How Social Norms Change: From Private Doubt to New Expectations

Social norms can appear remarkably stable because people reproduce them every day. Individuals follow a rule, observe others following it and infer that the behaviour is widely accepted. Repetition then makes the rule feel natural, even when it is historically recent or privately disliked by many of the people complying with it.

Yet social norms can change dramatically. Behaviours once treated as ordinary can become unacceptable, while practices that were once stigmatised can become routine. Expectations around smoking, workplace behaviour, family roles, health, dress, safety, communication and many other areas have shifted within individual lifetimes.

The central puzzle is how a rule that reinforces itself can become unstable.

Research on social norm change suggests that the answer usually involves more than persuading individuals one by one. Norms depend partly on people's expectations about other people: what they believe others normally do, what they think others approve of and what social consequences they expect if they behave differently. Contemporary research therefore treats norms as dynamic social systems rather than fixed cultural rules. A 2024 Royal Society review describes social norms as unwritten rules that evolve with group processes, institutions and changing social conditions.

Norm change begins when those expectations stop reinforcing the old behaviour. Private disagreement becomes visible. Alternative behaviour becomes socially credible. Networks transmit evidence that others are changing. Institutions alter incentives. New language allows previously hidden experiences to be discussed. Sometimes these forces accumulate slowly; sometimes they combine to produce rapid change.

Private disagreement can exist beneath public conformity

One reason a norm can look stronger than it really is is that public behaviour does not always reveal private belief.

People may personally dislike a practice but continue following it because they believe most other members of their group support it. Each person's conformity then becomes evidence for everybody else that the norm remains accepted.

This pattern is often described as pluralistic ignorance.

Imagine a workplace in which employees dislike a culture of answering messages late at night. Each employee observes everyone else responding and assumes colleagues expect constant availability. Because nobody wants to appear less committed, everyone continues responding. The visible behaviour suggests strong support for a rule that many individuals privately dislike.

The same mechanism can operate in schools, families, professions and communities. A person may disagree with a norm but decide that violating it carries too much reputational or social risk.

This explains why information about private opinion can sometimes destabilise an established rule.

Anonymous surveys, confidential employee feedback, private voting or credible public testimony can reveal that what appeared to be consensus was partly coordinated compliance. Once people learn that others also have doubts, dissent becomes less socially isolating.

But private disagreement alone rarely changes a norm.

People also need evidence that behaving differently is possible.

Early adopters therefore play an important informational role. Activists, professionals, artists, community leaders or ordinary peers may be among the first people willing to behave differently in public. Because they move before the alternative has become safe or normal, they often face greater social costs than later adopters.

Their significance is not only that they are courageous or visible. They provide information.

One person taking parental leave in a workplace may demonstrate that doing so is possible. A senior colleague openly admitting uncertainty can signal that asking questions does not necessarily indicate incompetence. A family discussing a previously avoided health issue can show others that direct conversation is survivable.

The first visible departures from a norm therefore create examples that later participants can evaluate.

People respond not only to what is normal now, but to where the norm appears to be going

Traditional social-norm messages often emphasise current prevalence: “most people do this” or “most people approve of this behaviour.”

That strategy can work when the desired behaviour is already common. But it creates a problem when the behaviour an organisation hopes to encourage remains a minority practice. Claiming that everybody already follows it would be inaccurate.

Research on dynamic norms offers another possibility.

Instead of communicating only what most people currently do, a dynamic-norm message communicates a direction of change: more people are beginning to adopt the behaviour.

In a series of experiments, psychologists Gregg Sparkman and Gregory Walton found that information about changing behaviour influenced participants even when the new behaviour had not yet become the majority norm. In one field experiment, communicating that people were increasingly reducing meat consumption increased meatless orders at a café. Their findings suggest that people respond not only to current social prevalence but also to evidence about where they expect social behaviour to move in the future.

The mechanism makes intuitive sense. People are not merely conforming to the present; they may anticipate future expectations.

If a practice is becoming more common, adopting it no longer feels like joining a permanently marginal group. It can instead feel like moving toward an emerging norm.

This does not mean that adding phrases such as “more people are changing” will automatically transform behaviour. The information has to be credible, relevant to the group whose approval matters and connected to behaviour people can realistically perform.

False norm claims can also backfire by undermining trust.

The broader lesson is that social expectations contain a temporal dimension. People care not only about what their community considers normal today but also about what appears likely to become normal tomorrow.

Networks determine how quickly alternative behaviour becomes credible

People rarely observe an entire society directly. They observe other people through networks.

Family members, friends, colleagues, neighbours, classmates, professional peers, religious communities and online groups provide much of the social information from which norms are inferred.

This means that a behaviour can become normal within one network while remaining unusual elsewhere.

A new professional practice might spread quickly within one industry but barely appear in another. A social attitude may become common among younger urban networks while remaining uncommon in older or geographically distant groups. National averages can hide these very different local realities.

Network structure also affects how change spreads.

For simple information, hearing something once may be enough. But adopting behaviour that carries social risk often requires reinforcement from several trusted people. One friend behaving differently may look eccentric. Several respected peers doing the same thing can make the behaviour appear socially viable.

This is why highly visible individuals do not automatically change norms. A celebrity may reach millions of people, but someone's close professional or family network may matter more when the behaviour involves identity, reputation or everyday cooperation.

The social cost of acting early is central.

Some people require relatively little reassurance before changing. Others move only when they see considerable evidence that members of their own relevant group will accept the new behaviour.

These differences create adoption thresholds.

When enough low-threshold individuals move first, they can supply the evidence required by people who need more reassurance. Those people then change, providing still more evidence to the next group. Under certain conditions, this can create a cascade.

Laboratory experiments led by Damon Centola and colleagues have demonstrated that committed minority groups can produce tipping dynamics in social conventions. The experiments provide evidence that coordinated minority behaviour can sometimes trigger rapid changes in group convention once sufficient social support accumulates.

But the popular idea that every society has one fixed numerical “tipping point” goes beyond what the evidence establishes.

A threshold observed in a controlled experimental setting should not be treated as a universal law governing attitudes toward smoking, gender, workplace conduct or political behaviour. Real societies contain multiple overlapping networks, institutions, identities, power differences and sanctions.

The relevant threshold can therefore vary enormously depending on who is changing, who is watching and what early adopters risk losing.

Institutions can turn emerging behaviour into an established expectation

Norms are informal rules, but formal institutions can profoundly influence them.

Laws, workplace policies, school rules, professional standards, market incentives and platform design alter the environment in which people decide whether to follow or violate a norm.

Consider smoke-free public spaces.

A legal restriction does more than create a fine. It changes where smoking is visible. It changes what nonsmokers expect when entering a restaurant or workplace. It gives people a formal basis for objecting. Over time, behaviour that once seemed routine can become socially unusual even in situations where direct enforcement is weak.

This illustrates the interaction between formal rules and informal expectations.

A law that people consider illegitimate may produce compliance only when enforcement is visible. A rule consistent with emerging social expectations can eventually become partly self-enforcing because ordinary participants begin rewarding compliance and disapproving of violations.

Institutions can also make new behaviour practical.

A company cannot meaningfully encourage parental leave while penalising employees who use it. A city cannot expect cycling to become normal if roads make cycling unusually dangerous. A health campaign cannot simply ask people to seek treatment when services are inaccessible.

This is why norm-change campaigns sometimes fail even when their message is persuasive: the practical environment still rewards the old behaviour.

UNICEF's guidance on social norms emphasises precisely this problem. Behaviour can be shaped by a mixture of individual motives and social expectations, so programmes need to diagnose whether a practice is genuinely norm-driven before deciding how to intervene.

A campaign aimed at personal attitudes will have limited effect if the real barrier is fear of social punishment.

Conversely, changing formal rules alone may achieve little if the informal group still strongly rewards the old practice.

Durable change often occurs when incentives, institutions and social expectations begin pushing in the same direction.

Language and media can change what people believe others think

Norms can persist partly because people lack language for describing an experience.

A person may repeatedly encounter a workplace behaviour that feels wrong without possessing a recognised term for it. Once a concept enters common language, individuals can compare experiences that previously seemed disconnected.

Terms related to harassment, consent, accessibility, burnout, discrimination and environmental harm can perform this function.

Naming does not prove that one interpretation is correct or settle every disagreement about how a concept should be applied. But it changes the informational environment.

People can identify patterns, compare experiences and make claims that institutions are able to recognise.

Language therefore functions as infrastructure for social change.

Media can amplify this process by making behaviour visible at a much larger scale.

Investigative journalism may reveal that a supposedly isolated problem occurs systematically. Entertainment can make previously unfamiliar identities or lifestyles more visible. News coverage can show that people in other organisations or communities are already acting differently.

Repeated representation can weaken the belief that an alternative is rare.

But visibility can also mislead.

A highly active minority on social media may create the impression that a behaviour is widespread. Viral incidents can make unusual actions seem common because they are memorable. Coordinated campaigns can deliberately exaggerate the appearance of consensus.

This creates an important distinction between attention and prevalence.

Something can dominate online discussion without being common in the population. If people mistake visibility for frequency, media can distort perceived norms rather than accurately reveal them.

This matters because communicating that harmful behaviour is widespread can sometimes produce an unintended normalisation effect.

A campaign warning that “everyone is doing X” may be intended to alarm the public but can simultaneously tell uncertain observers that the behaviour is ordinary.

Effective communication therefore has to consider not only whether a message condemns a behaviour, but what the message implies about how many other people engage in it or approve of it.

Social change can occur through generations, crises and organised norm entrepreneurship

Some norms change slowly through population replacement.

Younger cohorts may enter adulthood with different expectations from older generations. As their share of workplaces, institutions and electorates increases, aggregate behaviour changes even if many individuals retain the beliefs they formed earlier.

This process can produce major long-term shifts.

But generational replacement is only one mechanism.

Adults also revise expectations in response to migration, education, economic change, new relationships, law, technology and experience. Treating social values as permanently fixed by youth underestimates people's capacity to adapt.

Crises can accelerate this process because they disrupt routines.

Wars, pandemics, disasters and economic shocks can make existing forms of coordination temporarily impossible. People experiment because the old rule no longer works. Practices that previously faced resistance can spread rapidly when circumstances suddenly make them useful or necessary.

But emergency behaviour does not automatically become a durable norm.

Once the crisis ends, old incentives and routines may return. A change is more likely to persist when people discover lasting benefits, institutions support the new arrangement and participants begin expecting others to continue it.

Social movements, governments, professional organisations, companies and religious or community leaders can also deliberately attempt to change norms. Scholars sometimes call such actors norm entrepreneurs.

They may frame an existing practice as inappropriate, publicise alternatives, reward early adopters and create new language around the desired rule.

The term itself carries no moral endorsement.

Norm entrepreneurs can promote public health or harmful exclusion. They can broaden rights or encourage prejudice. The sociological question is not whether the proposed norm is automatically good but how actors attempt to change what their audiences consider normal or acceptable.

This also explains why backlash frequently accompanies norm change.

A changing norm can threaten identity, status or routines that people consider morally important. Those who support the old expectation may interpret visible change not as improvement but as evidence that society is moving in the wrong direction.

Resistance can therefore intensify when change becomes sufficiently visible to seem consequential.

Norm change is rarely a smooth movement from old to new. Adoption, resistance, institutional conflict and partial reversal can all occur along the way.

Measuring norm change requires separating beliefs, expectations and behaviour

One of the easiest mistakes in studying social norms is assuming that behaviour reveals everything.

Suppose people suddenly begin complying with a new workplace policy. That could mean they now personally support the policy. It could also mean that they fear sanctions while privately disapproving.

Alternatively, people may privately approve of a new behaviour but continue following the old rule because they believe their peers still reject the alternative.

Several different things therefore need to be measured separately.

Researchers may ask what a person personally believes. They may ask what the person thinks other people do. They can ask what the person believes others approve of. They can also observe what the person actually does.

These measures are not interchangeable.

The source draft correctly emphasises that behaviour, personal attitudes and expectations about other people's attitudes can move at different speeds.

This distinction matters enormously for intervention design.

Suppose nearly everyone privately agrees that a harmful practice should stop but each person believes most neighbours still support it. Another moral-awareness campaign may contribute little. The missing information is not whether the behaviour is harmful; it is whether people realise that others share their opposition.

Conversely, people may express strong approval for a healthier behaviour while lacking the infrastructure, money or time needed to perform it. A social-norm campaign cannot solve the practical constraint.

Good diagnosis therefore begins with several questions:

What do people actually do?

What do they personally approve of?

What do they think other people do?

What do they think other people approve of?

What happens socially when someone behaves differently?

And what practical barriers make the alternative difficult even when people support it?

UNICEF's social-norm programming guidance explicitly warns that collective behaviours have multiple possible drivers and that interventions need to establish whether social expectations are genuinely maintaining the practice before choosing norm-focused strategies.

Norm change can be uneven, reversible and much less dramatic than it looks

When norms change, they rarely change everywhere simultaneously.

A practice may disappear quickly within a professional network but remain common elsewhere. Urban and rural communities can move at different speeds. Different generations, religious groups, occupations or social classes may develop different expectations.

Migration can also combine normative systems rather than simply replacing one with another.

This is why national averages can conceal important variation.

A survey showing broad acceptance of a new norm does not mean every individual encounters that expectation in daily life. The people whose approval matters most may still belong to a network where the old rule dominates.

Change can also reverse.

Visible supporters may disappear. Institutions may stop reinforcing the new expectation. A crisis that produced temporary cooperation may end. Organised opposition may successfully reactivate identities associated with the previous rule.

A durable norm therefore involves more than short-term behavioural compliance.

People must begin expecting that others will continue following the new rule even when the original campaign, charismatic advocate, policy intervention or emergency is no longer present.

This helps explain why some changes appear sudden only in retrospect.

Years of private doubt may precede visible dissent. Small groups may develop new practices before wider society notices. Institutions may slowly alter incentives. New language may spread through professional and cultural networks.

Then, once enough pieces align, public behaviour changes rapidly.

The visible shift looks abrupt. The preparation was not.

Social norms change when people begin expecting something different from one another

A social norm is more than an individual preference and more than a frequently repeated habit.

It is part of a coordinated social expectation.

People behave partly in response to what they believe others normally do, what they think others consider appropriate and what consequences they expect if they depart from those rules. UNICEF similarly describes social norms as perceived informal rules shaped by beliefs about what others do and approve of.

This is why norms can survive private disagreement.

It is also why they can sometimes change rapidly.

Hidden opposition becomes visible. Early adopters demonstrate alternatives. Networks spread social evidence. Institutions change incentives. Language makes previously diffuse problems discussable. Media alters visibility. New generations and crises disrupt established expectations.

None of these mechanisms guarantees change on its own.

Norms are strongest when behaviour, approval, institutions and social sanctions reinforce one another. Durable transformation often requires that the same forces gradually begin reinforcing a different expectation.

The deepest lesson is therefore that social change is not always achieved by convincing each individual separately.

Sometimes many individuals have already changed their minds.

What has not yet changed is what they believe everyone else expects.

When that expectation moves, the social rule itself can begin to move with it.