Social Cohesion Explained: What Holds Societies Together and What Pulls Them Apart

Social cohesion is the capacity of people who are different from one another to remain connected enough to cooperate, trust important institutions and resolve conflict without treating every disagreement as a reason to abandon the shared social order. A cohesive society does not require everyone to have the same religion, ethnicity, political beliefs, income, lifestyle or understanding of history. It requires enough common ground that people continue to recognise one another as legitimate members of the same society even when their interests collide.

This distinction matters because social harmony can be misleading. A country may appear orderly because dissent is suppressed, communities rarely interact or powerful institutions prevent conflict from becoming visible. That is not necessarily cohesion. Durable cohesion is tested when disagreement actually occurs: when an election is lost, taxes rise, migration increases, economic conditions deteriorate, communities compete for resources or government makes a decision that some citizens strongly oppose.

The World Bank defines social cohesion through shared purpose, trust and willingness to cooperate, distinguishing relationships within communities, relationships across different communities and relationships between people and the state. This definition is useful because it prevents cohesion from being reduced to friendliness. Strong families and tightly connected local groups can coexist with deep distrust between ethnic communities, political camps or citizens and public institutions.

Social cohesion therefore involves at least three questions. Do people trust and support those close to them? Can they cooperate with people outside their immediate group? Do they believe public institutions treat them fairly enough to remain legitimate? A society becomes fragile when one or more of these relationships deteriorates severely.

The issue has become increasingly important. The UN World Social Report 2025 warns that economic insecurity, persistent inequality, declining trust and social fragmentation are destabilising societies. It reports that more than half of the global population has little or no trust in government and argues that equity, economic security and solidarity need to become central components of social policy. At the same time, OECD evidence shows that trust in institutions and the quality of ordinary social connections vary substantially across countries and demographic groups.

Social cohesion is therefore not an abstract sociological ideal. It affects whether societies can make collective decisions, absorb economic shocks, manage cultural diversity, respond to disasters and maintain democratic institutions when political competition becomes intense.

What Social Cohesion Means—and What It Does Not Mean

There is no single universally accepted definition of social cohesion. Researchers, governments and international organisations emphasise different dimensions depending on whether they are studying conflict, inequality, trust, citizenship, social networks, institutions or development.

The World Bank’s current framework is particularly practical because it divides cohesion into relationships within groups, across groups and between people and structures of authority. These dimensions are sometimes described in related research as bonding, bridging and linking relationships, although those concepts are also associated with the broader literature on social capital.

Bonding relationships connect people who already share important characteristics or identities: relatives, close friends, religious communities, ethnic groups or neighbourhood networks. These relationships can provide emotional support, practical help and a sense of security.

Bridging relationships connect people across social boundaries. A workplace may bring together people from different religions. A public school may connect families from different classes or communities. Sports clubs, universities, neighbourhood associations and professional organisations can create relationships that cross otherwise separate social worlds.

Vertical or linking cohesion concerns relationships between citizens and institutions. People interact with governments through schools, courts, police, taxation, welfare systems, local authorities, elections and administrative services. If these encounters repeatedly appear arbitrary or discriminatory, formal citizenship may coexist with deep institutional distrust.

Strong social cohesion requires more than powerful bonding networks. A society can contain extremely cohesive political camps, sectarian communities or ethnic groups while being fragmented as a whole. Members may trust one another intensely yet view outsiders with hostility.

That is why cohesion should not be confused with conformity. A country where people fear expressing disagreement may appear calm, but silence generated by repression is not the same as cooperation generated by legitimacy. Cohesion becomes more meaningful when people can disagree, criticise institutions and still accept one another’s membership and basic rights.

Nor is cohesion identical to social capital. Social capital usually refers more specifically to networks, relationships, trust and norms that make cooperation possible. Social cohesion is broader because it also includes questions of belonging, institutional legitimacy, inclusion, inequality and whether society remains sufficiently integrated across group boundaries.

A useful way to think about the distinction is that social capital describes resources contained in relationships; social cohesion describes the wider condition of a society in which those relationships are distributed strongly enough across important boundaries to sustain cooperation.

Trust, Belonging and Social Connections Are the Core Infrastructure

Trust reduces the amount of effort required for everyday cooperation. If people generally expect neighbours, businesses or institutions to behave predictably, they do not need to verify every interaction from the beginning. Contracts become easier to enforce socially and legally, information moves more efficiently and cooperation becomes less psychologically expensive.

OECD research treats both interpersonal trust and institutional trust as important dimensions of social and economic life. Its trust framework distinguishes trust in other people from trust in public institutions and links both to broader questions of governance, well-being and cooperation.

Institutional trust deserves particular care because democratic trust should not mean unquestioning obedience. Citizens can distrust a particular government, criticise policies or protest against leaders while still trusting courts, elections or administrative procedures enough to use them. In fact, healthy democratic systems require scrutiny as well as confidence.

The OECD’s 2024 Trust Survey found that across participating OECD countries, 39% reported high or moderately high trust in national government while 44% reported low or no trust. Only 41% believed government generally used the best available evidence when taking decisions, and trust was strongly influenced by whether people believed institutions were accountable, competent, fair and open to citizen influence.

The 2025 edition of Government at a Glance reports a similarly uneven picture. Across surveyed OECD countries, around 39% expressed high or moderately high trust in national government, compared with 54% for courts and the judicial system, 45% for local government and the civil service, and only 37% for national parliaments. Trust therefore varies not only between countries but between institutions within the same society.

Belonging adds another dimension. A citizen can possess equal legal rights on paper while repeatedly receiving social signals that they are considered an outsider. When people believe that public institutions do not represent individuals like them, the emotional connection between formal citizenship and social membership weakens.

Belonging does not require a single culture. A cohesive society can contain multiple religions, languages and regional identities. What matters is whether individuals believe there is a credible place for them within the shared political and social order.

Everyday social relationships also function as a form of infrastructure. Friends, relatives, neighbours, community groups, religious organisations, unions, clubs and voluntary associations move information and assistance through society. They provide help during illness, unemployment, disasters or personal crises and create environments in which people practise reciprocity before a national emergency ever occurs.

The OECD’s 2025 report on social connections illustrates both the strength and fragility of this infrastructure. Across OECD countries, more than two-thirds of respondents interacted with friends or family at least daily and around 90% reported having someone they could rely on in a time of need. At the same time, around 10% felt unsupported, 8% reported having no close friends in the European OECD countries covered, and 6% experienced loneliness most or all of the time during the previous four weeks.

The same report notes that face-to-face contact has declined over a longer period in a number of OECD countries even as remote communication increased. Digital contact can sustain valuable relationships, but the change raises wider questions about what happens to public life when people spend less time in shared physical spaces.

Libraries, parks, markets, schools, sports grounds, community organisations and walkable public spaces therefore matter socially as well as physically. OECD work increasingly describes such places as social infrastructure because they create opportunities for repeated interaction rather than simply providing services.

Inequality, Economic Security and Fairness Shape Whether Cooperation Feels Legitimate

Economic inequality does not automatically destroy social cohesion. Many societies tolerate significant differences in income when citizens believe economic opportunities remain credible, essential services work and rules are applied reasonably fairly.

The problem becomes more serious when inequality is interpreted as evidence that institutions serve some groups permanently while asking others to carry disproportionate burdens. Material inequality then becomes political and psychological inequality.

The World Social Report 2025 places economic insecurity, inequality and declining trust at the centre of what it describes as a global social crisis. More than 2.8 billion people live on between $2.15 and $6.85 per day, leaving a very large share of humanity economically vulnerable even when they are above the extreme-poverty threshold. The report also states that two-thirds of the global population live in countries where income inequality is increasing.

Economic insecurity matters because people living close to financial crisis experience institutions differently from those who can absorb shocks. A rent increase, illness or temporary unemployment may become existential rather than inconvenient. Under those conditions, politics can increasingly be understood as a contest over which group receives protection and which group pays.

Economic security alone does not produce cohesion. Wealthy societies can be deeply polarised. But basic security makes cooperation easier because every public decision is less likely to be experienced as a struggle for immediate survival.

Procedural fairness is equally important. People care about outcomes, but they also care about whether rules were applied consistently, whether decisions were explained and whether they had a meaningful opportunity to be heard.

This helps explain why two citizens who receive the same material outcome can judge an institution differently. Someone who receives an unfavourable decision through a transparent procedure with clear reasons may still accept the legitimacy of the process. A person who receives a favourable outcome through obvious favouritism may conclude that the system itself is unreliable.

The OECD Trust Survey finds that ordinary experiences with public agencies—including whether people expect applications for services or benefits to be treated fairly—are important drivers of trust in government institutions. This is a reminder that social cohesion is not built only through constitutions, national speeches or patriotic rituals. It is built every time a citizen encounters a school, court, municipal office, police officer or public-service system.

Inclusion is the distributional side of the same problem. Formal membership has limited meaning when groups cannot realistically access education, employment, housing, transport, public services or political voice.

Exclusion tells citizens that there are two forms of membership: membership recognised by law and membership recognised in practice.

When those diverge persistently, cohesion becomes difficult to sustain.

Diversity Does Not Automatically Weaken Social Cohesion

The relationship between diversity and cohesion is often oversimplified. It is common to assume that societies are cohesive when people resemble one another and fragmented when populations are culturally diverse.

History provides too many counterexamples for that rule to work.

Relatively homogeneous societies can be divided by class, region, ideology, generation or political identity. Diverse societies can sustain strong institutions and extensive cooperation when citizenship, rights and opportunity are credible.

The more important question is how differences are organised.

If people from different groups attend completely separate schools, live in sharply segregated neighbourhoods, consume entirely separate information and interact mainly within their own communities, opportunities for bridging relationships remain weak. Strong in-group solidarity may then coexist with distrust of outsiders.

If people encounter one another repeatedly in workplaces, universities, public transport, neighbourhood spaces and civic organisations, difference becomes more likely to be ordinary rather than exceptional.

None of this means contact automatically creates harmony. Contact can reproduce hierarchy when one group consistently holds more status or power. Institutions therefore influence whether interactions take place under conditions that reinforce equality or confirm exclusion.

A shared national identity can contribute to cohesion when membership is broad enough to include citizens with different backgrounds. National identity allows strangers to imagine that they owe one another some degree of solidarity simply because they participate in the same political community.

It becomes divisive when national identity is defined so narrowly that citizens belonging to minority religions, ethnicities or linguistic communities are treated as permanent guests. A person repeatedly told that they do not fully belong may eventually stop believing that national institutions represent them.

This is why shared purpose is often more useful than forced sameness. Citizens do not need identical cultural identities to cooperate around functioning schools, predictable law, public safety, economic opportunity, disaster response, clean public spaces or infrastructure.

Practical cooperation can itself generate solidarity. People who successfully solve a shared local problem may build trust across differences that ideological appeals never overcame.

Democratic Institutions Need to Make Disagreement Survivable

Conflict is normal in political life. People disagree about taxation, immigration, religion, welfare, education, policing, climate policy, land use and countless other issues.

The important question is whether institutions give those disagreements a legitimate place to go.

Courts convert some conflicts into legal arguments. Elections convert political competition into periodic choices. Legislatures negotiate among interests. Trade unions and professional bodies represent organised groups. Local councils provide forums for neighbourhood disputes. Civil-society organisations allow citizens to mobilise without immediately turning political disagreement into violence.

A society becomes more fragile when large groups conclude that these channels cannot protect their interests or change anything.

This does not mean every institutional decision must satisfy everyone. No functioning political system can produce unanimous outcomes. Cohesion instead requires losers to retain enough confidence in the process to remain participants.

Political alternation is therefore one of the clearest tests of democratic cohesion.

A party loses an election. Its supporters are disappointed, perhaps deeply angry. But they believe they can organise, criticise the government and compete again. The losing side remains part of the nation rather than being treated as an enemy population.

The same principle applies to election winners. A cohesive democracy requires winners to understand that electoral victory gives temporary governing authority, not permanent ownership of the state.

When both sides begin treating every election as an existential struggle in which defeat means exclusion from the political community, institutional trust deteriorates rapidly.

This is where cohesion differs fundamentally from conformity. A society with real cohesion can tolerate opposition because disagreement does not automatically threaten the right to belong.

Media, Digital Platforms and Shared Reality

Modern social cohesion is increasingly shaped by the information environment.

Traditional mass media never produced one perfectly shared national conversation, but they often created overlapping sets of public facts. Digital platforms have made information much more abundant while fragmenting the pathways through which people encounter it.

Different groups can now receive different political stories, statistics, videos and interpretations through recommendation systems optimised for engagement.

This does not mean social media inevitably destroys cohesion. Digital platforms can help minority communities find one another, connect people across geography, circulate emergency information and create new forms of political participation.

The problem appears when societies lose not only agreement about values but agreement about basic observable reality.

The UN World Social Report 2025 identifies misinformation and disinformation as factors worsening distrust and social fragmentation. If citizens cannot agree whether an event occurred, whether an institution released a particular document or whether a piece of media is authentic, ordinary political disagreement becomes much harder to resolve.

Artificial intelligence further complicates the environment because convincing text, images, audio and video can be produced at low cost. The response cannot simply be censorship or expecting every citizen to become a professional fact-checker.

Institutions need credibility, transparency and timely communication. Citizens need media literacy. Platforms need systems for limiting clearly manipulative or fraudulent behaviour without converting disagreement into prohibited speech.

Social cohesion therefore increasingly depends on something that older theories sometimes treated as background infrastructure: shared methods for deciding what evidence deserves trust.

Crisis Reveals Whether Social Cohesion Actually Exists

Social cohesion is easiest to claim during stable periods. Crisis tests whether it is real.

Pandemics, floods, earthquakes, economic crashes, wars and large-scale displacement require people to accept decisions whose benefits and burdens are uneven. Resources may need to be rationed. Movement may be restricted. Taxes may rise. Some regions may receive assistance before others.

Trust becomes operational.

Do people believe official warnings? Will they cooperate with evacuation orders? Do communities share resources? Do citizens believe emergency assistance is being distributed according to need rather than political loyalty?

Crisis can strengthen solidarity when people experience a genuine common purpose. It can also deepen existing divisions when some groups believe they are being ignored, blamed or asked to bear disproportionate costs.

The World Bank places social cohesion directly within its work on fragility, conflict, displacement and resilience because trust across communities and between citizens and institutions can affect whether societies withstand shocks without escalating into violence.

This is another reason governments should not begin thinking about cohesion only after a crisis starts.

Trust accumulated through years of competent, fair everyday administration becomes a resource during emergencies.

Trust repeatedly damaged before a crisis cannot always be manufactured through emergency messaging when the disaster arrives.

How Social Cohesion Can Be Strengthened

There is no single “social cohesion policy” because cohesion is produced by relationships, institutions, economic conditions and physical environments simultaneously.

The most effective approach is therefore usually a portfolio of interventions rather than one national campaign about unity.

Reliable public services matter because competence affects trust. Citizens are more willing to accept collective obligations when schools function, administrative procedures are understandable and public authorities respond predictably.

Equal treatment matters because discrimination undermines the credibility of shared citizenship. Anti-discrimination law is important, but enforcement and everyday institutional behaviour matter as well.

Economic security matters because extreme vulnerability makes cooperation more difficult. The World Social Report 2025 specifically argues for equity, economic security and solidarity as foundations for rebuilding social trust.

Participation matters because people are more likely to accept institutions they can influence. The OECD Trust Survey finds substantial relationships between trust and whether citizens believe they have a voice in public decision-making. Consultation becomes counterproductive, however, when people are repeatedly asked for views and then receive no explanation of what happened to them.

Public spaces matter because bridging relationships need places to form. Libraries, parks, sports clubs, community centres, schools and accessible public transport bring people into ordinary contact outside their closest social circles. OECD’s 2025 work on social connections explicitly identifies investment in social infrastructure as a promising direction for policy.

Housing and urban planning matter because severe residential segregation can reduce opportunities for meaningful interaction. If income, ethnicity and access to services map onto entirely separate neighbourhoods, social distance becomes physical distance.

Civil society matters because not every relationship should run through government. Associations, unions, religious organisations, charities, clubs and voluntary groups create independent networks through which citizens learn cooperation.

Media literacy and credible public information matter because political cooperation becomes extraordinarily difficult when citizens operate from incompatible factual realities.

Finally, reciprocity matters. People are more willing to sustain collective systems when they believe obligations and benefits are distributed according to understandable principles. Social solidarity becomes difficult when any group believes it is permanently expected to contribute while another group permanently receives without responsibility—or when vulnerable groups believe society demands contribution while denying them real opportunity.

Cohesion therefore requires both protection and participation.

How Social Cohesion Is Measured

There is no single social-cohesion score capable of summarising an entire society accurately.

Researchers therefore measure cohesion indirectly through combinations of indicators such as interpersonal trust, confidence in institutions, social support, belonging, volunteering, civic participation, discrimination, loneliness, perceptions of fairness and willingness to cooperate with other groups.

This multidimensional approach is necessary because indicators can move in different directions.

A society might have strong family networks but weak institutional trust.

Another might have reliable public institutions but increasing loneliness.

A third might show strong national identity but severe hostility between political camps.

OECD research on social connectedness highlights precisely this measurement problem. All OECD countries collect some information on social connections, but definitions and indicators differ considerably, making direct comparison difficult. Loneliness, social support and frequency of social interaction are not measured consistently across surveys.

World Bank work similarly recommends examining cohesion across several dimensions rather than attempting to infer everything from one indicator. Its framework distinguishes relationships within communities, across communities and between citizens and people or institutions holding power.

Measurement should also look below national averages.

A country can display reasonable overall trust while particular communities experience systematic exclusion. One region may have reliable public services while another has little confidence in the state. Young people may feel differently about institutions from older generations.

Aggregate stability can therefore hide local fragmentation.

A strong measurement system asks not only “How cohesive is the country?” but also “Who experiences belonging, trust and institutional legitimacy—and who does not?”

Frequently Asked Questions About Social Cohesion

What is social cohesion? Social cohesion is the capacity of people and groups to cooperate, maintain trust and remain connected to shared institutions despite social, cultural or political differences.

How does the World Bank define social cohesion? The World Bank describes it through shared purpose, trust and willingness to cooperate within communities, across different communities and between people and the state.

What are the main dimensions of social cohesion? Useful frameworks distinguish cohesion within groups, cohesion across group boundaries and vertical cohesion between citizens and institutions.

What is horizontal social cohesion? Horizontal cohesion concerns relationships among individuals and social groups, including trust, cooperation and interaction across community boundaries.

What is vertical social cohesion? Vertical cohesion concerns relationships between citizens and institutions, including trust, fairness, accountability and whether people believe institutions represent them.

Is social cohesion the same as social capital? No. Social capital focuses particularly on networks, trust and norms that facilitate cooperation. Social cohesion is broader and also includes belonging, inclusion, institutional legitimacy and relationships across society.

Does diversity reduce social cohesion? Not automatically. Diverse societies can be cohesive when institutions are fair, interaction crosses group boundaries and people share credible citizenship. More homogeneous societies can still be deeply divided by class, politics or region.

Does inequality weaken social cohesion? Persistent inequality can weaken cohesion when people interpret it as unfair or as evidence that institutions benefit some groups disproportionately. The UN World Social Report 2025 links inequality and economic insecurity with declining trust and fragmentation.

Why is trust important for social cohesion? Trust reduces the cost of cooperation and makes people more willing to rely on other people and institutions without verifying every interaction independently.

Is institutional trust the same as supporting the government? No. Citizens can criticise a government while maintaining trust in elections, courts or administrative procedures. Democratic trust should coexist with scrutiny.

Why does belonging matter? People are more likely to support shared institutions when they believe they are recognised as legitimate members of the society rather than permanent outsiders.

What is bonding social capital? Bonding relationships connect people within relatively similar or close groups, such as families, close friends or tightly connected communities.

What is bridging social capital? Bridging relationships connect people across social, ethnic, religious, class or other group boundaries.

Can strong communities still produce a fragmented society? Yes. Individual groups can be highly cohesive internally while distrusting or opposing one another. Society-wide cohesion requires enough relationships across those boundaries.

How do public spaces affect cohesion? Parks, libraries, schools, markets, community organisations and similar places provide opportunities for repeated interaction and can function as social infrastructure.

How does social media affect cohesion? Digital platforms can create communities and connections but can also fragment information environments, intensify polarisation or spread misinformation. The effect depends on platform design, institutions and user behaviour.

How is social cohesion measured? Researchers usually combine measures of trust, social connections, belonging, discrimination, civic participation, social support and confidence in institutions because no single indicator captures the concept completely.

What can governments do to strengthen social cohesion? Important levers include reliable public services, fair procedures, meaningful citizen participation, economic security, equal rights, accessible social infrastructure and policies that reduce persistent exclusion.

Can a dictatorship have social cohesion? It may display order or strong national identification, but suppression of disagreement makes it difficult to know whether cooperation is voluntary and durable. Cohesion should not be confused with enforced conformity.

Why does political alternation matter? A cohesive democracy allows people who lose elections to remain legitimate members of the political community and believe they can compete again through lawful procedures.

Why does crisis matter for social cohesion? Disasters and economic shocks reveal whether people trust institutions, share obligations and cooperate beyond their immediate groups. Strong pre-existing relationships and institutional credibility can improve resilience.

Social Cohesion Is the Ability to Remain a Society When Agreement Ends

The easiest time for people to cooperate is when their interests already align.

That is not the real test of social cohesion.

The test comes when interests diverge.

One region wants more public spending while another wants lower taxes. Religious groups disagree about cultural rules. Younger and older generations compete over priorities. Political parties alternate in power. Migration changes neighbourhoods. Economic shocks force difficult distributional choices.

A cohesive society does not prevent these conflicts.

It prevents every conflict from becoming a struggle over whether the other side still deserves membership.

That requires trust, but not blind trust. Citizens need enough confidence in institutions to use them while retaining the right to scrutinise them.

It requires belonging, but not cultural sameness. People need to feel that the public realm includes them without having to erase every difference.

It requires social connections, but not only inside families or ideological communities. Relationships must cross enough boundaries that society does not dissolve into internally cohesive but mutually hostile camps.

It requires economic security and fairness because solidarity becomes increasingly difficult when large groups believe the social contract works only for others.

And it requires institutions capable of processing disagreement. Courts, elections, councils, associations, unions, public administration and civil society give conflict somewhere to go other than withdrawal or violence.

Current evidence shows why this has become an urgent policy question. The UN warns of declining trust and widening social fragmentation. OECD data show substantial distrust of national governments alongside meaningful gaps in social support and connection. The World Bank increasingly treats cohesion as part of resilience because communities and institutions cannot manage conflict or crisis effectively when relationships among citizens and the state have already collapsed.

The goal is therefore not to create a society without arguments.

A society without visible disagreement may simply be a society in which disagreement has become dangerous.

The stronger objective is a civic order in which people can argue fiercely about what should happen next while continuing to recognise that they still have to decide that future together.

That is what social cohesion ultimately means: the capacity to stay in relationship when agreement is no longer easy.