows that exist in society, with elites, social groups, and the media playing a role in generating and shaping public opinion. Within a democratic society, a majoritarian public opinion that coalesces around some issue is the most important, as it is the most likely to lead to political action, and therefore spurs the most discussion, contestation, and general interaction. And at its extreme end, this majoritarian public opinion can form into norms of behaviour, which are powerful inasmuch as they constrain individuals to conform to said behaviour or face social censure.
All this being said, a generation of ubiquitous polls presented by social science and the media has also indoctrinated both scholars and the general public into believing that public opinion is an aggregated concept. And inasmuch as public opinion is to an extent a shared social construct, there is certainly a place for incorporating in a belief of individualized aggregated opinions, as long as there is a realization that this cannot stand alone. There is something to be said also about not divorcing a concept too dramatically from its common usage and understanding, however flawed. What I am proposing, therefore, is not to abandon the aggregation aspect of public opinion’s conceptualization, but rather to enrich it with these additional components.
Public opinion as the product of the interaction of political actors in a polity may still feel too vague. What does this product look like? What are its components? It is difficult to answer such a question without resorting to generalities, as difficult as answering the question of how to describe a microcosm of public opinion—the mood of a crowd. You can feel the mood, you know everyone else can sense that same mood, you all know it exists, but yet it remains hard to describe exactly what that mood consists of. Extending this analogy to try and help us understand this conceptualization in a more useful way, imagine that public opinion is the mood of a country. Just like in a crowd, the mood is created by the interaction of those within the crowd. And just like in a crowd, some voices matter more than others. The voice of the people addressing the crowd matter more than the random murmurs of approval or disapproval from people in the back. And like a crowd, the reaction of the greater part of the crowd will set the mood more than a smaller minority, but even a smaller minority will affect the mood if they happen to have greater leverage to make their voices heard.
Works Cited
Adcock, R., & Collier, D. (2001). Measurement Validity: A Shared Standard for Qualitative and Quantitative Research. American Political Science Review, 95(3), 529–546. https://doi.org/10.1017/S0003055401003100
Blumer, H. (1948). Public Opinion and Public Opinion Polling. American Sociological Review, 13(5), 542–546. https://doi.org/10.2307/2087146
Bryce, J. B. (1910). The American commonwealth (Abridged ed. for the use of colleges and high schools.). Macmillan Co.
Gamson, W. A. (1996). Media discourse as a framing resource. In A. N. Crigler (Ed.), The psychology of political communication (pp. 111–132). University of Michigan Press.
Herbst, S. (1993). Numbered voices: How opinion polling has shaped American politics. University of Chicago Press.
Iyengar, S., Peters, M. D., & Kinder, D. R. (1982). Experimental Demonstrations of the “Not-So-Minimal” Consequences of Television News Programs. The American Political Science Review, 76(4), 848–858. https://doi.org/10.2307/1962976
Lazarsfeld, P. F. (1957). Public opinion and the classical tradition. Public Opinion Quarterly, 21(1), 39–53.