Interdependence explains how and when crises impact social cohesion Authors: Jesse Koster, Martin Lang, Paul A M Van Lange Published: 2026 ABSTRACT AbstractCrises may bring people together or drive them apart. However, it is not clear whether crises undermine or enhance social cohesion or how such effects might develop over time. We compare four types of crises—natural hazards, pandemics, economic downturns, and wars—to explain their effects on social cohesion from an interdependence theoretical perspective. We argue that social cohesion depends on how much people depend on each other and the government, the (un)equality of these dependencies, and whether people’s interests conflict or align. Based on our theory-guided review, we conclude that the four crises are likely to differ in how (and when) they might affect social cohesion within societies. These insights may help societies better prepare for future crises by clarifying when they may heighten cohesion or division.Subject terms: Human behaviour, Social sciences, Psychology Introduction Crises are a basic fact of social life. Notable examples include natural hazards such as wildfires, floods, or earthquakes; the COVID-19 pandemic; economic crises such as the 2009 Great Recession; and international conflicts and wars affecting the lives of millions of civilians across the world. Apart from material, financial, and sometimes physical harm, they often pose challenges to health and well-being, and can have implications for social connection, interpersonal trust, feelings of security, and confidence in social institutions1,2. While the devastating consequences of crises are well-documented, it has been less clear whether and how crises affect social relations. In particular, does a crisis generally enhance or undermine social cohesion? And do different crises affect social cohesion differently? Some scholars argue that crises may foster cooperation and social cohesion3,4, while others argue that crises may undermine political trust and social identities, thereby reducing social cohesion1,5,6. Yet, previous studies paid little attention to the distinctive aspects of different crises that affect the societal dynamics during a crisis. Understanding the trajectories and time horizons over which crises affect social cohesion is crucial for both advancing our theories of societal well-being, and offering insights and suggestions on how to best prepare for, cope with, and resolve different crises. The goal of this article is to provide insight into the similarities and differences among crises that could help explain their differential effects on social cohesion. In doing so, we focus on two broad questions. First, what explains the different effects that crises may have on social cohesion? And second, what explains the ebbs and flows in cohesion over time during a particular crisis? We adopt an interdependence analysis to address these questions, specifying the nature of interdependent relationships between individuals, groups, and society7,8. Specifically, we argue that a crisis may be characterized by (change in) three interdependence features, including degree of interdependence (how strongly individuals/groups depend on each other), asymmetry in dependence (whether one group is more dependent on another group), and the degree to which their interests align versus conflict. We close by applying the interdependence analysis to four distinct crises, including pandemics, natural hazards, economic crises, and wars, to predict and understand how these crises may differentially affect social cohesion. We seek to review empirical evidence relevant to our interdependence theoretical analysis of crises and their effects on social cohesion. At the same time, we came to realize that there is a lack of research to support (or disconfirm) our proposed hypotheses. Therefore, our goal is not only to review scientific findings, but also to advance theory by providing several predictions that can be tested in future research. Some of our theory-based predictions are addressed in specific time-related patterns of how social cohesion may develop before, during, and after a crisis. We provide an analysis of each crisis and will address research paradigms to test these predictions. Social cohesion in times of crises The importance of social cohesion While crises affect diverse facets of social life, we focus on the effects on social cohesion since cohesion creates the fundamental fabric of a well-functioning society. Specifically, we conceptualize social cohesion as a condition of a society characterized by both subjective perceptions (operationalized as feelings of trust, solidarity, and shared identity) and objective behaviors (operationalized as participation in communal life, prosocial actions, cooperation, and norm compliance)9. Importantly, these features operate at two levels: horizontal, referring to trust and solidarity among individuals and groups within society, and vertical, referring to trust in and perceived legitimacy of social and political institutions. This definition differs from policy-focused approaches, which often treat liberal values such as multiculturalism or equal rights as indicators of cohesion9,10; while such policy-oriented conceptions are important from a governance perspective, our focus is on the everyday, lived experience of cohesion and on how it is shaped by different crises. Similarly, social cohesion is not reducible to a single component, such as interpersonal trust or to closely related constructs such as social capital or institutional confidence. Rather, it encompasses a broader constellation of social relations, psychological orientations, and cooperative behaviors that sustain societal functioning. In this conceptualization, social cohesion encompasses both an identity component that recognizes the societal unit to which cohesion refers and the positive behaviors and feelings directed toward members of that unit11. Importantly, while cohesion is expressed in individual attitudes and behaviors, it is fundamentally a societal-level phenomenon, emerging from patterns of interpersonal and person-institution interactions. Individuals differ in how strongly they experience cohesion, but the construct gains meaning when aggregated to characterize a group, community, or society. More formally, we define social cohesion by the degree to which relationships within and between individuals, groups, and institutions in a society are characterized by (a) strong levels of trust, (b) a sense of shared identity, (c) consensus about basic norms and values, and (d) cooperative interaction (and the relative absence of noncooperative interaction), including involvement in society at large. Operationalizations of social cohesion can include well-established self-report scales of trust, specific questions about participation in communal and societal life, as well as a variety of economic games and other decision tasks that seek to capture trust and cooperation within and between groups and institutions in a society12–14. Social cohesion may play a critical role in determining how well a society recovers during and after crises. For instance, during the COVID-19 pandemic, higher levels of perceived social cohesion were associated with better mental health recovery following lockdowns6. Similarly, communities with stronger social cohesion are more confident that their community will recover quickly after natural hazards15. These effects are likely driven both by horizontal trust and helping behaviors, and by vertical trust in the institutions coordinating the response. However, while pre-existing levels of social cohesion are important, our focus is on how different crises shape cohesion and its individual components. An interdependence analysis of crises A crisis has been defined as “an event that has the potential to cause a large detrimental change to the social system and in which there is a lack of proportionality between cause and consequence”16. Characteristics of a crisis include its unexpected nature, urgency, associated uncertainty, and societal impact17. We propose that to understand the differential impact of crises on social cohesion, the complexity of different crises needs to be described by the context, as well as the perception of the crises18. Some crises call for concerted effort to help reduce the consequences of a crisis, while other crises may cause a divide between people, or at least groups of people, within society. For example, during natural hazards, people may quickly come to realize that those affected by the disaster “are in this together” and so they need to help each other. Such solidarity, as well as collective resilience, has been documented across qualitative, survey, and field-based research, including the strengthening of shared social identities, and increases in aid and support among the affected communities19–22. In contrast, an economic crisis may pose different challenges since some people are much more affected by the crisis than others. Likewise, the existence of a crisis, its potential consequences, and the seriousness of those consequences, are not always a matter of society-wide agreement. For example, the climate change crisis, as a slowly unfolding threat, initially faced considerable skepticism, with many not being convinced that we are facing a crisis at all23–25. Such perceptions may be persistent, and they may cause a divide in society among those who believe in climate change and those who do not. The context and perception of different crises are critical to promoting, undermining, or not affecting social cohesion at all, or affecting only some of its aspects (e.g., horizontal or vertical). Thus, our basic assumption is that a greater understanding of context and perception helps us predict and understand how crises affect social cohesion, and how these effects develop over time. Put differently, how can we “diagnose” the context of the crisis and how people come to see it, to understand the crisis’s consequences for social cohesion? The four system-wide crises we study are inherently social. Individuals are interdependent in effectively dealing with the crisis, in that they require coordination and cooperation from individuals within their groups and societies. One theoretical assumption we advance is that different crises present different collective-action dilemmas that affect how people coordinate and cooperate based on their social ties and governing institutions. Understanding the interdependence between individuals and their social ties and institutions is essential to explaining how the differential effects of crises impact social cohesion, especially how people feel and behave towards each other and the governing institutions. One promising approach to this end is interdependence theory, which offers a framework for the essential dimensions that define social situations7,8,26. Three dimensions are particularly relevant in the context of crises. The first dimension refers to interdependence, the degree to which individuals influence one another’s outcomes. For example, during war, people may become more dependent on one another for survival, whether on the battlefield or at home in war zones. Greater levels of interdependence may increase horizontal social cohesion, including a willingness to trust and help one another26. Yet, such interdependence may also negatively impact social cohesion, especially if it is not voluntary and forced top–down. For example, during COVID-19, “flattening the pandemic wave” relied on citizens to police each other (e.g., to get vaccinated, to wear a mask), which may breed mistrust and suspicion27,28. The second dimension focuses on the asymmetry in dependence, raising issues of differences in vulnerability and power. For example, compared to people from economically marginalized groups, the rich and resourceful are typically better able to take care of themselves and their family2,29; or during a pandemic, vulnerable groups—such as the elderly and individuals with respiratory conditions—face greater health risks. Such imbalances may be detrimental to horizontal social cohesion, especially when the less-affected group needs to significantly limit their quality of life to help the affected/dependent groups. Finally, the degree of corresponding versus conflicting interests is critical to understand the crisis’s impact on social cohesion. In some crises, people’s interests are more likely to align, for example, when mutual help and support are essential for coping with a crisis, such as when dealing with the initial devastating effects of a natural hazard. In the aftermath of natural hazards, individuals are collectively motivated to protect their health, ensure the safety of their compatriots, and minimize damage to their homes and infrastructure. However, crises may also foster conflicting interests. The obvious example may be found in war, where opposing groups have directly conflicting goals. Less obvious examples include those who do not wish to participate in pandemic safety protocols due to perceptions of personal freedom, or those who exploit the crisis for personal gain by selling essential goods at outrageously high prices. Importantly, crises also have a significant impact on the nature of interdependence between the people and the government (vertical social cohesion). Similar to dependence between people, the level of dependence between citizens and the government typically increases as a result of a crisis. People rely on the government to implement policies to navigate crises, whereas the government relies on its people to follow the law and volunteer to help out. The asymmetry in dependence often shifts, with the people becoming significantly more dependent on the government than before, increasing the government’s power. This imbalance often leads to public resentment, causing people to protest or riot in response to the government’s increased power2,29. However, when a government is perceived as handling a crisis effectively—whether due to propaganda or genuine competence—vertical social cohesion, along with public trust in government and compliance with governmental mandates, may increase30,31. Regarding the covariation of interests’ dimension, the interests of people and the government should generally align during crises. However, conflicting interests may arise when, for example, governments believe that the interests of specific groups in society need to be prioritized or when strict measures are needed to cope well with a crisis30,31. Thus, much like interdependence between people, crises significantly affect the interdependence between the people and the government. A closer look at four crises: natural hazards, pandemics, economic crises, and wars Our main assumption is that it is unrealistic to make general claims about the consequences of crises for social cohesion, because crises differ in key dynamic features that shape social cohesion and how it develops over time. Each crisis brings about a unique pattern of interdependence among citizens that may help predict how social cohesion may be affected at the onset of a crisis, during a crisis, and after a crisis. We advance an interdependence analysis of the four crises to examine their differential impact on social cohesion and their time-related patterns. Natural hazards Looking at the effects of natural hazards (defined as natural processes and phenomena that may cause loss of life, injury, property damage, social and economic disruption, or environmental degradation32) on social cohesion, a literature review concluded that despite their destructive context, natural hazards generally foster cohesive communities33. For example, an earthquake’s aftermath was associated with increased social cohesion (assessed through several socio-economic variables such as trust and life satisfaction) in Chile, but cohesion slowly decreased when the threat became less salient11. The same effect was observed in a controlled lab-in-the-field study conducted in Houston, USA, where evacuees from Hurricane Katrina displayed higher levels of cooperation in dictator and public goods games compared to “normal populations” used in other studies34. However, in contrast to single studies, a meta-analysis investigating the impact of earthquakes over 185 countries found that earthquakes increase the likelihood of intrastate conflict (which may reflect both horizontal and vertical cohesion), and that these effects are larger for earthquakes of higher magnitude35. Another meta-analysis yielded similar results: deviations from normal temperatures and rainfall consistently predicted a higher risk of conflict36. One potential qualifier for these divergent results is the time horizon of the societal impact. While some natural hazards are temporary setbacks (floods and earthquakes) that raise prosocial action, slow-onset or protracted hazards, such as droughts, unfold gradually and tend to generate sustained competition over scarce resources. Another qualifier is the perceived immediacy and certainty of those threats: while earthquakes, floods, or fires have immediate reality propagated by media outlets, hazards with delayed consequences might be perceived as less urgent or even contested, especially when their most severe effects are anticipated rather than directly experienced (as is the case with climate-change-related hazards). From an interdependence perspective, these differential patterns can be explained by the interacting effects of strong interdependence and high degrees of corresponding (rather than conflicting) interests in the case of a short-term crisis. At the onset of such a crisis, people are typically strongly dependent on each other for urgent help. For example, after a wildfire, people may need to help provide medical care and shelter for one another. Since natural hazards often have localized disaster zones, people can proactively cooperate in minimizing the damage from natural hazards, such as by fortifying riverbanks during incoming floods. The joint engagement both manifests and strengthens the interdependence between people and can be directly observed and experienced37. At the national scale, natural hazards are distinctive in that they typically affect only specific regions of a country and do so over a limited time frame. While the mutual interdependence between those in the affected area is usually symmetrical, the rest of the nation often offers temporary help in the form of charitable donations or volunteering at the affected sites. Such temporary and affordable help not only supports the people affected by disaster but also can create a wave of solidarity at the national scale (everyone contributes a little bit), effectively strengthening social bonds within a nation and temporarily decreasing political frictions and other grievances19,33. The presumed low cost of collective action for the unaffected region also affects the perceived conflict of interests, which is rather minimal—since the contributions are voluntary. Of course, sometimes, the help is unevenly provided, especially when certain groups are disadvantaged for the benefit of the majority. For example, during flash floods, the government may open anti-flood barriers and destroy some villages to flatten the incoming flood wave and save larger cities from flooding. While such actions might lead to a greater good from the national perspective, the sacrificed regions have obviously conflicting interests. Here, trust in public institutions, as well as any potential compensation, would be crucial in minimizing the potential negative effects on both horizontal, and especially vertical social cohesion. Similarly, after the basic needs are met and people are generally safe, conflicts of interest may arise when people need to restore and rebuild collectively shared resources or public goods. Again, in such cases, the less affected may promote public funds aimed at restoring public goods, while the more affected are likely to prefer funds that alleviate their personal suffering. Such circumstances are likely to undermine trust and create tension. In summary, the asymmetry of interdependence and conflicting interests (both inter-personal and between people and the government) are exacerbated in long-term natural hazards such as droughts, water shortages, and high seasonal variations in food production. When such natural hazards affect the whole nation, or the resources are so scarce that the unaffected regions cannot afford to help, people may try to protect their own self-interests, which may erode horizontal and vertical social cohesion. The COVID-19 pandemic A large body of behavioral science evidence highlights the relevance of several core components of social cohesion for collective responses to infectious disease outbreaks. A review of 747 studies spanning 142 countries found consistent support for the role of shared social identity, prosocial norms, and trust in promoting preventive health behaviors during the COVID-19 pandemic, although effect sizes were small to moderate and varied across contexts31. In contrast, this review also found that political polarization, misinformation, and conspiracy beliefs were consistently associated with lower compliance with public health measures. These findings suggest that social cohesion often is an important facilitator of early collective mobilization during COVID-19, which is mostly in line with early theoretical work38. These general patterns become more understandable when examined through an interdependence perspective, which highlights how levels and asymmetries in dependence, alongside the costs of cooperation, may evolve during such a crisis. Some examples from specific national contexts illustrate how interdependence dynamics unfolded during the COVID-19 pandemic. For instance, a longitudinal study conducted in Romania found that solidarity levels increased during lockdown initially, then decreased as uncertainty persisted, and increased again following effective state responses and lockdown normalization39. This suggests that perceived uncertainty and effective institutions shape prosocial lockdown behaviors. The same study also found similar increases in perceived solidarity in Hungary, though not in Spain, highlighting the context-specific nature of these effects. Other studies found that social cohesion decreased during a COVID-19-related lockdown, only to increase again once the lockdown was lifted. One study, conducted in Germany, found that social cohesion, measured through trust, sense of belonging, social interactions, and social engagement, decreased during lockdown but increased again once the lockdown was lifted6. A study conducted in England showed similar results: overall levels of neighborhood cohesion were lower in June 2020—several months into the pandemic—than during any pre-pandemic period2. This reduction in perceived social cohesion was especially high in deprived communities, certain ethnic minority groups, and among those whose professions are more easily replaced. These group-level differences suggest that prolonged lockdown restrictions may increase asymmetries in dependence, as well as the cost of cooperation. From an interdependence perspective, a pandemic such as COVID-19 can be interpreted dynamically according to systematic changes of the evolving crisis. At the onset, people are overwhelmed by strong experiences, either through media exposure or through direct experiences. Although physical distance increases, people become more aware of others’ vulnerability, particularly high-risk groups such as seniors, and may be willing to temporarily decrease their life quality to engage in prosocial behavior. As a case in point, volunteering in hospitals and other organizations in China rapidly increased after the onset of COVID-1940. However, pandemics such as COVID-19 are not localized events. They affect the whole nation, and despite the asymmetrical vulnerability (elderly and chronically ill disproportionally affected), everyone is a potential victim. Given the initial insecurity, the interdependence strongly aligned with personal goals: “not only am I helping others by getting a vaccine, I am also protecting myself.” Some studies suggest increased perceptions of interdependence during COVID-19, and that these perceptions may have helped to keep social cohesion stable over time41. Nevertheless, unlike the joint action mitigating the aftermath of natural hazards, the prosocial activity mitigating the pandemic was focused on behavioral change of individuals, and less on behavioral change of well-established groups, which may have weakened perceptions of interdependence. For example, relative to a natural hazard that calls for joint activities, such as sharing resources, COVID-19 included fewer collective activities that enhance experiences of interdependence. Over time, the asymmetry in interdependence was amplified. First, younger people might have felt less vulnerable to the virus and that they needed to sacrifice more than older people. Likewise, business owners, restaurants, artists, and public services, who rely on the physical presence of many people to earn a living, had to limit their services due to governmentally imposed restrictions, leading many to bankruptcy and a severe conflict of interests42. Second, there were bigger differences in attitudes, beliefs, and ideology relevant to vaccination, with the vaccinated viewing the unvaccinated as posing risks to them and society, and the unvaccinated feeling marginalized and excluded, as their access to public spaces and travel was restricted43. COVID-19 illustrates that, particularly early on, people were faced with strong interdependence, some asymmetry in dependence, and a relatively low conflict of interests. Many perceived the situation as a shared burden in which some groups needed more protection than others, and short-term social cohesion (both as mutual help and conforming to governmental mandates) appears to have increased in the early stages of the pandemic31. As the costs of cooperation grew over time, especially for younger people, business owners, and people strongly opposed to vaccination, both horizontal and vertical social cohesion weakened. While cohesion increased within some groups, between-group cohesion declined, accompanied by rising polarization44. Economic crises While research on economic crises and social cohesion is limited, the effects of economic crises on social cohesion appear to be negative. Economic crises place financial strain on individuals, which is associated with lower levels of interpersonal trust, especially toward marginalized groups (such as immigrants)45. In addition, economic crises are related to higher unemployment rates, which in turn reduce social interactions among men, although no such effects were observed for women46. In Spain, the 2008 economic crisis had a particularly detrimental impact on the social and economic security of vulnerable groups29. Policy responses involving reductions in social spending weakened inclusion policies and, in turn, the institutional basis of social cohesion. Because the association between economic crises and social cohesion has received little empirical attention, it is challenging to analyze this type of crisis in terms of interdependence theory. Unlike natural hazards, economic crises typically cannot be immediately mitigated by collective effort, and there often is no joint endeavor for people to boost their cohesion. Like pandemics, economic crises require individualized responses, but since there is no life threat for the whole nation, those individualized responses do not amplify the helping of others and building cohesion. While people become more interdependent, relying on their social networks as safety nets, dampening the impacts of economic crises, the redistribution of resources in terms of loans and credits, together with high unemployment, may severely damage interpersonal trust and social cohesion if those loans cannot be paid on time. The key consequences of economic crises are increases in asymmetry in dependence. In most societies, the effects of an economic decline are more disruptive for the less resourceful than for the affluent. For example, during the Great Recession of 2007–2009 in the United States, house prices fell by 21% nationwide, leaving people with homes worth less than the mortgages they were paying. While the wealthy were able to weather the crisis and recover their assets, the less wealthy lost around half of their net worth47. These effects were propagated into the asymmetry of dependence on the government, where the low-income and vulnerable households needed state-mediated help. It could be further deduced that whenever there is an asymmetry in the dependence on the government that redistributes public funds, conflicts of interest would be amplified between those who need help and those who are well-off and might prefer directing public funds into other services. Such disparities may lead to scapegoating, particularly when the groups most affected by the crisis share other salient characteristics, such as ethnicity, employment sector, or immigration background. The less-affected people may come to view the affected as “outgroups” who unnecessarily drain the public resources. These processes may lead to a belief in a just world: “People deserve what they get, and get what they deserve”48,49. This belief may even be shared among the less resourceful50. Taken together, from an interdependence perspective, an economic crisis increasingly undermines social cohesion. The pattern unfolds linearly over time, and because of mechanisms such as belief in a just world, it is possible that the effects of the crisis on the least resourceful are somewhat invisible but most detrimental – and therefore especially challenging from a governmental or policy standpoint. War The social consequences of war and international conflict are not very well-understood51. However, recent findings provide some clues. For example, one study found that individuals who experienced more intense war violence during the Sierra Leone civil war were more likely to join political and community groups52. Another study from Uganda found that past violence predicted increased political engagement among ex-combatants53, and communities exposed to civil-war violence in Nepal showed greater cooperation in economic games54. And perhaps most importantly, a meta-analysis of 16 studies found that exposure to war violence is positively associated with a great variety of indices of social cohesion, including social group participation, community leadership/participation, prosocial behavior in experimental games, as well as knowledge and interest in politics3. Experimental studies further showed that intergroup conflict may boost in-group cohesion, whilst also increasing hostility toward outgroup members55–57. Yet, other research found that war and conflict mostly harm social cohesion1. For example, a study in Israel found that combat exposure predicted significantly lower political participation among ex-combatants58. Furthermore, when victims of a civil conflict played a trust game, they were less trusting of their neighbors59 and trust strongly decreased during and after the civil war in Uganda, after which it slowly increased back in the ten years following the war5. Moreover, people who were exposed to violence during World War II as children showed significantly lower generalized trust as adults60. Thus, wars can have different effects. From an interdependence perspective, one key question is whether a war always enhances the level of interdependence in combination with alignment of interests—two interdependence ingredients that are expected to promote social cohesion. Most wars are probably characterized by these two features. Beliefs and actions aimed at protection and eliminating threat, or even seeking justice and retaliation, bring about a mindset of social cohesion, especially when the joint action (helping others/cooperating) is immediate and localized. Similar to natural hazards, people directly depend on each other during combat, and this interdependence is tangible. At the same time, this may be a somewhat optimistic diagnosis. Wars can also promote strong conflict of interests, especially when “free-riding” and collaboration with the enemy become realistic61. Moreover, a conflict of interest may be promoted by differences in historical experiences and societal or political beliefs relevant to a particular group identity, such as the long-lasting social conflicts as a result of different wartime alliances following the civil war in Mozambique62. If there is optimism left, it is for post-war restoration and rebuilding that calls for strong boosts in social cohesion. It likely turns into a pattern of interdependence similar to the one that people may also experience at the onset of a natural hazard. Thus, war is likely to undermine social and political trust, although group membership appears to play a significant role. Regarding the effects within communities, war sparks an initial increase in social cohesion out of sheer necessity. Longer-term effects of war on social cohesion and trust within communities are difficult to generalize, as they depend on various factors, including the war’s duration, the level of psychological distress experienced, the proximity of the enemy, and the strategies and responses of both the community and the government1. Furthermore, similar to pandemics, wars may shatter existing communities and social relationships, due to forced migration and displacement, negatively impacting social cohesion built on generations of interdependent relationships. With a view of all four crises, we conclude that the time-related effects of a crisis on social cohesion are quite context-dependent. Sometimes this is even true for a specific crisis. Box 1 outlines how the interdependence perspective can be examined empirically across the different crisis contexts mentioned in this article. The interdependence lens through which we have reviewed the literature has generated several testable predictions about how different crises shape social cohesion over time. As different crises differ substantially in terms of impact, duration, scope, and institutional involvement, our predictions may not be tested through a single method. Thus, we propose a strategy of converging operations, combining several different methods from multiple disciplines. Regarding integrative tests, we recommend scenario experiments to establish causal (although tentative) evidence regarding the association between crises and social cohesion. This includes trust games, public goods games, and resource dilemmas often used in behavioral sciences, which have been shown to capture variation in cooperation and trust across crisis contexts1,83. To complement integrative experiments, crisis-specific tests can examine social cohesion in real-world settings using surveys, structured interviews, or experimental games conducted in crisis-affected populations51. These studies can capture how cooperation, trust, and perceived dependence vary across crises and over time, including across countries, and can directly assess interdependence features, such as mutuality of dependence and corresponding/conflicting interests, using existing measures84. Last but not least, historical, archival, and ethnographical approaches may help illuminate associations between various interdependence features, social cohesion, and different types of crises, particularly with respect to their longer-term impact during and after a crisis85. Reflections and extensions Building on the evidence reviewed, we summarize our core theoretical predictions in Fig. 1. The purpose of this model is to initiate a theoretical discussion of the basic features of crises that would help us predict how a particular crisis may affect social cohesion and its development over time. We should acknowledge that our analysis represents an oversimplification of the specific predictions we advance in this article. An illustration of how natural hazards, pandemics, economic crises, and war may affect social cohesion at the onset of a crisis, during a crisis, and after a crisis. Note: Shaded background zones (blue = pre-crisis, amber = mid-crisis, green = post-crisis) delineate the three temporal phases; vertical dotted lines mark the onset and end of a crisis; horizontal lines represent pre-crisis baseline social cohesion. Each crisis type is represented by two thick lines—solid and dashed—corresponding to the two empirically best-supported contextual conditions identified in the reviewed literature. Thin lines represent empirical uncertainty, as well as possible contextual variation in this dynamic. All trajectories were generated from synthetic data for illustrative purposes only. First, the empirical literature provides tentative evidence, and it is especially challenging to use the evidence to predict time-specific patterns in the development of social cohesion. For example, even though there is growing evidence that natural hazards can bring about increases in social cohesion, this view is not universally accepted. Some studies indicate that the effects on social cohesion may depend on policies regarding resource distribution, where the more vulnerable (and less affluent) experience the more severe consequences (even if social cohesion within their groups is increasing)29. At the same time, there is more unequivocal support for the idea that a pandemic such as COVID-19, after an initial period of increased cohesion, undermines social cohesion, even though communication and policy by the government may play an important role. Second, while this research focused on individual crises, crises can also occur simultaneously, and sometimes even cause one another. An example is the massive economic recession following the COVID-19 pandemic, which impacted employment rates across the globe63, and increased long-term social unrest64. Another example is the Russian invasion of Ukraine, which started in the late stages of the COVID-19 pandemic, creating extreme strain on Ukraine’s healthcare system and resources. In addition, the simultaneous crises affected mental distress: increasing depression, social withdrawal, and anxiety. Relating these findings to social cohesion, the onset of a new crisis while a society is still recovering from the last is likely to negatively impact social cohesion due to an increased scarcity of resources, deepened wealth inequalities, and heightened mental distress. Third, the social effects of crises often depend on governmental responses. Successful state reactions reduce the harmful effects of a crisis, and thereby their effects on social cohesion. For example, by reducing uncertainty during COVID-1939 or by reducing wealth inequality by adopting more progressive policies29. In addition, open communication by acknowledging responsibility and fostering dialogue may minimize reputational damage and public blame towards the government65. While the direct effects that effective crisis responses have on social cohesion are less clear, they may reduce the harmful effects inherent to crises, and have a positive impact on trust in government, thereby increasing—or preserving—social cohesion. Fourth, experiences of a crisis may be influenced by risk perception66. For example, at the onset of COVID-19, people fell prey to unrealistic optimism by systematically underestimating the severity of the pandemic31. Unfamiliarity and limited experience with a particular event can lead people to systematically underestimate risks for that event in the future67. During a crisis, such underestimation of risks may become weaker because novel (and vivid) information is often impactful. Not only may people increasingly observe or experience health-related consequences, but people may also increasingly adapt by, for example, wearing face masks, which may also help to recognize risks. At some point, some people might systematically overestimate risks. One key element may be social influences conveyed by media or “bubbles” where risk aversion and conversation can magnify risk perception. For example, at the turn of the century, many people expected a computer breakdown due to the millennium bug (or Y2K Bug), the widely shared expectation that computers were not able to transition from 1999 to 2000, which was assumed to potentially cause large-scale problems for hardware banking, medical devices, and the overall economy68. Finally, crises and social cohesion interact in a dynamic, bidirectional manner that is sensitive to initial conditions. For instance, societies characterized by high baseline trust in institutions might better comply with top–down imposed restrictions during a pandemic, and such coordinated adherence can accelerate the mitigation of the crisis (e.g., rapid suppression of infection rates). The resulting low mortality may reinforce social cohesion through a shared perception of collective efficacy and mutual support. Conversely, in highly polarized societies, government-imposed restrictions may be met with resistance or outright noncompliance, driven by distrust or hostility toward governing bodies. This lack of coordinated action can exacerbate the crisis and erode the very social bonds required for an effective collective response. Broader implications Our theoretical framework suggests that societies need to focus on the degree to which individuals and groups learn from previous crises, and especially whether individuals, communities, and governments provide suitable circumstances for social cohesion to grow and thrive. The circumstances for successful crisis management are reducing asymmetry in dependence and reducing conflicts of interest, while the level of interdependence should be relatively high so that people retain a feeling that “we are in this together.”7. Typically, crises increase the level of dependence between people, while the asymmetry in dependence may often shift in favor of certain societal groups. While interdependence theory does not provide a definitive answer, it highlights the importance of social cohesion and the broader notion that fluctuations in social cohesion may be rooted in changes in interdependence that individuals and groups within societies face69. Since crises vary in their impact on different individuals and groups within a society—some unfold as widespread disruptions, while others disproportionately affect specific segments of the population or particular regions—they pose different challenges to maintaining social cohesion. Shorter, localized crises where people can cooperate to mitigate the crisis—or when some segments of society can help out—are more likely to uphold cohesion despite the asymmetry of interdependence. At the same time, scattered crises that affect the whole nation unevenly are more likely to exacerbate existing divides, especially if the required response lies in the hands of individuals rather than cooperative, localized units (e.g., building an anti-flood barrier vs helping the national economy). These issues are related to a broader question of whether a crisis involves top–down enforcement or bottom–up enforcement. Some crises involve top–down enforcement (e.g., everyone has to wear a mask in public at the onset of COVID-19), whereas other crises call for immediate bottom–up initiatives (e.g., the local help and coordination that is needed after a natural hazard). Crises that involve bottom–up enforcement may be more likely to promote social cohesion as cooperation is largely voluntary, while top–down enforcement of cooperation may create more polarization, protest, and reactance70,71. Decreases in social cohesion during (and sometimes after) a crisis may help individuals and governments to recognize the importance of social cohesion in a society. Programs that foster social contact between neighbors may be critical to build trust and cooperation in social networks72,73. Similarly, creating community facilities may be critical, an example being the installation of warehouses to store water, food, and other essentials to cope with the days after an emergency (such as an electricity breakdown). Likewise, the provision of vaccines, the availability of facemasks, and other resources, are important to the communities and organizations in which people mostly live their lives. We also emphasize the importance of key community-building activities that people from differing groups can share in order to avoid “bowling alone” and build social safety nets that increase the resilience of individual households. For example, sports such as soccer or basketball, even as spectators, are clearly a binding force that brings people together74,75. The feeling of a cooperative, helping community should be nurtured in local environments. A second implication is that some consequences of crises are exceptionally long-lasting. Consider post-traumatic stress after warfare, increased mental issues as a result of isolation during a pandemic, and a decreased quality of life following a natural hazard76–78. In some cases, a crisis also leads to long-lasting positive outcomes, particularly when the response to a crisis leads to better policies and laws for a community’s inhabitants18,79. Some enduring effects of crises involve threats to intergroup relations such that intergroup trust and cooperation decline, whereas distrust, hostility, and even hate may grow and persist1,78. This is most obvious in the crisis of war. However, less intense forms of intergroup conflict may even come into being when the conflict of interest is far less evident. A compelling case in point is friction and hostility between the vaccinated and unvaccinated groups during COVID-19, or when a natural hazard creates a scarcity of resources. Note that these conflicts may continue long after the crisis is resolved, but that issues of distrust and lack of cohesion may still remain serious challenges for societies. We also want to outline the need for an interdisciplinary approach to understanding crises and social cohesion. Our interdependence analysis focuses on features such as degree and symmetry in dependence and the alignment of interests among citizens and between citizens and authorities. This is grounded in game theory and social exchange theory, and thus can be viewed as rooted in economics, sociology, and psychology8. However, there are various theories or models that are related to interdependence theory but focus on alternative, interdisciplinary concepts relevant to understanding causes and consequences of crises. One approach is the social-ecological frameworks that focus on variation in natural resources, climate, and its opportunities and vulnerabilities80. For example, variation in climate can bring about specific challenges (seasonal variation in temperature or rainfall) that may shape how strongly people need to work or plan together81. Geographical history can also bring about specific hazards (e.g., drought versus flooding) that might inspire a specific culture of getting together in the form of rituals and narratives that may be binding82. A second approach may consist of cultural-evolutionary frameworks, such as those focusing on a cultural analysis of tight versus loose norms83. For example, some societies were less strict in norm enforcement at the onset of COVID-19, which helped explain the relatively high number of cases and deaths during that early period of time84. A third approach derives from social-political frameworks that focus on governance, institutions, and wealth. Examples of the latter include theorizing on social capital72,73 or economic governance and resource management85. For example, a crisis such as natural hazards may be best approached when authorities orient themselves to local management (e.g., at the level of districts) whereas a pandemic or an economic crisis requires a national or international orientation. Concluding remarks To fully understand crises such as natural hazards, pandemics, economic downturns, and wars, it is important to acknowledge that crises differ in the nature of the interdependence that individuals and groups face at the onset of the crisis, during the crisis, and after the crisis. Such diversity speaks against one-size-fits-all approaches to understanding and potential solutions. Our interdependence theoretical approach also underlines that government policies and communication strategies are crucial to help reduce asymmetry in dependence and conflict of interests to create optimal circumstances that facilitate, rather than undermine, social cohesion. Communities can also make contributions by initiating local activities. On the more positive side, crises can be catalysts for societal change79. While crises are bound to have adverse effects on the welfare of individuals and communities, sometimes the longer-term consequences may be positive. The realization that people need each other to effectively cope with natural or human-made threats is one that may serve as a strong reminder that the most serious problems in society can only be solved together. The studies cited in Box 1 are listed here for ref. 86–88. Supplementary information Transparent Peer Review file Author contributions J.K. drafted the manuscript and contributed to the conceptualization, as well as to several revisions of the paper. M.L. contributed to the conceptual development and writing and re-writing and to the figures. P.A.M.v.L. proposed and supervised the project, contributed to the writing and revisions, as well as to the conceptual framework. All authors contributed to revisions and approved the final manuscript. Peer review Peer review information Communications Psychology thanks Evangelos Ntontis and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Marike Schiffer. A peer review file is available. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Jesse Koster, Paul A. M. van Lange. Supplementary information The online version contains supplementary material available at 10.1038/s44271-026-00471-2. References Associated Data Supplementary Materials Transparent Peer Review file