Moral Panics and the Sociology of Fear

Sometimes a society becomes intensely, collectively afraid of a threat that is much smaller than the fear — and the fear itself becomes a social force, reshaping law, policing and the treatment of a category of person.

This recurring process has a name, a model, and a large literature, and it is the point where the sociology of deviance meets the sociology of fear. It also has serious critics, and a course committed to evenhandedness has to give both the concept and its critics their due — because the concept is powerful, frequently abused, and genuinely contested.

A bank holiday, a seaside town, and a threat that grew in the telling.

In the mid-1960s, on English bank holidays, groups of young people gathered at seaside resorts — two loosely defined youth styles, Mods and Rockers, with different tastes in music, clothing and transport. There were scuffles. Some minor damage. A few arrests. A number of bored young people on cold beaches, with occasional trouble.

The reporting was something else entirely.

The press covered it as an eruption of violence threatening the social order. The language was of "riots", "orgies of destruction", "battles". The scale was inflated; isolated incidents were assembled into a picture of a wave; and the two youth groups were cast as folk-devils — visible, identifiable enemies of decent society.

And the reaction produced the thing it feared.

Because the coverage told young people where the action was and what role they were expected to play, more of them came to the next bank holiday. Because the police expected trouble, they deployed heavily and made more arrests, which confirmed the picture. Because the courts treated it as a crisis, sentences were exemplary, which made more news. The reaction amplified the phenomenon — a self-fulfilling prophecy at the scale of a society (see 7.6.4, 10.1.3), in which the fear of a deviance helped produce more of it.

Stanley Cohen, studying this, named the process a moral panic, and defined it in terms that have organised the field ever since: a condition, episode, person or group of persons emerges to become defined as a threat to societal values and interests; its nature is presented in a stylised and stereotypical fashion by the mass media; the moral barricades are manned by editors, politicians and other right-thinking people; socially accredited experts pronounce their diagnoses and solutions; and ways of coping are evolved or resorted to.

Read the definition carefully, because the key word is one it does not contain: proportion. A moral panic is not merely a reaction to a threat. It is a reaction out of proportion to the threat — and the disproportion is the whole of the concept, and also the whole of the trouble with it.

Two questions the concept forces, and the second is harder than it looks.

First: why does a society fasten on this threat, at this time, with this intensity — usually out of proportion to the danger? This is the productive question, and the answers are illuminating.

Second: how do you establish that a reaction is disproportionate? This is the hard question, and it is where the critics attack — because "disproportionate" implies you know the true size of the threat, and if you knew that, measuring the reaction against it would be straightforward. In practice the analyst's judgement that a reaction is "excessive" can smuggle in the analyst's own view of how serious the threat really is (see 7.1.2 on value judgements dressed as findings), and a threat one analyst calls a panic another calls a proportionate response to a real danger.

This tension runs through the whole lesson. The concept is powerful precisely because it names a real and recurring process. It is dangerous precisely because "moral panic" can become a way of dismissing any concern one does not share — a rhetorical move rather than an analysis. Holding the concept while respecting the critics' point is the discipline here, and it is 7.9.1's proportionate-belief requirement applied to a concept about proportion itself.

The model

The stages and the elements, stated so they can be applied and tested.

The classic sequence has recognisable stages. A group or behaviour is identified as a threat. The media amplify it in stylised form. Public concern rises. The authorities respond — new laws, more policing, exemplary punishment. The response, sometimes, amplifies the phenomenon (the deviance amplification spiral). And the panic eventually subsides, leaving behind changed laws, changed institutions, or a changed target.

A later formulation identified five defining attributes, which turned the concept into something more testable — and this is the version to use, because it lets you check whether a given episode qualifies rather than just asserting it.

Concern — a heightened level of concern over the behaviour of a group and its consequences.

Hostility — increased hostility toward the group, who become folk-devils, an enemy of respectable society.

Consensus — a broad agreement that the threat is real and serious (though not necessarily universal).

Disproportionality — the concern is greater than the threat warrants; more people believe themselves at risk than the evidence supports, and the resources mobilised exceed the danger. This is the defining and the contested attribute.

And volatility — panics erupt suddenly and subside suddenly, often without the underlying condition having changed.

An episode that has all five is a moral panic in the technical sense. The value of the five-attribute version is that it makes the concept diagnosable — you can ask, of a given episode, whether each attribute is present — rather than a label to be applied to whatever alarms one considers overblown.

Who drives them, and who benefits — three models.

The interest-group model: panics are driven by the middle-level actors who benefit from them — the media, who gain audiences; the moral entrepreneurs and campaigners, who gain influence and funding; the professionals, who gain jurisdiction (see 9.5.2). On this account panics are produced by the ordinary incentives of the institutions that amplify them, and they are the media dynamics of 9.6.1 — news values favouring the dramatic, the novel and the threatening — operating on a target.

The elite-engineered model: panics are deliberately manufactured by powerful groups to divert attention from real problems or to justify expanded control. This is the strongest version and the least supported — it requires a degree of coordination and intent that is usually absent, and it has the functionalist problem (benefit is not a mechanism — see 5.1.1, 8.4.3). Panics are more often emergent than engineered, though elites certainly exploit them once underway.

And the grassroots model: panics express genuine, widespread anxiety that finds a focus. On this account the fear is real even when the target is displaced — the anxiety is about something (social change, economic insecurity, loss of status) and fastens onto a folk-devil that stands in for it. This is the deepest version, and it connects to the classic study of how a panic over one crime can be, underneath, a society working out its anxieties about something else entirely — using the fear of a threat to authorise a broader tightening of control.

The honest synthesis: most panics involve all three — real underlying anxiety (grassroots), amplified by actors who benefit (interest-group), and exploited by elites once underway (a weak version of elite-engineering). Assigning a panic wholly to any one is usually wrong (see 7.6.2 — multiple causes, and the temptation to pick the one that fits your politics).

The criticisms, taken seriously

The concept is genuinely contested, and the criticisms are strong enough that this course must give them full weight.

One — the disproportionality problem. As the puzzle noted, calling a reaction disproportionate presupposes knowing the right proportion. For some threats the disproportion is clear (a panic over a threat that demonstrably harmed almost no one). For others it is exactly what is in dispute — and the analyst who declares a panic may simply be someone who does not share the concern. The concept works cleanly only where the true scale of the threat is knowable and small, and much of its application is to cases where it is neither.

Two — it can dismiss real harms. Some things labelled "moral panics" turned out to be real and serious — early warnings that were treated as hysteria. A concept that trains you to read alarm as overblown will sometimes train you to dismiss a genuine threat, and the history includes cases where the "panic" was closer to the truth than the complacency. Child abuse, some public health threats, and some environmental dangers were dismissed by people confident they were witnessing a panic.

Three — it can condescend. Declaring the public's fear disproportionate positions the analyst above the frightened public, and the "grassroots" anxiety being displaced is treated as false consciousness (see 8.6.2 on the underclass framing — the same condescension). People's fears are sometimes well-founded, and treating them as symptoms to be explained away is both an analytical error and a political one.

Four — the media landscape has changed. The classic model assumes a few mass-media gatekeepers manufacturing a consensus (see 9.6.1). In a fragmented environment (see 9.6.2), panics are more contested, more partial, and countered in real time — there is rarely the single consensus the model assumes, and panics now often run in opposite directions in different media ecosystems simultaneously.

Where this leaves the concept: it names a real process — the amplification of a threat, the creation of folk-devils, the expansion of control out of proportion to danger — and it is one of the discipline's genuine contributions. And it must be used as a hypothesis to be tested against the five attributes, not a verdict to be pronounced on any alarm one finds excessive. The disciplined user asks whether the disproportionality can actually be established, and is willing to conclude that a given fear is proportionate.

The fear of crime

The finding that connects this Topic to the last, and to the media Part.

The fear of crime does not track crime. This is one of the most robust findings in the field, and it is the empirical heart of the sociology of fear.

People's fear of crime correlates weakly with their actual risk of victimisation (see 7.6.2). The groups most afraid — the elderly, women — are frequently among the least victimised for the crimes they fear most; the groups least afraid — young men — are frequently the most victimised. Fear and risk come apart, which means fear is not a rational readout of danger.

And fear tracks the media environment more closely than it tracks crime (see 9.6.1 on cultivation, the mean-world perception). Through the two decades in which crime fell across the rich world (see 10.2.2), the fear of crime in many countries did not fall correspondingly, and public belief that crime was rising persisted through the entire fall. People believed crime was getting worse while it was getting dramatically better — a gap between perception and reality as large as any in this course.

Why fear persists when crime falls has several documented components: the media's continued supply of dramatic crime content regardless of the rate (news values — see 9.6.1); the way fear expresses anxieties that are not really about crime at all (the grassroots model — insecurity displaced onto a legible threat); and the political usefulness of fear, which few actors have an incentive to reduce.

And the consequences are real. Fear of crime, whatever its relation to crime, changes behaviour and authorises policy: it empties public space, justifies surveillance and harsher sentencing (see 10.2.3, 10.3.1), concentrates on the folk-devils the panics produce, and falls hardest on the groups already most policed (see 8.4, 10.2.3). The fear does real work even when the threat does not warrant it — which is why a sociology of control has to study the fear as well as the crime.

Because control expands through fear, and the fear is systematically decoupled from the danger — which means a citizen has to be able to assess a fear as carefully as a fact.

Three questions to carry.

Does the reaction have the five attributes? Concern, hostility, consensus, disproportionality, volatility. Use them as a diagnosis, not a label — and be genuinely willing to conclude that a given alarm is a proportionate response to a real threat, because sometimes it is.

Can the disproportionality actually be established? This is the concept's load-bearing and most abusable element. If the true scale of the threat is unknown or contested, "moral panic" is a hypothesis, not a finding — and calling something a panic because you do not share the concern is 7.1.2's error, a value judgement dressed as an analysis.

And whose fear, of what, benefiting whom? Fear is produced by actors with incentives (media, campaigners, professionals, politicians), expresses anxieties that may not be about its ostensible object, and authorises control that falls on the usual targets. Follow the fear to what it displaces and what it licenses.

One closing observation that draws the Topic together. The fear of crime rose, or held, while crime fell — which is the single cleanest demonstration in the Part that the response to deviance is not a rational function of the deviance. It is produced by the media environment, by political incentives, by displaced anxiety, and by the machinery that benefits from amplification.

Which returns us to where the Part began (see 10.1.1): deviance is a property of the response, and the response is patterned, studiable, and frequently out of all proportion to the act. The Mods and Rockers on a cold beach did very little. The reaction did a great deal — it made law, made careers, made folk-devils, and made more Mods and Rockers. The sociology of control is, in the end, the sociology of that gap — between what people do and what is done about it — and closing your own perception of that gap, so that your fears track dangers and your alarms survive the five-attribute test, is as much a part of the sociological imagination as anything in this course.

Mods and Rockers on English beaches did little — scuffles, minor damage, a few arrests — and the press reported an eruption threatening the social order, casting the youths as folk-devils. The reaction amplified the phenomenon: coverage drew more youths, expectation drew more police, and exemplary sentences made more news — a self-fulfilling prophecy at the scale of a society. Cohen named this a moral panic, and its key word is the one the definition omits: proportion.

The concept forces two questions: why a society fastens on a threat out of proportion to the danger, and — harder — how disproportionality can be established, since calling a reaction excessive presupposes knowing the true size of the threat. "Moral panic" can become a way of dismissing any concern one does not share.

The five-attribute version makes it diagnosable: concern, hostility (folk-devils), consensus, disproportionality (defining and contested), and volatility. Three models of who drives them: interest-group (media, campaigners and professionals who benefit — the news dynamics of 9.6.1), elite-engineered (strongest and least supported, with the functionalist problem), and grassroots (real anxiety displaced onto a folk-devil) — with most panics involving all three.

The criticisms are strong: the disproportionality problem (works cleanly only where the threat's scale is knowable and small); the risk of dismissing real harms (some "panics" were early warnings that were right); condescension toward the frightened public; and the fragmented media landscape that no longer manufactures a single consensus. The concept names a real process and must be used as a testable hypothesis, not a verdict.

And the fear of crime does not track crime: it correlates weakly with actual risk (the most afraid are often the least victimised), tracks the media environment more closely, and persisted or rose through the two-decade crime drop — the cleanest demonstration that the response to deviance is not a rational function of the deviance. Fear does real work regardless of the threat — emptying space, authorising surveillance and harsher sentencing, falling on the usual targets — which is why control has to be studied as the sociology of the gap between what people do and what is done about it.

Moral panic — a reaction to a threat that is out of proportion to the threat, with folk-devils and expanded control.

Folk-devil — the visible, stereotyped enemy figure a panic constructs.

Deviance amplification spiral — the process by which reaction increases the deviance it reacts to.

The five attributes — concern, hostility, consensus, disproportionality, volatility; the diagnostic version.

Disproportionality — the defining and most contested attribute; presupposes a knowable threat scale.

Interest-group / elite-engineered / grassroots models — panics as produced by beneficiary actors, by deliberate elite design, or by displaced genuine anxiety.

Moral entrepreneur — a campaigner who works to establish a rule or a threat as public concern.

Fear of crime — a fear that correlates weakly with risk and tracks the media environment more than the crime rate.

The perception–reality gap — the persistence of belief that crime is rising through a period when it fell.

The disproportionality problem — the difficulty of establishing excess without presupposing the threat's true size.

One — diagnose with the five. Take a current alarm and check it against concern, hostility, consensus, disproportionality and volatility. Decide honestly whether it qualifies — including the possibility that it is proportionate.

Two — test the disproportionality. For that alarm, ask whether the true size of the threat is knowable. If it is not, "panic" is a hypothesis, not a finding.

Three — find the folk-devil. For one panic, identify the stereotyped enemy figure and what broader anxiety it might stand in for.

Four — check fear against risk. For a crime you fear, find out your actual likelihood of experiencing it. The gap is the sociology of fear.

Five — follow the beneficiaries. For one fear in circulation, list who gains from its amplification — media, campaigners, professionals, politicians. Then ask what the fear licenses.