Skip to main content
v2026.11,986 entries · CC-BY 4.0

Spiral of Silence Theory: Evidence, Measurement and Criticisms

Noelle-Neumann’s 1974 spiral of silence theory: how it is measured, what Pew’s 2014 survey and a 66-study meta-analysis found, and the main criticisms.

Written and maintained by CASRAI Editorial Board

Published

On this page

The spiral of silence theory, proposed by the German political scientist Elisabeth Noelle-Neumann in 1974, holds that people who think their opinion is in the minority tend to stay quiet, while people who think theirs is shared speak up. Because the quiet side becomes less visible and the vocal side more visible, the apparent balance of opinion tilts further, and the cycle repeats. For researchers it is a testable claim about the link between perceived opinion climate and willingness to express an opinion, and it has been tested for five decades.

This guide treats the theory as a research object: what was claimed, how it has been measured, and what two large bodies of evidence (a national Pew survey and a 66-study meta-analysis) show. For the field it belongs to, see our overview of what communication studies is.

What is spiral of silence theory?

Spiral of silence theory is a model of public opinion formation. It says that individuals continuously gauge which views are gaining or losing support, and that they adjust whether to voice their own view according to that perception. People who see their view as widely shared speak with confidence; people who see it as unpopular become reserved. The result is that the majority position looks louder, and the minority position weaker, than the underlying distribution of opinion may warrant (Wikipedia, “Spiral of silence”, summarising the literature).

The theory was published as Noelle-Neumann, E. (1974), “The Spiral of Silence: A Theory of Public Opinion”, Journal of Communication 24(2), 43-51 (DOI 10.1111/j.1460-2466.1974.tb00367.x, confirmed in the Crossref record). Pew Research Center’s 2014 report cites that article as the origin of the idea and describes the tendency as people not speaking up about policy issues in public when they think their view is not widely shared.

Which scholar is associated with the spiral of silence theory?

Elisabeth Noelle-Neumann. Secondary summaries describe the theory as growing out of her work on West German federal elections in 1965 and 1972, where a small share of voters appeared to shift toward the perceived winner late in the campaign.

How the mechanism works

The literature describes the mechanism in four linked parts. Each is a separate thing a study can measure, which is why the theory has produced so many different research designs.

  • Perception of the opinion climate. Noelle-Neumann called people’s unconscious reading of where opinion stands a “quasi-statistical organ”. Studies operationalise it as a respondent’s estimate of how many others agree with them, or of which side is gaining.
  • Fear of isolation. The motive usually named is fear of being isolated from others for holding an unpopular view. Whether this is the real driver is contested: other proposed explanations include political interest, self-efficacy, knowledge, bandwagon effects and doubt.
  • Willingness to speak. The outcome variable. It can be measured as a stated willingness, a hypothetical-situation response, or actual expression such as a social media post.
  • The media. Noelle-Neumann treated the mass media as central to what people believe the dominant opinion is, whether or not the portrayal is accurate. Some scholars argue that the immediate social micro-climate matters more than the media.

The theory also names groups that do not follow the pattern. The “hard core” are nonconformists already rejected for their beliefs who have little left to lose by speaking, and the “avant-garde” are intellectuals, artists and reformers who speak out because they think they are ahead of their time.

Evidence 1: the Pew Research Center survey (2014)

The most-cited early test of the theory on social media is Hampton, Rainie, Lu, Dwyer, Shin and Purcell, “Social Media and the ‘Spiral of Silence'” (Pew Research Center, 26 August 2014). It was a nationally representative telephone survey of 1,801 US adults aged 18 and over, fielded 7 August to 16 September 2013 in English and Spanish, with 901 landline and 900 cell phone interviews. The margin of error is plus or minus 2.6 percentage points for the full sample, and plus or minus 3.3 for the 1,076 respondents who use social networking sites. The topic was the Snowden revelations about NSA surveillance.

The headline findings are summarised in the data table below:

  • 86% of Americans were willing to discuss the surveillance issue in person, but only 42% of Facebook and Twitter users were willing to post about it on those platforms.
  • Of the 14% who were not willing to discuss it in person, 0.3% were willing to discuss it on social media, so social media did not act as an alternative forum for people who stayed silent offline.
  • Respondents who believed their coworkers agreed with them were 2.92 times more likely to join a workplace discussion. For family dinners the odds ratio was 1.90, for friends at a restaurant 1.42, and for Facebook 1.91.
  • Confidence in one’s own knowledge, intensity of opinion and interest in the topic each also raised willingness to speak.

The authors are explicit about limits. The survey covered a single public issue, did not test why people stayed silent (the authors only speculate about fear of isolation and persistent online posts), and, in their words, “our data cannot definitively demonstrate” that online silence causes offline silence. For a research design this makes the report a model of what a one-issue, cross-sectional test can and cannot establish.

Evidence 2: the 2018 meta-analysis

Matthes, Knoll and von Sikorski, “The ‘Spiral of Silence’ Revisited: A Meta-Analysis on the Relationship Between Perceptions of Opinion Support and Political Opinion Expression”, Communication Research 45(1) (DOI 10.1177/0093650217745429), pooled 66 studies with more than 27,000 participants. They report a significant positive relationship between perceived opinion climate and opinion expression of r = .10. Three results matter most for study design:

  • The relationship was not weaker in online than in offline expression environments.
  • It did not vary with the number of targets of expression, the targets’ opinion, the characteristics of the opinion climate, or the design, measurement and sample characteristics.
  • The largest silencing effect, r = .34, appeared when participants spoke to family, friends or neighbours about obtrusive issues.

The authors conclude that the relationship is stronger and more robust than previously thought. A correlation of .10 is small, and the subgroup figure of .34 is much larger, so the headline value is best read as an average over very different situations rather than as the strength of the effect in any one setting.

Reading the two sources together

The two studies answer different questions and are not directly comparable. Pew measured stated willingness on one issue for a US adult sample at one time; the meta-analysis pooled many designs across contexts. Both point the same direction on one practical question: the strongest effects sit in close social circles, not in anonymous public forums. Pew found perceived coworker agreement had the largest odds ratio and its social media users were less willing to speak in person; Matthes and colleagues found the largest correlation when talking to family, friends and neighbours.

Question for a new study What the sources show
Is the effect weaker online? Meta-analysis: not weaker than offline (Matthes et al., 2018).
Where is it strongest? Talking with family, friends or neighbours about obtrusive issues, r = .34 (Matthes et al., 2018).
Does perceived agreement matter for willingness to speak? Yes, in a 2013 US survey on one issue: odds ratios of 1.42 to 2.92 across settings (Pew, 2014).
Can the data show causation? The Pew authors say their data cannot definitively demonstrate it.

What to decide when you design a spiral of silence study

The sources above point to a short list of design choices that change what a study can claim. They come from how the cited studies were built and from the critiques below, not from a single standard method.

  1. Which issue. The meta-analysis singles out obtrusive issues in close circles as the strongest case; Pew tested one issue only.
  2. How opinion climate is measured. Pew asked about perceived agreement among specific groups (coworkers, family, friends, Facebook friends). Studies that ask only about a general public climate are measuring something different.
  3. How willingness to speak is measured. A stated willingness to join a discussion is not the same as an observed post. Scheufele and Moy (2000) questioned how willingness to speak out is operationalised.
  4. Which controls. Pew’s regression controlled for gender, age, education, race and marital status, and found knowledge confidence, opinion intensity and interest each mattered.
  5. What causal language to use. Cross-sectional associations support statements about association only.

If you code media content as part of a study of opinion climate, see content analysis and Krippendorff’s alpha for reliability. For publishing and sharing results, see social media for researchers.

What are the criticisms of the spiral of silence theory?

Scholarship has challenged the theory on the mechanism, the measurement and the context. The following criticisms are named in secondary summaries of the literature (Wikipedia, “Spiral of silence”, read 2026-11-21) and in the primary articles cited:

Criticism Source as summarised
Willingness to speak out is operationalised in questionable ways, and the conformity experiments the theory leans on lack a moral component. Scheufele and Moy (2000), International Journal of Public Opinion Research 12(1), 3-28
Fear of isolation from close circles is stronger than fear of isolation from the wider public. Moy, Domke and Stamm (2001)
People with a positive self-concept and no sense of shame speak regardless of perceived climate. Ross (2007)
Nervousness about speaking was linked more strongly to silence than fear of isolation. Ho and McLeod (2008)
On social media, vocal minorities can be comfortable voicing unpopular views online, weakening the theory. Chaudhry and Gruzd (2019)
Issue type (enduring, emerging, transitory) affects willingness to speak. Yeric and Todd

The meta-analysis complicates the last-but-one row: across its 66 studies the relationship was not weaker online, so single-platform findings of weakening should be read against the pooled result.

Spiral of silence theory on social media

Three kinds of evidence exist. The Pew survey found that people willing to discuss an issue in person were mostly not turning to social media instead. Studies summarised in the literature describe users who received strong negative reactions to political posts later censoring themselves. Others find that social bots can shape the apparent opinion climate. These findings are not uniform, and the meta-analytic result of no online weakening is the best available summary of the average.

Can you provide an example of the spiral of silence theory?

The 1965 West German election is the original example: Noelle-Neumann’s work found a late shift of 3 to 4% of voters toward the perceived winner, a pattern repeated in 1972, per the secondary summary read for this page. A modern example is the Pew finding itself: among social network users, 42% were willing to post about the Snowden NSA surveillance issue, against 86% of all adults who were willing to discuss it in person.

Illustrative composite, not a real case: a researcher in a lab meeting believes a statistical method is flawed but sees every colleague endorsing it. If she estimates that her view is in the minority, the theory predicts she stays quiet, and each quiet colleague in her position makes the endorsement look more unanimous. This shows the mechanism; it is not evidence for it.

Common misunderstandings

  • It is not proof that minorities are silenced. The meta-analytic correlation is r = .10 overall. It is an average tendency, not a law.
  • It is not only about social media. The theory dates from 1974, and Pew itself notes the pattern was documented in pre-internet communication studies.
  • Fear of isolation is a proposed driver, not a settled one. Competing explanations exist, as listed above.
  • Pew did not show silence online causes silence offline. The authors state their data cannot show causation.

Frequently asked questions

What is spiral silence theory?

It is Elisabeth Noelle-Neumann’s 1974 theory that people who perceive their opinion as unpopular tend to stay quiet, which makes the dominant opinion look stronger and encourages further silence. In research it is tested as the link between perceived opinion climate and willingness to express an opinion.

Which scholar is associated with spiral of silence theory?

Elisabeth Noelle-Neumann, a German political scientist, who published “The Spiral of Silence: A Theory of Public Opinion” in the Journal of Communication in 1974.

What are the criticisms of the spiral of silence theory?

Critics question how willingness to speak is measured (Scheufele and Moy, 2000), whether fear of isolation is the cause, whether close circles matter more than the public (Moy, Domke and Stamm, 2001) and whether social media changes the picture (Chaudhry and Gruzd, 2019). See the table above.

Can you provide an example of the spiral of silence theory?

The original example is the 1965 West German election. A modern empirical one is Pew’s 2013 survey: 86% of US adults were willing to discuss NSA surveillance in person, but only 42% of Facebook and Twitter users were willing to post about it.

Is the effect weaker on social media?

A 2018 meta-analysis of 66 studies found the relationship between perceived opinion support and expression was not weaker online than offline.

Sources

Data: Spiral of silence: key empirical figures from the Pew 2014 survey and the Matthes 2018 meta-analysis

Selected quantitative results from two sources that test spiral of silence theory: Pew Research Center's 2013 national survey on the Snowden NSA revelations (published 2014) and the Matthes, Knoll and von Sikorski 2018 meta-analysis of 66 studies. Compiled by listing each reported figure with its measure, unit, study and population; no values were estimated.

Period covered: 2013/2018. Coverage: United States (Pew); multiple countries (meta-analysis).

Spiral of silence: key empirical figures from the Pew 2014 survey and the Matthes 2018 meta-analysis
MeasureValueUnitSourcePopulation or scope
Willing to discuss the issue in person86percent of respondentsPew Research Center 2014US adults, n = 1,801, fielded 7 Aug-16 Sep 2013
Willing to post about the issue on Facebook or Twitter42percent of usersPew Research Center 2014Facebook and Twitter users in the sample
Willing to discuss on social media among those unwilling in person0.3percentPew Research Center 2014The 14% unwilling to discuss in person
Odds of joining workplace discussion if coworkers perceived to agree2.92odds ratioPew Research Center 2014US adults, logistic regression with demographic controls
Odds of joining discussion on Facebook if Facebook friends perceived to agree1.91odds ratioPew Research Center 2014Facebook users
Studies pooled66studiesMatthes, Knoll and von Sikorski 2018Meta-analysis, more than 27,000 participants
Overall correlation of opinion climate perception with opinion expression0.1rMatthes, Knoll and von Sikorski 2018All 66 studies
Largest correlation, talking to family, friends or neighbours about obtrusive issues0.34rMatthes, Knoll and von Sikorski 2018Subgroup of the meta-analysis

Sources

Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · full access

Ask about Spiral of Silence Theory: Evidence, Measurement and Criticisms

Full access: large monthly usage, up to 10 source passages behind each answer, and MCP and API access for Claude Code, Codex and other CLIs. Cancel any time.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →