Social cohesion describes the degree to which people in a society can cooperate, trust institutions and one another, and maintain a workable sense of shared membership despite disagreement. Social fragmentation describes the weakening or separation of those relationships into disconnected, distrustful or antagonistic groups.
Neither concept is simple. A diverse society is not necessarily fragmented, a homogeneous society is not automatically cohesive, and intense political disagreement does not by itself establish that social cohesion has collapsed. For conflict analysis, the important question is whether social divisions become organised, mutually hostile and difficult for ordinary institutions to manage.
This guide explains what social cohesion and social fragmentation mean, how researchers try to measure them, why the concepts matter for political stability, and how they relate to ConflictMeter’s narrower internal-conflict framework.
What is social cohesion?
Social cohesion is best understood as a multidimensional condition rather than a single attitude. Definitions vary, but most approaches include some combination of:
- interpersonal trust — whether people expect others to behave cooperatively;
- institutional trust — confidence in political, legal and social institutions;
- belonging — whether people feel included in a wider political or national community;
- solidarity — willingness to accept obligations toward other members of society;
- inter-group relations — whether differences between groups are manageable or antagonistic;
- participation — engagement in civic, community or political life;
- legitimacy — acceptance that disputes can be settled through shared rules and institutions.
A cohesive society does not require consensus. Democracies depend on disagreement. Cohesion concerns whether disagreement can remain compatible with cooperation and a shared political framework.
What is social fragmentation?
Social fragmentation refers to the weakening of connections across society and the growth of separation between groups, institutions or communities.
Fragmentation can take different forms. It may be geographic, economic, political, ethnic, religious, generational, informational or cultural. What matters analytically is not the existence of differences but whether those differences become socially or politically isolating.
Possible signs include:
- declining trust between major groups;
- political identities becoming increasingly exclusive;
- parallel information environments with little common factual ground;
- declining confidence in institutions seen as neutral arbiters;
- reduced interaction across social or political boundaries;
- growing belief that rival groups are illegitimate or threatening;
- withdrawal from shared civic institutions.
Social cohesion is not the same as homogeneity
This distinction is essential. Diversity does not equal fragmentation.
Societies can contain substantial ethnic, religious, linguistic or cultural diversity while maintaining strong institutions, high interpersonal trust and effective mechanisms for resolving disputes. Conversely, relatively homogeneous societies can become intensely polarised around class, region, ideology, religion or political identity.
Conflict analysis therefore should not infer cohesion from demographic similarity or fragmentation from demographic difference. The relevant question is how groups relate to one another and to common institutions.
How does social cohesion relate to political stability?
Higher cohesion can make political systems more resilient because disputes are more likely to remain within accepted institutions. People may strongly oppose one another while still believing elections, courts, legislatures, negotiations or peaceful protest are legitimate ways to resolve conflict.
When cohesion weakens, several stabilising assumptions can erode:
- rivals may be seen as enemies rather than legitimate opponents;
- political losses may be viewed as existential rather than temporary;
- institutions may be treated as captured by one faction;
- compromise may be interpreted as betrayal;
- rumours and hostile narratives may circulate more easily within separated information networks.
These conditions can increase political risk, but they still do not make conflict inevitable. Institutional quality, leadership, economic conditions, mobilisation capacity and triggering events also matter.
What causes social fragmentation?
Fragmentation is usually cumulative. Common contributors can include:
- persistent political polarisation;
- rapid economic or demographic change;
- regional inequality;
- segregated communities or institutions;
- weak confidence in government or media;
- highly partisan information environments;
- perceived unfairness or status loss;
- political strategies that reward mobilisation around group identity;
- major crises that distribute costs unevenly.
The same factor can have different effects in different countries. Economic change may be absorbed through trusted institutions in one society and interpreted as evidence of deliberate exclusion in another.
How is social cohesion measured?
No single indicator captures cohesion. Researchers typically combine several types of evidence.
Survey evidence can measure interpersonal trust, confidence in institutions, perceived discrimination, attitudes toward other groups and national belonging.
Behavioural measures may examine civic participation, association membership, voting, volunteering or interaction across group boundaries.
Institutional measures can assess political inclusion, rule of law, government effectiveness or the perceived fairness of public institutions.
Network and information measures can examine whether communities occupy separate media ecosystems or whether political communication is highly factionalised.
Conflict and event data can capture demonstrations, communal clashes or political violence, although these are better treated as outcomes or activation signals than as pure measures of cohesion.
Because the concept is broad, a credible index should explain which dimension it measures rather than presenting one variable as “social cohesion” in its entirety.
Social cohesion versus political polarisation
Polarisation and cohesion overlap but are not identical.
Political polarisation concerns distance or antagonism between political camps. A society can be politically polarised yet retain strong institutional legitimacy and non-political social ties. In that case cohesion may be strained without collapsing.
Fragmentation becomes more serious when political divisions align with multiple other boundaries—social, geographic, ethnic, religious or informational—and when trust across those boundaries declines.
Social fragmentation versus narrative fragmentation
Narrative fragmentation is narrower still. It concerns the extent to which people inhabit incompatible or separated public narratives and information environments.
Narrative fragmentation can contribute to wider social fragmentation, but the two should not be treated as synonyms. People may consume different media yet continue to cooperate socially and trust shared institutions. Conversely, severe group distrust may exist even where people consume much of the same information.
Why social cohesion matters in conflict analysis
Large-scale internal conflict generally requires more than dissatisfaction. It requires political mechanisms through which grievances become organised, polarised and difficult to contain.
Social cohesion matters because it affects those mechanisms. Strong cross-group ties and legitimate institutions can provide channels for compromise. Weak ties and mutually hostile group identities can make mobilisation easier and compromise harder.
But cohesion is not a deterministic “peace variable”. Strongly cohesive groups can mobilise against other cohesive groups. Internal solidarity within factions may increase at the same time that cohesion across society declines.
For that reason, conflict analysis often needs to distinguish within-group cohesion from between-group cohesion.
How ConflictMeter treats related concepts
ConflictMeter does not include a single variable called “social cohesion”. Instead, related evidence appears across several theoretically distinct parts of the model.
Factionalism addresses organised political and group division. Loss of faith in normal politics concerns confidence that existing political mechanisms can resolve major grievances. Narrative fragmentation is retained as a separate bounded accelerant rather than being treated as equivalent to conflict.
The model also separates these structural pressures from Activation, because a fragmented society is not necessarily mobilised and a highly mobilised society is not necessarily socially fragmented in every dimension.
The full architecture is described in the ConflictMeter Methodology.
Why ConflictMeter does not equate diversity with conflict
ConflictMeter’s public methodology deliberately avoids the claim that demographic diversity itself causes civil conflict. The relevant mechanisms concern status change, political downgrading, factionalism, institutional confidence and mobilisation.
This distinction matters analytically and ethically. A demographic characteristic is not a political outcome. The same demographic pattern can coexist with very different levels of political trust, inclusion, mobilisation and conflict across countries.
How social cohesion fits into political risk analysis
For country and political risk analysis, cohesion is one layer among several. Analysts may need to examine:
- whether political competition remains legitimate;
- whether institutions are trusted across major groups;
- whether group boundaries are becoming politically salient;
- whether mobilisation is increasing;
- whether economic or policy shocks are reinforcing those divisions.
A useful assessment therefore asks not only “how divided is society?” but which divisions matter politically, how organised they are, and whether institutions can still manage them.
Can social cohesion improve?
Yes. Cohesion is not fixed. It can improve through credible institutions, cross-group cooperation, successful conflict resolution, economic recovery, shared civic projects and political incentives that reward coalition-building rather than permanent antagonism.
It can also deteriorate rapidly after crises or polarising events. That is why trend information is often more informative than a one-off snapshot.
Frequently asked questions
What is social cohesion in simple terms?
Social cohesion is the ability of people and groups in a society to coexist, cooperate and resolve disagreements within a broadly shared institutional framework.
What is social fragmentation?
Social fragmentation is the weakening of shared ties and the growth of separation, distrust or antagonism between groups, communities or institutions.
Does diversity reduce social cohesion?
Not necessarily. Diversity and cohesion are different concepts. The key questions concern trust, inclusion, group relations and the legitimacy of shared institutions.
Is polarisation the same as fragmentation?
No. Polarisation describes distance or hostility between camps. Fragmentation is broader and can include social separation, institutional distrust and disconnected information or community networks.
Does low social cohesion cause civil war?
No single cohesion measure can predict civil war. Low cohesion may contribute to conditions associated with conflict, but mobilisation, organisation, institutions, security dynamics and triggering events also matter.
Explore related ConflictMeter analysis
See the Methodology, compare countries in the Rankings, or browse the Research Archive for country and thematic analysis.
Go beyond the headline score
A headline reading raises a question. A Country Intelligence Report helps you investigate it: explore the structural drivers, historical comparisons, alternative interpretations and evidence towatch for your chosen country. Choose from 81 country-specific PDFs, at £49 each with no subscription.