Chapter 6 Key Takeaways

The Central Question

Does public opinion exist as a coherent entity that surveys discover? The answer is nuanced: public opinion is a social construct that is real in its consequences, but not a fixed, pre-existing quantity that measurement simply reveals.

Core Theoretical Frameworks

Converse's Non-Attitudes (1964)

  • Many survey responses, especially on unfamiliar policy topics, are non-attitudes — fabricated on the spot with no stable underlying content
  • Ideological constraint (the logical coherence of positions across issues) is rare in mass publics
  • Panel data shows many respondents give randomly different answers to the same question across survey waves
  • True ideologues constitute only about 2-4% of the electorate

Zaller's RAS Model (1992)

  • Receive: Citizens vary in how much political communication they receive (higher for engaged citizens)
  • Accept: Citizens filter communication through predispositions (resist inconsistent arguments)
  • Sample: When answering, citizens sample from currently accessible considerations
  • Key implication: opinion is constructed at the moment of answering, not retrieved from a fixed storage
  • Context, framing, and recent events shape what considerations are accessible — and therefore what opinion is expressed

The Thermostatic Model (Wlezien/Erikson)

  • Public opinion reacts against the direction of policy change — like a thermostat
  • When government does more of X, public demand for more X decreases; when government does less, demand increases
  • Opinion is not a fixed backdrop but a dynamic feedback signal to government action
  • Evidence: ACA support spiked when repeal was threatened; defense spending opinion tracks actual defense spending levels

The Measurement Construction Problem

  • Survey questions do not discover pre-existing opinions — they participate in constructing the opinions they measure
  • "Welfare" vs. "assistance to the poor": 10-20 point gap in support for the same policy
  • Different framings activate different considerations, producing different expressed opinions
  • No framing is fully neutral; every question embeds assumptions

Aggregation Problems

  • Any aggregate "public opinion" number embeds contested choices about whose voice counts
  • Likely voter polls vs. all-adult polls produce different results with different political implications
  • Arrow's Impossibility Theorem: no aggregation method satisfies all reasonable democratic criteria simultaneously
  • Transparency about aggregation choices is an ethical, not just methodological, requirement

Social Distortions of Expressed Opinion

Phenomenon Mechanism Political Example
Social desirability bias Respondents answer to look good Overstating voter turnout
Spiral of silence Minority opinion suppressed by perceived majority Under-reporting support for stigmatized candidates
Acquiescence bias Tendency to agree with any statement Yes-saying to leading questions

A Working Theory of Public Opinion for Analysts

  1. Opinions are distributions, not fixed points — any individual's "true" position is a probability cloud, not a location
  2. Context is data — question wording, order, and mode are not contaminants; they are part of what you measure
  3. Stability varies — partisan identity is stable; specific policy positions on unfamiliar issues are volatile
  4. Elite cues drive change — look for elite communication shifts when you see rapid opinion change
  5. Aggregation is political — be transparent about whose voices your sample captures
  6. Silence is data — suppressed opinions are part of the landscape you need to account for

Key Scholars and Works

  • Walter Lippmann — Public Opinion (1922): the pseudo-environment
  • Philip Converse — "The Nature of Belief Systems in Mass Publics" (1964): non-attitudes, ideological constraint
  • John Zaller — The Nature and Origins of Mass Opinion (1992): RAS model
  • Christopher Wlezien / Robert Erikson — Thermostatic model of opinion and policy
  • Elisabeth Noelle-Neumann — Spiral of silence (1974)
  • Adam Berinsky — Non-attitudes and political knowledge

The Bottom Line

Public opinion is a social construct that is real in its consequences — it shapes campaigns, policies, and political careers — but it is not a fact about the world the way a physical measurement is. It is an approximation, always context-dependent, always incomplete, always shaped by the instruments of measurement. The best political analysts hold both truths simultaneously: the numbers matter, and the numbers are constructions.

"This number here — 47% support. That's not a fact about the world the way the speed of light is a fact about the world. It's a fact about how a particular population of people responded to a particular question, asked in a particular way, at a particular time." — Dr. Vivian Park