History, Philosophy of Public Opinion and Public Opinion Research

[17:13 2/8/2007 4984-Donsbach-Ch01.tex] Paper: a4 Job No: 4984 Donsbach: Public Opinion Research (SAGE Handbook) Page: 7 7–24

PART I

History, Philosophy of PublicOpinion and Public Opinion

Research


Section 1

The Nature of Public Opinion


1The Public and Public Opinion

in Political TheoriesV i n c e n t P r i c e

The origins of our modern conception ofpublic opinion are usually traced to lib-eral democratic theories of the eighteenthcentury, with precursors reaching all theway back to ancient Greece (Palmer, 1936).And yet the connections between empiricalpublic opinion research and political the-ory have been remarkably loose. Despitethe encouragement of leading researcherssuch as Berelson (1952), Lazarsfeld (1957),and Noelle-Neumann (1979), public opinionresearchers have only recently taken up thetask of trying to integrate empirical andphilosophical models (e.g., Herbst, 1993;Price & Neijens, 1997; Althaus, 2006).

This chapter explores some fundamentalconnections between public opinion researchand democratic theories, with several interre-lated aims: (a) illustrating briefly the historicalspan of democratic theories and the widerange of views they adopt with respect to citi-zens, publics, public opinion and governance;(b) considering some of the normative modelsimplicit in public opinion research; and(c) exploring some of the enduring theoreticaltensions, dialectics, and debates that empirical

research might conceivably help to inform,if not resolve.1 In view of a general modelof democracy as collective decision making,this chapter considers the variable sorts ofexpectations democratic theories harbor forpolitical leaders, news media, publics, andcitizens.

ENTWINED CONCEPTS: PUBLIC,OPINION AND DEMOCRACY

The concept of public opinion emerged duringthe Enlightenment, but the separate conceptsof the public and opinion have much olderhistories, each with a range of meanings thatcontinue to inform their use to the presentday (Price, 1992). Opinion was used primarilyin two ways. In an epistemological sense,opinion indicated a particular and to someextent inferior way of knowing, distinguishinga matter of judgment (an “opinion”) from amatter known as fact or asserted on faith. Ina second sense, the term was used to indicateregard, esteem, or reputation (as in holding ahigh opinion of someone). Both senses relate


12 THE SAGE HANDBOOK OF PUBLIC OPINION RESEARCH

to the notion of judgment, though in the onecase the emphasis is on the uncertain truth-value of something believed, whereas in theother the emphasis is on a moral dimensionof judgment, that is, approval or censure.As we shall see, political theories variouslyseize upon one or the other of these sensesof “opinion,” at times emphasizing cognitionand knowledge and at others moral sensibilityor sentiment. The term public, from the Latinpublicus meaning “the people,” similarly hadseveral discernable meanings. In some of itsearliest uses it referred to common access,with areas open to the general populationdeemed public (Habermas, 1962/1989). In asecond usage, public referred to the commoninterest and common good, not in the sense ofaccess (or belonging to) but rather in the senseof representing (that is, in the name of) thewhole of the people. Thus the monarch underthe theory of royal absolutism was the solepublic figure, representing by divine right theentirety of the kingdom in his person (Baker,1990).

The compound concept public opinioncame into widespread use only in theeighteenth century and as the product ofseveral significant historical trends, primarilythe growth of literacy, expansion of themerchant classes, the Protestant Reformation,and the circulation of literature enabled bythe printing press. An ascendant class ofliterate and well-read European merchants,congregating in new popular institutions suchas salons and coffee houses and emboldenedby new liberal philosophies arguing for basicindividual freedoms, began to articulate acritique of royal absolutism and to asserttheir interests in political affairs (Habermas,1962/1989). In early usage, public opinionreferred to the social customs and mannersof this growing class of prosperous “men ofletters” but by the close of the century it wasbeing used in an expressly political context,often in conjunction with cousin phrases suchas “common will,” and “public conscience.”Baker (1990) argues that with the dissolutionof absolute monarchical power, both thecrown and its opponents alike invoked publicopinion as a new source of authority and

legitimacy, largely in rhetorical fashion andwithout any fixed sociological referent. Hencethe term remained, in some sense intention-ally, vague. It was linked quite explicitly withfree and open discussion of political affairsamong educated men of financial means.Yet it often acquired (as in the writings ofRousseau, 1762/1968) an abstract and almostsuper-human quality as an expression ofthe common will, divined through reasoneddebate, and framed as a powerful new tribunalfor checking and thus controlling, as rightwould have it, the actions of the state.

Despite these communitarian origins,however, the concept of public opinion cameto acquire much of its contemporary meaningfrom its deployment in the work of laterliberal thinkers, particularly “utilitarian”philosophers such as Mill (1820/1937) andBentham (1838/1962). While continuingto argue for full publicity of governmentaffairs and strongly advocating freedom ofexpression, these analysts saw the polity lessas the coming together of separate mindsreasoning together toward a shared, commonwill than as a collection of individualsattempting to maximize their own interestsand utilities. The harmonization of theseconflicting interests was best achieved notthrough public reasoning to any consensualconclusion, but instead through rule bymajority, requiring regular election andplebiscite, with the state functioning as areferee to individuals and groups vying toachieve their economic and political ends.“A key proposition,” writes Held (1996,p. 95), “was that the collective good couldbe realized only if individuals interacted incompetitive exchanges pursuing their utilitywith minimal state interference.” Thus publicopinion was wedded to the liberal idea ofan unregulated “marketplace of ideas,” withthe majority view, ascertained through a freepopular vote, as its operational definition.

The early development and use of theconcept of public opinion, then, were part andparcel of the Enlightenment project to replaceEuropean monarchies with civil democra-cies. What the Enlightenment accomplished,according to Peters (1995), was to transform


THE PUBLIC AND PUBLIC OPINION IN POLITICAL THEORIES 13

the classical assembly of the people—inAthenian democracy a physical, face-to-faceforum—into a mass-mediated, fictive bodyconstituted by newspapers bringing peopletogether, not in physical space but in sharedstories and conversations at a distance. “Theimagined public is not, however, imaginary:in acting upon symbolic representations of‘the public’ the public can come to existas a real actor” (p. 16). Implicitly, notionsof the public and public opinion followedthe complete arc of thinking about just whatforms such “imagined assemblies” might take,from highly communitarian formulations ofthe public as a fluid and amorphous groupof freely associating citizens willing to thinkand debate in consideration of the good ofthe whole community, to highly individualistformulations equating it with the mass ofcitizens freely pursuing their personal andgroup interests as they wished, and bymajority vote aggregating those interests tochoose wise political leaders.

NOT ONE, BUT MANY, DEMOCRATICTHEORIES

Despite references to “democratic theory” and“classical democratic theory” that imply somesort of unified conception of democracy, writ-ings on the subject offer myriad competingmodels. Indeed, while democracy is generallyheld to mean “rule by the people,” therehas been historically some dispute over thedefinition of “the people,” and, even moreso, over just what it means for them to“rule” (Lively, 1975). Held’s (1996) reviewidentifies no fewer than a dozen variations.He describes four basic models, appearingroughly in chronological order—fifth-centuryAthenian democracy, with its sovereignassembly of the whole citizenry; republican-ism, from its Roman and Italian Renaissancemanifestations through the Enlightenmentconceptions of Rousseau; eighteenth- andnineteenth-century liberal democracy, withits commitment to individual rights andelectoral representation; and Marxist models

of direct democracy, predicated on completeeconomic and political equality.

These were then supplemented andexpanded by twentieth-century models,drawing in various ways upon all fourbasic formulations but principally from therepublican and liberal traditions (Habermas,1966). Among these are theories Held(1996) names competitive elitism, neo-pluralism, legal democracy and participatorydemocracy. Each in various ways resultedfrom grappling with perceived problems ofthe public in the face of modern industrial life.These perceived ailments of the body politicincluded: a poorly informed and emotionalmass citizenry subject to demagogueryand manipulation; widening inequalities inprivate economic, and hence political, power;expanding centralization of governmentand bureaucratic regulation; a growingand pervasive lack of citizen concern forthe collective welfare; and the politicalwithdrawal of citizens who feel inefficaciousand effectively disenfranchised.

Worry over the emotionality and irra-tionality of ordinary citizens, and a nearcomplete lack of confidence in their abilityto discriminate intelligently among variouspolicies, led some democratic theorists tofear that catering to a “popular will” wouldprove at the least inefficient and at the worstdisastrously unstable, particularly in times ofcultural and political stress. Contemplating acomplex industrial world that had collapsedinto international confusion and warfare,and despairing any hope of wisdom inpopular democracy, both Lippmann (1922)and Schumpeter (1942) argued that anindependent, expert bureaucracy was neededto aid elected representatives in formulatingand administering intelligent public policy,and also that public influence on policymatters should be strictly limited. In makingthe case for a “leadership democracy” or“competitive elitism,” Schumpeter (1942,p. 269) proposed that citizens’ choicesshould extend only to periodic selectionof “the men who are able to do thedeciding.” Lippmann (1922, p. 32) argued thatexpert advisors with unrestricted access to



information should make the “unseen facts”intelligible to political decisions makers and“organize public opinion” for the press andthe citizenry.

Such minimalist conceptions of democracyequate it with any system offering competitiveelections, often placing considerable distancebetween the decisions of governing elitesand the desires of the masses. Pluralistformulations, which became ascendant inAmerican political science in the 1950sand 1960s, accept many of the minimalists’views of citizens but emphasize the role ofintermediary interest groups and quasi-elite“issue publics” in maintaining a competitivebalance of power and providing a critical“linkage” function in tying popular wishes togovernmental decisions. Analysts includingAlmond (1950) and Key (1961) invokedthe concept of “issue publics” (or “specialpublics”) to explain how policy in demo-cratic societies can, despite wide swaths ofinattention and ignorance in the citizenry,nevertheless respond to public opinion in afairly rational manner. It stands to reasonthat, because politics routinely gives way tomore pressing matters of family, work andrecreation, people should focus their attentionon just a few matters of the most direct interestand importance. Nonetheless, for most issuesat least a segment of the population is arousedand interested enough to learn, discuss andform opinions. Issue publics represent thesmall, policy-oriented segments within themass polity that attend to particular problems,engage their political leaders and the mediaover these issues, and demand some degreeof elite responsiveness and accountability.Elections by themselves do not ensure astable and publicly responsive democraticstate; rather, it is a multiplicity of contendingminority interests, which, in pressing theirclaims, are able to bargain for policy accom-modations. Hence modern democracies, atleast those offering relatively open electoralsystems and guarantees of civil liberties thatprotect contending minority interests, are“polyarchies” (Dahl, 1971), where politicalpower is effectively disaggregated and wherespecific policies are—unlike in the elite

model proposed by Schumpeter—anchoredto popular wishes through politically activesegments of the citizenry (Held, 1996).

Pluralist conceptions of disaggregated andin some sense “fairly” distributed powerin society were challenged by many. Somecritics cast the model as elevating a descriptiveaccount of contemporary Western democra-cies to the status of a normative theory, andin so doing enshrining the status quo. Others(e.g., Pateman, 1970) argued that social,financial and political resources, includingknowledge and efficacy, are so maldistributedin the population that many groups in societylack the ability to mobilize. Assumptionsmade by liberal theory that people are “freeand equal,” argued Pateman, do not squarewith actual social and economic disparities,which effectively undermine any formalguarantees of equal rights. True democracyrequires that such inequities be ameliorated,and that the active participation of allsegments of society be fostered in democraticinstitutions of all kinds, which must be fullyopen and publicly accountable (Barber, 1984).

Participatory democratic theorists argue,drawing upon the communitarian notions ofRousseau and other “developmental repub-licans,” that political autonomy arises fromcollective engagement in political actionand discussion. As Dewey (1927, p. 208)had earlier proposed in rebutting Lippman’s(1922) withering attack on citizens, “theessential need [is] improvement in the meth-ods and conditions of debate, discussion,and persuasion.” The problem, many writerssubmit, is that the mass media transformpolitics into a kind of spectator sport. Opinionpolls and popular referenda, despite theirdemocratic aims, merely amplify defectiveopinions formed without any meaningfulpublic debate. The result is a citizenryconverted into a body that consumes politicalviews disseminated by elites through themass media, rather than an autonomous,deliberating body that discovers its ownviews through conversation. The sovereign,reasoning public is displaced by a massaudience assembled around political spectacle(Mills, 1956; Habermas, 1962/1989).



These conditions, however, are not viewedas inevitable. Were people more broadlyempowered, this line of argument runs, theywould become politically transformed: “theywould become more public-spirited, moreknowledgeable, more attentive to the interestsof others, and more probing of their owninterests” (Warren, 1992, p. 8). The actof deliberating, in many treatments (e.g.,Gutmann & Thompson, 1996) is thought tobe especially transformative: it fosters mutualrespect and trust, leads to a heightened senseof one’s value as part of an active politicalcommunity, and stimulates additional foraysinto political engagement. The presumedvalue of discussion in stimulating and engag-ing the citizenry has thus figured heavily inrecent proposals for revitalizing the modernelectorate. Participatory democratic theoryin general and “deliberative democracy”theories in particular have emerged in tandemwith a multi-faceted critique of contemporarysocial and political life (e.g., Fishkin, 1991).

Participatory democratic theory is coun-tered by another contemporary trend in polit-ical philosophy that draws its inspiration notfrom classical republican and communitariannotions but instead from democratic theory’sliberal foundations. Much of the emphasisin liberal democratic theory has to do withdelineating the rights of the citizen against thestate, and balancing and distributing power toavoid its untoward concentration of power inthe hands of any single actor or alignment ofactors. Proponents of legal democracy (e.g.,Hayek, 1979), who are sometimes called neo-liberals, view state efforts to ameliorate socialinequities as inevitably coercive and likelyto come at the expense of individual liberty.In this view, democracy is valuable primarilyin its protection of individual liberty; andthe more expansive the state, the larger itslegislative and bureaucratic reach, the moregrave the dangers to freedom. The potentiallycoercive powers of the state must conse-quently be highly circumscribed by the rule oflaw. The most legitimate means of collectivechoice and—thus the basis for any genuinelyliberal society, legal democrats argue—is thefree-market; and this mechanism operates

best when unencumbered by governmentintervention and regulation (e.g., Friedman,1962).

Twentieth-century models of democracyhave thus moved beyond classical notions ingrappling with ever more complex industrialand corporate societies; yet they continueto range from the highly communitarian tothe highly individualistic in their conceptionsof the public and public opinion, drawingfreely from several centuries of philosophicalinquiry. “Democratic theory is in a state offlux,” writes Held (1996, p. 231), “There arealmost as many differences among thinkerswithin each of the major strands of politicalanalysis as there are among the traditionsthemselves.” As Price (1992, p. 2) hasnoted, connecting the concepts public andopinion represented an attempt by liberaldemocratic philosophy to unite the “one” andthe “many,” to devise ways of producingcoordinated, collective action out of disparateand conflicting individual choices. It did soby turning to the idea of democracy, that is,collective decision making through discussionand debate among members of the citizenry,under conditions of openness and fairness. Yetthe particular mechanisms of decision makingproposed by democratic theorists have alwaysvaried widely.

DEMOCRACY AS COLLECTIVEDECISION MAKING

A useful matrix for conceptualizing thecomplex, temporally extended process ofcollective decision making was proposed byPrice and Neijens (1997). Their matrix servesour particular purposes here by illustrating andsummarizing a very wide range of possiblecollective decision-making processes, andmyriad roles the public might play.

Price and Neijens note general similaritiesbetween traditional models of the stagesthrough which public opinion develops (e.g.,in the work of Bryce, 1888) and the phasesof decision making later adopted by decisionanalysts and policy researchers. Five mainphases of collective decision making can



be distilled. First is the process of elicitingvalues, sometimes called the “problem”stage, which involves recognizing a matterof collective worry or concern, and thenarticulating various goals thought to beimportant in addressing the issue. Next isa phase that involves developing options orproposals for resolving the problem, andsifting these down into a small set ofpotentially viable alternatives. Once thesehave been developed, decision makers turnto estimating consequences of selecting oneover another option, a task that often fallsto technical and policy experts. The fourthstage involves evaluating the alternatives,with advocates of competing options activelyengaged in persuasive appeals aimed atgarnering both public and elite support, andthe issue typically receiving broad mediaattention through news coverage and opinionpolling. This public debate ultimately leadsto the making of a decision, either throughbureaucratic or governmental action or insome cases by electoral choice.

The Price and Neijens decision matrixcrosses each of these five stages with six dif-ferent groups of actors in a democratic societywho may be implicated to varying degrees atany particular phase of the process: political

leaders, technical experts, interest groups, thejournalistic community, attentive publics andmuch larger mass audiences (see Figure 1.1).Political leaders, policy experts and interestgroups comprise the political “elites,” bothwithin and outside the sphere of formalgovernment, who play active roles throughoutall phases of decision making. Members of thepress serve as critical conduits for informationand opinion exchange between these elites,their followers in attentive publics, and muchlarger mass audiences.

Large-scale, democratic choices are espe-cially complicated—due not only to theinteractive engagements of each of thesemyriad groups, but also because the processdoes not necessarily unfold in any neatlylinear fashion. It is often a rather ambiguousand politically-charged affair, far less rationalthan the formal stage-model would imply.While the model suggests that the discovery ofproblems gives rise to solutions, for example,Price and Neijens (1997) note that the entireprocess can be turned on its head wheninterest groups or political leaders adhere toideologically favored political “solutions” andmerely lie in wait opportunistically for theright “problems” to which they can readily beapplied to appear on the scene. Despite these

Elicitation ofgoals/ values

Developmentof options

Estimation ofconsequences

Evaluation of options

Decision

Political leaders

Technicalexperts

Interestgroups

Reportersand editors

Attentive publics

Massaudiences

Figure 1.1 The collective decision-making process—matrix of phases and participants.Reproduced from Price and Neijens (1997, p. 342) with permission from Oxford UniversityPress and the World Association for Public Opinion Research

rathi.t

Highlight

rathi.t

Text Box

Au: Please check if the credit line is OK as set.



complications, the matrix helps to summarizethe full range of possible interactions thatmight potentially feed into decisions made bydemocratic states. It also visually reinforcestwo main dimensions underlying differingconceptions of democratic decision making.

Elite/mass relationships

Comparisons of activities across the verticaldimension of the matrix—from politicalleaders and technical policy experts at the topto mass audiences at the bottom—capture therelative degree to which the process is “top-down” or “bottom-up” in nature. At one endof the theoretical spectrum, elite models ofdemocracy propose that collective decisionmaking unfolds best when it is largelytechnocratic, with elected leaders and expertpolicy advisors deciding the relevant courseof collective action and then organizing publicopinion for the masses (a position embracedas noted above by Lippmann, 1922, in view ofwhat he considered irremediable deficienciesin both the public and the press, for whichhe saw little hope). At the other end of thespectrum are models of direct or participatorydemocracy. More communitarian in spirit, assuggested earlier, they advocate a strong andengaged role for ordinary citizens across allphases of the collective decision-making pro-cess (e.g., Pateman, 1970). All seek some sortof “linkage” from top to bottom (or bottomto top); but the degree of looseness of thelinkages desired and the preferred means bywhich they are to be achieved is quite variable.

The nature of mass involvement

Comparisons of activities across the horizon-tal dimension of the matrix—from elicitationof goals and values, to developing options,estimating consequences, evaluating options,and finally deciding a course of action—capture the relative degree to which theprocess attempts to respond to general popularviews about desirable end states, on the onehand, or aims at soliciting far more focusedpublic evaluations of policy alternatives on

the other. A number of democratic theories—while placing most of the burden fordeveloping, debating, and evaluating policyoptions on elite political leaders, experts,and interest groups—nonetheless propose thatordinary citizens should play critical rolesin conveying, if not highly directive viewson specific policies, at least general signalsof popular values and desires. “Minimal”democratic models view periodic selectionand removal of political leaders as a sufficientmeans of public participation. Other theoriesargue for the more regular and substantialinvolvement of ordinary citizens, for instancethrough referenda on specific policy actions(a primary role advocated for public opinionpolls by Gallup & Rae, 1940). Differentpolitical theories, then, seize upon one orthe other of the two traditional senses ofopinion discussed above: Some seek citizens’knowledgeable contributions (their informedpreferences for particular policies), whileothers seek merely to ground elite decisionmaking in popular moral sensibilities or broadjudgments related to a governing regime’soverall success in meeting the citizens’ basicneeds. Some ask the public to think carefullyabout exactly what the government is doing;others are more concerned with leaders’legitimate public standing (“opinion” hereequated with popular regard or reputation).

Various democratic theories, then, placea range of expectations and demands onthe shoulders of citizens. They range fromrelatively top-down or “weak” forms ofdemocracy to bottom-up, “strong” forms(Barber, 1984); and they range from mod-els positing that ordinary citizens are bestconsulted by seeking diffuse judgments ofsatisfaction with elite performance to modelsthat seek much more direct and detailed publicinput on the substance of pressing policyquestions.

Polls as policy referenda

Implicit in contemporary understandings ofpublic opinion and opinion polling, Priceand Neijens (1997) and Althaus (2006)submit, is a particular decision-making model.



Mass audiences enter the process at theevaluation phase, during which time theyfollow elite debate over a limited numberof options and are asked, via polling, toregister opinions as to which they prefer.However, this informal “policy referendum”model can be seen as problematic, evencontradictory (Althaus, 2006). If membersof the mass audience have no engagementin the process until they are asked theiropinions at the evaluation phase, then itplaces quite heavy and perhaps unreasonableburdens on the press to inform their previously(perhaps habitually) unengaged audiences atthis juncture. Even assuming these burdensare met, the capacity for sovereign citizenjudgments may be heavily circumscribed,both because they have at their disposallittle or no knowledge of alternatives thatwere considered and rejected (or indeednot considered) by elites, and because theyare unlikely to fathom the consequences ofvarious options (aside from whatever can begleaned from political contestants as theyattempt to recruit supporters for their side;Price & Neijens, 1997).

The decision-making matrix suggests atleast two potential remedies to this prob-lem, each consistent with a rather differentnormative-theoretical approach to democracy.Despairing of any expectation for intelligentmass contributions at the evaluation stage,one might shift the focus of mass engagementto the very first, problem-oriented phaseof decision making. Citizens may not becompetent to judge the intricacies of policy,this line of reasoning goes, but they maybe fully capable of telling elite decisionmakers what bothers them, what needs policyattention, and what they most desire interms of collective outputs. Such a modelemphasizes public agenda setting over themonitoring of policy alternatives. Alterna-tively, one might propose, as do deliberativetheorists, that ordinary citizens would be fullycapable of rendering intelligent judgments,if only they enjoyed a different communi-cation apparatus for doing so and were nothamstrung by the conventional press and

polling model as presently institutionalized.Hence the “deliberative poll,” which seeks tounite the mass-representative capabilities ofprobability sampling with something very likethe Athenian assembly (Fishkin, 1991), alongwith related notions of citizen juries, shadowassemblies, and the like.

The omnicompetent straw man

A theme running throughout our discussiondeserves to be stated explicitly at thisjuncture. Empirical opinion research in thetwentieth century—though often framed asrebutting classical democratic theory—in factbore out the low expectations of most pre-empirical theorists, documenting the shallowdiffusion of political information across theelectorate, low levels of popular politicalknowledge, and the tendency of mass beliefsystems to exhibit poorly integrated or weakly“constrained” opinions across different issues(Converse, 1964). Contrary to many claimsthat “classical” democratic theory calledfor omnicompetent citizens, however, themajority of social-philosophical writers of theeighteenth and nineteenth centuries largelyeschewed any expectation that many ordinarypeople would bother to spend more thana modest amount of time thinking aboutpolitics and public policy (Pateman, 1970).AsBryce (1888) and others had long suggested,most people, most of the time, are weaklyif at all engaged in political issues ofthe day. Schudson (1998), after examiningmodels of citizenship over the course ofAmerican history, argues that the ideal of aninformed citizen is actually the product ofearly twentieth century progressive thought.So empirical renderings of citizen ignorance,if they indeed undercut a “classical” theory,may actually address a relatively recentone (hypostasizing Lippmann’s critique ofcontemporary American progressive hopes asa critique of “democratic theory”). Althaus(2006) sums up the matter by pointingto two “false starts” in public opinionresearch: the idea that opinion surveys arebest used to assess government policies, and



the idea that popular disinterest in politicsis a grave and unanticipated problem fordemocratic rule.

RESEARCHING COLLECTIVE DECISIONMAKING

It would be difficult at this point to con-clude that empirical public opinion researchhas convincingly overturned any partic-ular democratic theory. It has arguablyhelped, however, to refine various con-cepts, and has at times called certainphilosophical-theoretical assumptions intoquestion. Significant amounts of surveyresearch have accumulated, for example,detailing the nature of mass political engage-ment (e.g., Verba, Schlozman, & Brady,1995) and the diffusion of political infor-mation, (e.g., Delli-Carpini & Keeter, 1996).This work highlights important inequitiesin both knowledge and participation, offersclues as to their origins, and considersvarious ramifications for democratic practice.As Held (1996, chap. 6) recounts, surveyresearch inAmerica and Britain proved centralin early arguments supporting pluralisticdemocratic theory, but also, as it turnedout, provided evidence of the broad socio-economic inequalities and cultural chasms inpolitical resources marshaled by critics of thepluralistic model.

A full, perhaps even a satisfying integrationof empirical opinion research with democratictheory is beyond the scope of this chapter.Still, as a way of concluding our discussion,we can paint in broad strokes a few ofthe key ways empirical studies and demo-cratic theories might profitably inform eachother.

The empirical contours of “opinion”

Price (1992) notes that while some sociol-ogists adopted an organic, discursive modelof public opinion more or less alignedwith republican theory, developments inattitude measurement and survey research

techniques in the 1920s and 1930s deflectedpublic opinion research onto a much moreindividualistic trajectory. This trajectoryhas proved occasionally contentious (mostnotably when Blumer in 1948 attacked thefield for having entirely missed the mark);however, the operational definition of publicopinion as the aggregated attitudes of a pop-ulation gained wide, indeed nearly universalacceptance. At any rate, in pursuing the studyof individual attitudes and opinions over ahalf century, the field has inarguably accu-mulated a considerably refined understandingof both.

Many of the most profound developmentshave been methodological in origin. In theearly days of opinion research, pollsterstended to view instabilities and inaccuraciesin survey responses as mere artifacts ofmeasurement (Sudman & Bradburn, 1974).However, a shift toward more theoreticallyoriented research in opinion measurement,which began in the late 1970s, led toan understanding that many variations insurvey responses were far from random.Over the past few decades, research hastried to develop comprehensive models ofthe way people respond to survey ques-tions, drawing heavily from theories ofcognitive processing (➔Designing effectiveand valid questionnaires; ➔Psychology ofsurvey responses). The clear trend has beento interpret opinion responses, not as self-evidently interpretable, but in light of howrespondents react to wording or contextchanges, how they respond to rhetoricalmanipulations, how they are influenced bysocial perceptions, and how the responsesvary across groups in the population. Therehas also been conceptual clarification of therange of phenomena relevant to opinionexpression, with researchers examining notonly opinions (e.g., preferences related topolicy matters or public officials), but alsobroad underlying values and attitudes, beliefs,perceptions of groups, and the complex rela-tionships among these (➔Different conceptsof opinions and attitudes; ➔Identifying valueclusters in societies).



In any democratic decision-making processone can imagine, the public’s opinions must atsome point be gathered. Empirical researchoffers extensive guidance, far beyond any-thing speculation might offer, on how to ask.However, empirical research does not, in andof itself, offer any guidance on what to ask.That is properly the role of democratic theorywhich, in return for technical guidance, canoffer the field some normative direction—inemphasizing, say, expressions of basic wantsand desires, or demands for elite action onproblems seen as pressing, over the usual“approval voting” on policies of the day(Althaus, 2003).

The Internet may presage another importantdevelopment for public opinion research.Despite continuous methodological improve-ments, survey research has generally con-sisted of randomly sampled, one-on-one,respondent-to-interviewer interactions aimedat extracting pre-coded answers or shortverbal responses. Web-based technologies,however, may now permit randomly con-stituted respondent-with-respondent groupconversations integrating general-populationsurvey methods and focus-group techniques(Price, 2003). The conceptual fit between suchconversations and the phenomenon of publicopinion, itself grounded in popular discussion,renders it theoretically quite appealing (➔Theinternet as a new platform for expressingopinions/a new kind of public sphere).

The empirical contours of“the public”

Although sublimated, the concepts of publicopinion as an emergent product of widespreaddiscussion, and of the public as a dynamicgroup constituted by the give-and-take ofdebate and deliberation, have never beenentirely absent from public opinion research.Early scientific analysts, most prominentlyAllport (1937), found the notion of publicopinion as an emergent product of discussiondifficult to grasp empirically and problematicin a number of respects, and hence came toaccept mass survey data as the only workableempirical rendering of public opinion. Yet the

extent to which general population surveysthemselves render a valid representation ofthe public has been questioned by scholarsof many stripes. Opinions given to pollstersand survey researchers—often unorganized,disconnected, individual responses formedwithout the benefit of any debate—haveindeed been called “pseudo” public opinion(Graber, 1982).

These debates echo enduring republi-can/liberal tensions in democratic theory,which has variously cast “the public” as oneor another of any number of sociologicalentities: a complex of groups pressing forpolitical action (i.e., interest groups); peopleengaged in debate over some issue; peoplewho have thought about an issue and knowenough to form opinions (whether or notthey have been engaged in conversation ordebate); groups of people who are followingsome issue in the media (i.e., audiencesor attention aggregates); an electorate; anagglomeration of all citizens; the generalpopulation of some geopolitical entity; or evensome imagined community in the minds ofcitizens. These varying conceptions impli-cate a number of empirical phenomena—conversations, the holding of opinions, mediause, knowledge, participation, the perceivedclimate of opinion—as criterial attributes.And each of these phenomena has beenstudied, some of them quite extensively, inempirical research. In one way or another,normative theories will only make contactwith public opinion research if we are able tofind the public (or publics) as conceptualizedin theory.

The study of public knowledge serves asa case in point, one that drives directly atissues of rationality and equity, and indirectlyat how we define the public. Suppose wedismiss general-population survey results asexpressing, not true public opinion, butinstead rather thoughtless, lightly rooted “top-of-the-head” reactions to some issue. Howwould our reading of public opinion look if weconfined “the public” to only knowledgeablecitizens? Would it render a substantiallydifferent portrait of public preferences? Per-haps surprisingly, Page and Shapiro (1992)



argue “no.” Despite the relative incoherenceof many sampled opinions, when surveydata are aggregated they reveal essentiallyrational collective preferences, since mostof the thoughtless “noise,” the flotsam andjetsam of mass pseudo opinions, ends upcanceling out. Such collective rationalityis reassuring to pollsters; however, it doesnot necessarily solve the problems arisingfrom a large number of uninformed votersin the population (Delli-Carpini & Keeter,1996; ➔Studying elites vs. mass opinion).Recently Althaus (2003) has demonstratedthat, at least on some issues, systematicinequalities in knowledge distribution amonggroups in the population can distort evenaggregate readings of public opinion. Andbecause political knowledge is a resource(just like financial capital) that underwritesparticipation and facilitates mobilization, theimplications of its distribution in societyextend far beyond the impact on pollingresults. In pluralistic formulations of demo-cratic decision making, government policiesare linked to mass preferences throughrepresentative issue publics. Although theymay vary in size and composition from issueto issue (Krosnick, 1990), issue publics maybe drawn disproportionately from a generallywell-educated, attentive and knowledgeablestratum of the population (at best one-fifth ofthe electorate at large, by most methods ofaccounting; see e.g., Delli-Carpini & Keeter,1996). These are not just empirical linesof inquiry; they take on deep theoreticalmeaning when viewed through the prismof one or another model of democraticdecision making. The public is a complexblending of “active” and “passive” segments,of “engaged” citizens and mere “spectators.”The size and representative composition ofthese segments, which surely changes acrossissues and over time, indexes in many waysthe health of a democracy.

The empirical contours of“the citizen”

Implicit in any model of democracy is amodel of the citizen: a set of assumptions

about what motivates him or her, about hercognitive capacities, about his behavioraltendencies. Here again we find significantopportunities for empirical research anddemocratic theory to inform one another, withthe latter proposing what to look for, and theformer serving to refine and correct theoreticalassumptions.

Fundamental to the project of understand-ing citizens is some recognition that they are,in large part, products of their surroundingpolitical culture. Consequently, understand-ing them requires two tasks: learning howthey are at present,and learning how, underdifferent conditions, they might be. A fittingillustration is provided by participatory demo-cratic theory, developed as it was withthe understanding that many citizens arepoorly informed, politically apathetic andinefficacious, but also in the belief that thesevery people could be transformed througheveryday democratic praxis into differentand more productive citizens.In its delibera-tive variant, this theoretical model proposesthat public discussion serves to broadenpublic perspectives, promote tolerance andunderstanding between groups with divergentinterests, and generally encourage a public-spirited attitude.

Advocates of deliberative theory arepresently legion, but its fundamental propo-sitions are not without critics (e.g., Hibbing &Theiss-Morse, 2002), and they have beenincreasingly subjected to empirical scrutiny(e.g., Fishkin & Luskin, 1999; Mutz, 2006;Price, in press). Group discussion has, afterall, been known to produce opinion polar-ization, shifts in new and risky directions,and other undesired outcomes. Disagreementmay also be fundamentally uncomfortablefor citizens, particularly those uncertain oftheir views and feeling ill-equipped to defendthem. Some have argued that encouragingcitizen discussion, despite its democraticintentions, will make reaching out to thedisenfranchised, who tend to lack statusand deliberative ability, even more difficult(Sanders, 1997). As deliberative theory isplayed out in actual practice and as empiricalresearch accumulates, we should come to



better understand conditions of discussion thatfacilitate or retard democratic aims. Com-parisons of citizen behavior across differentcontexts—local, national, and international—should also prove highly informative.

Empirical monitoring of collectivedecision making

There is another, perhaps even more importantway in which public opinion research anddemocratic theory should intersect. Some50 years ago, Hyman (1957) pointed outthat opinion research tended to pursue,using sociologist Robert Merton’s phrase,“theories of the middle range.” While thisstrategy stood to produce useful and valuablepsychological insights, Hyman opined, ithad potential liabilities as well. “We mayconcentrate on the trivial rather than theimportant,” Hyman worried, “We may eveninstitutionalize the neglect of some importantpart of our ultimate larger theory” (p. 56).What was needed to avoid these problems, hesuggested, was careful monitoring of large-scale social processes over time, with afocus on the relationship of popular think-ing to governmental processes and policyoutcomes.

In the terms adopted here, Hyman’s callis for the monitoring over time of key cellsin the decision-making matrix, as collectiveproblems are first identified and addressed,and as decisions work their way throughprocesses of social and political negotiation.Attention would be paid to the goals andinterests of each of the participants identifiedby the matrix, with the aim of determininghow—and indeed if—democratic mass–elitelinkages occur. This is admittedly a tallorder to fill. Yet here again, empirical publicopinion research has been evolving in thisdirection, albeit not always with the explicitconnections to democratic theories that itmight have marshaled. Research on agendasetting, for example, though very oftentethered to “middle-range” theoretical goals,has at times turned to big-picture questionsand produced interesting examinations overtime of the complex interactions of public,

press, and policy agendas. Some exemplaryworks in this tradition include the “agenda-building” research of Lang and Lang (1983),who examined the ways in which Watergatedeveloped as a public issue through persistentelite efforts, constrained by political eventsand contemporary currents in mass opinion;or the series of detailed case studies conductedby Protess and colleagues (1991), who stud-ied the ways investigative journalists oftencollaborate with public policy makers to seta “public” reform agenda, in some instancesapparently without much engagement ofattentive publics or mass audiences.

Price and Neijens (1997) suggest a largenumber of collective “decision-quality” con-cerns that might be empirically examinedin opinion research. These sorts of qualitycriteria—for example, the extent to whichthe problems addressed appear responsiveto popular concerns, the extensiveness ofpopular discussion and debate, the degreeto which those who are engaged representthe affected population, the generation ofdiffering viewpoints on the problem at hand,the degree to which the consequences ofchosen policies are clearly understood bythe public, or the degree to which theprocess is perceived as fair and legitimate—all have import for the democratic characterof the public opinion and policy making.Democratic theories construct various modelsof the way decision making ought to unfold,but empirical research is required to informjudgments about the way they actually unfoldin practice.

This brief overview has necessarily takena rather broad sweep at identifying some ofthe major lines of normative theoretical think-ing that feed into modern opinion research,suggesting just a few of the ways empiricaland philosophical inquiry might inform oneanother. Readers are encouraged, as theyconsider the many lines of study summarizedelsewhere in this volume, to look for otheruseful connections to democratic theory notexplored here. Finding and nurturing thoseconnections should help an already vibrantfield of research to become even morefruitful.



NOTE

1 Our review must of necessity be brief. Thoroughreviews of political theory can be found, for instance,in Held (1996). For a review of the history of theconcept of public opinion see Price (1992), Peters(1995), or Splichal (1999).

REFERENCES

Allport, F. H. (1937). Toward a science of public opinion.Public Opinion Quarterly, 1, 7–23.

Almond, G. (1950). The American people and foreignpolicy. New York: Harcourt.

Althaus, S. L. (2003). Collective preferences indemocratic politics: Opinion surveys and the willof the people. New York: Cambridge UniversityPress.

Althaus, S. L. (2006). False starts, dead ends, andnew opportunities in public opinion research. CriticalReview, 18, 75–104.

Baker, K. M. (1990). Public opinion as political invention.In K. M. Baker (Ed.), Inventing the French Revolution:Essays on French political culture in the eighteenthcentury (pp. 167–199). Cambridge, UK: CambridgeUniversity Press.

Barber, B. (1984). Strong democracy: Participatorypolitics for a new age. Berkeley: University ofCalifornia Press.

Bentham, J. (1962). The works of Jeremy Bentham(J. Browning, Ed.) (Vols. 1–11). New York: Russell &Russell (Original work published 1838–1843).

Berelson, B. (1952). Democratic theory and publicopinion. Public Opinion Quarterly, 16, 313–330.

Blumer, H. (1948). Public opinion and public opinionpolling. American Sociological Review, 13, 542–554.

Bryce, J. (1888). The American commonwealth (Vol. 3).London: Macmillan.

Converse, P. E. (1964). The nature of belief systemsin mass publics. In D. E. Apter (Ed.), Ideology anddiscontent (pp. 206–261). New York: Free Press.

Dahl, R. A. (1971). Polyarchy: Participation andopposition. New Haven: Yale University Press.

Delli-Carpini, M. X., & Keeter, S. (1996). What Americansknow about politics and why it matters. New Haven:Yale University Press.

Dewey, J. (1927). The public and its problems. New York:Holt, Rinehart & Winston.

Fishkin, J. S. (1991). Democracy and deliberation: Newdirections for democratic reform. New Haven: YaleUniversity Press.

Fishkin, J. S., & Luskin, R. C. (1999). Bringing deliber-ation to the democratic dialogue. In M. McCombs &

A. Reynolds (Eds.), The poll with a human face:The national issues convention experiment inpolitical communication. Mahwah, NJ: LawrenceErlbaum.

Friedman, M. (1962). Capitalism and freedom. Chicago:University of Chicago Press.

Gallup. G., & Rae, S. (1940). The pulse of democracy.New York: Simon & Schuster.

Graber, D. A. (1982). The impact of media research onpublic opinion studies. In D. C. Whitney, E. Wartella, &S. Windahl (Eds.), Mass communication reviewyearbook (Vol. 3, pp. 555–564). Newbury Park, CA:Sage.

Gutmann, A., & Thompson, D. (1996). Democracy anddisagreement. Cambridge, MA: Harvard UniversityPress.

Habermas, J. (1966). Three normative models ofdemocracy. In S. Benhabib (Ed.), Democracy anddifference: Testing the boundaries of the political(pp. 21–30). Princeton, NJ: Princeton UniversityPress.

Habermas, J. (1989). The structural transformation of thepublic sphere: An inquiry into a category of bourgeoissociety (T. Burger, Trans.). Cambridge, MA: MIT Press(Original work published 1962).

Hayek, F. A. von (1979). The political order of a freepeople. Chicago: University of Chicago Press.

Held, D. (1996). Models of democracy (2nd ed.).Stanford, CA: Stanford University Press.

Herbst, S. (1993). Numbered voices: How opinion pollinghas shaped American politics. Chicago: University ofChicago Press.

Hibbing, J. R., & Theiss-Morse, E. (2002). Stealth democ-racy: American’s beliefs about how governmentshould work. Cambridge, UK: Cambridge UniversityPress.

Hyman, H. H. (1957). Toward a theory of public opinion.Public Opinion Quarterly, 21, 54–60.

Key, V. O., Jr. (1961). Public opinion and Americandemocracy. New York: Knopf.

Krosnick, J. A. (1990). Government policy and citizenpassion: A study of issue publics in contemporaryAmerica. Political Behavior, 12, 59–92.

Lang, G. E., & Lang, K. (1983). The battle for publicopinion: The president, the press, and the polls duringWatergate. New York: Columbia University Press.

Lazarsfeld, P. F. (1957). Public opinion and the classicaltradition. Public Opinion Quarterly, 21, 39–53.

Lippmann, W. (1922). Public opinion. New York:Harcourt Brace Jovanovich.

Lively, J. (1975). Democracy. Oxford, UK: Blackwell.Mill, J. (1937). An essay on government. Cambridge, UK:

Cambridge University Press (Original work published1820).



Mills, C. W. (1956). The power elite. Oxford, UK: OxfordUniversity Press.

Mutz, D. C. (2006). Hearing the other side: Deliberativeversus participatory democracy. Cambridge, UK:Cambridge University Press.

Noelle-Neumann, E. (1979). Public opinion and theclassical tradition. Public Opinion Quarterly, 43,143–156.

Page, B. I., & Shapiro, R. Y. (1992). The rational public:Fifty years of trends in Americans’ policy preferences.Chicago: University of Chicago Press.

Palmer, P. A. (1936). Public opinion in politicaltheory. In C. Wittke (Ed.), Essays in history andpolitical theory: In honor of Charles Howard McIlwain(pp. 230–257). Cambridge, MA: Harvard UniversityPress.

Pateman, C. (1970). Participation and democratictheory. London: Cambridge University Press.

Peters, J. D. (1995). Historical tensions in the concept ofpublic opinion. In T. L. Glasser & C. T. Salmon (Eds.),Public opinion and the communication of consent(pp. 3–32). New York: Guilford.

Price, V. (1992). Public opinion. Newbury Park, CA:Sage.

Price, V., & Neijens, P. (1997). Opinion quality in publicopinion research. International Journal of PublicOpinion Research, 9, 336–360.

Price, V. (2003, September). Conversations at random:New possibilities for studying public opinion online.Invited paper presented to the Innovative ResearchMethodologies Symposium, New Research for New

Media, Institute for New Media Studies, University ofMinnesota.

Price, V. (in press). Citizens deliberating online:Theory and some evidence. In T. Davies &E. Noveck (Eds.). Online deliberation: Design,research, and practice. Chicago: University of ChicagoPress.

Protess, D. L., Doppelt, J. C., Ettema, J. S., Gordon, M. T.,Cook, F. L., Leff, D. R. (1991). The journalism ofoutrage: Investigative reporting and agenda buildingin America. New York: Guilford Press.

Rousseau, J. J. (1968). The social contract (M. Cranston,Trans.). Hammondsworth, UK: Penguin (Originalwork published 1762).

Sanders, L. M. (1997). Against deliberation. PoliticalTheory, 25, 347–376.

Schudson, M. (1998). The good citizen: A history ofAmerican civic life. New York: Free Press.

Schumpeter, J. A. (1942). Capitalism, socialism anddemocracy. New York: Harper and Brothers.

Splichal, S. (1999). Public opinion: Developments andcontroversies in the twentieth century. Lanham, MD:Rowman and Littlefield.

Sudman, S., & Bradburn, N. (1974). Response effects insurveys. Chicago: Aldine.

Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voiceand equality: Civic voluntarism in American politics.Cambridge, MA: Harvard University Press.

Warren, M. (1992). Democratic theory and self-transformation. American Political Science Review,86, 8–23.

History, Philosophy of Public Opinion and Public Opinion Research

Documents

Transcript of History, Philosophy of Public Opinion and Public Opinion Research