Intergroup and intragroup dimensions of COVID-19: A social identity perspective on social fragmentation and unity

COVID-19 is a challenge faced by individuals (personal vulnerability and behavior), requiring coordinated policy from national government. However, another critical layer—intergroup relations—frames many decisions about how resources and support should be allocated. Based on theories of self and social identity uncertainty, subjective group dynamics, leadership, and social cohesion, we argue that this intergroup layer has important implications for people’s perceptions of their own and others’ situation, political management of the pandemic, how people are influenced, and how they resolve identity uncertainty. In the face of the pandemic, initial national or global unity is prone to intergroup fractures and competition through which leaders can exploit uncertainties to gain short-term credibility, power, or influence for their own groups, feeding polarization and extremism. Thus, the social and psychological challenge is how to sustain the superordinate objective of surviving and recovering from the pandemic through mutual cross-group effort.

outlines some implications of our recent and current work to suggest how multilevel dynamics within and between groups, mediated through social identity, are impacting the effects of the pandemic on society.

Framing the Pandemic in Terms of Social Identity Theory
The social identity perspective (Abrams & Hogg, 2010;Tajfel & Turner, 1979) holds that group memberships at different levels of inclusiveness, and the identities they evoke, become salient in different contexts. The subjective prominence of these is affected by both cognitive (Turner et al., 1987) and motivational factors, such as self-enhancement (Tajfel & Turner, 1979) and desire for optimal distinctiveness (Brewer, 2003). Whether a particular social identity provides a focal basis for intentions and behavior is also responsive to external conditions. These conditions may be manifested in everyday interactions or norms, such as when others routinely treat one as belonging to a particular category or group. They may also be manifested institutionally, such as the presence of laws or public policies that focus on the actions of one's ingroup or a relevant outgroup (e.g., when all people from a particular college, city, or region are bound by stricter measures due to . Such conditions also expose realistic differences of status, power, or interest between groups (Sherif, 1966). How people interpret the implications of their group's situation is therefore framed not only by each group's objective or material circumstances, but also by intergroup relations-the comparisons and relationships between groups.
People's social identities-their sense of selfhood defined by their group memberships-provide psychologically meaningful frames within which to define themselves and to navigate the social world . Thus, people desire a degree of certainty around their social identity (Hogg, 2007;Hogg & Abrams, 1993). If the external environment becomes less stable and less predictable, self-certainty is likely to become more focal but also more vulnerable. Uncertainty about one's self and identity motivates stronger group identification and it also increases people's quest for positive group and intergroup outcomes associated with their ingroup memberships (Hogg, 2007(Hogg, , 2012; see also Kruglanski et al., 2021, for further discussion on threats to self).

Social Identity and Intergroup Relations at the Onset of Crisis
People tend to come together in time of crisis. For example, solidarity and social cohesion within groups often arise in the aftermath of natural disasters or mass tragedies (e.g., Hawdon & Ryan, 2011). During the ongoing COVID-19 pandemic, people spontaneously organized "mutual aid groups" in many countries (Ntontis & Rocha, 2020). Such increases in solidarity may arise from identity fusion, the feeling of "oneness" with the group, or the psychological elision of the personal and social self (Segal et al., 2018), which increases actions taken to protect the community (Paredes et al., 2020). Crises may therefore also reinforce the sense of community between social groups. Indeed, the perception of a shared and global traumatizing experience, combined with the necessity of common and coordinated responses, increases the perception of being "all in the same boat" regardless of previous divisions between social groups (Drury et al., 2016;Muldoon, 2020). In principle then, intergroup relations should improve once groups are no longer perceived as opposed but rather as united in confronting a shared challenge (see research on common ingroup identity, e.g., Gaertner & Dovidio, 2009).
We question, however, whether it is either feasible or psychologically functional for people or society to sustain a broadly inclusive level of selfcategorization that overlooks meaningful intergroup differences. Self-categorization theory assumes that social identity is defined through a metacontrast principle, which always requires the presence of a noningroup category. Optimal distinctiveness theory (Abrams, 2009;Brewer, 2003) underlines that people are averse to being defined by overly inclusive superordinate categories, and the common ingroup identity model (Gaertner & Dovidio, 2009) explicitly highlights that a dual (shared plus distinctive) identity is often needed as a basis for improving intergroup relations. Uncertainty-identity theory argues and shows that people are motivated to reduce self-and identity uncertainty, and that group identification satisfies this motivation (Hogg, 2007). Because it is the ingroup that most people refer to for relevant norms and beliefs, the group and level at which people identify has implications for actions such as maintaining physical distance, wearing face masks, being willing to take tests, report test results, self-quarantine, accept offers of or seek opportunities for vaccination, and so on. Let us consider a few of the relevant processes in more detail.

Expectations and Perception of Leadership
In difficult and uncertain times, people turn to their leaders for guidance. Different motivations underlie this tendency, notably a need to reduce uncertainty (Hasel, 2013) and a motivation to restore an indirect sense of perceived collective control when personal control is limited (Rothbaum et al., 1982). Consequently, citizens will first look for and support leadership that is stronger, more action-oriented ("crisis-responsive"), and perhaps more authoritarian than they would in normal times (Hasel, 2013;Hunt et al., 1999). Second, they will show greater trust in the leadership (e.g., political leaders) and greater acceptance of political decisions (Kay et al., 2008).
If people are more willing to seek and follow leadership, their openness to being influenced and guided seems likely to increase too. Therefore, it matters very much which leaders or types of leadership will prevail. A social identity theory approach holds that people are likely to gravitate towards leadership that provides a meaningful translation between social identity and the normative context for behavior (Haslam, 2020). Consequently, in a period of uncertainty, people will want leadership that (a) is believed to represent the prototypical values of the group, and (b) provides unambiguous indication of the ways group members should behave (new or reinforced norms), thereby reducing uncertainty. The leader is also likely to benefit from the situation because trust in leadership is generally greater amongst citizens who identify more strongly with the group and its leader (Tyler & Degoey, 1995) and share their political affiliation (Hooghe & Oser, 2017; see also Antonakis, 2021, for further discussion of leadership).
As a shared crisis commences, superordinate self-categorization can make national or global leaders highly salient. Such leaders' ability to frame the crisis as universal elevates how prototypical they are perceived to be of the superordinate group (e.g., nation), so that people perceive them as embodying that group's prescriptive norms (Hogg, 2001). The leader's influence is further enhanced through their legitimate authority to determine new norms (a conferral effect; Abrams et al., 2008). These effects are likely to be manifested as temporary increases in trust in leadership. For example, in the early part of 2020, opinion polls in Europe revealed pronounced increases in political trust across the political spectrum. This occured in Italy, a country marked by corruption and usually low political trust (Falcone et al., 2020), as well as in Switzerland (Radio Télévision Suisse [RTS], 2020) and in France, where political trust had dramatically decreased over the preceding year with the Gilets Jaunes movement (Ipsos, 2020).

Personal Adherence to Group Norms and Perceptions of Deviant Members
Government leadership in a crisis demands compliance with new norms and rules. People show more support and respect for these rules when they have greater trust in their government (Lalot et al., 2020;Levi & Stoker, 2000;Marien & Hooghe, 2011) and perceive the government's authority as legitimate (Tyler, 1990). Shared (e.g., national) identity also facilitates the adoption of protective behavior for the sake of the community if not the self (Kramer & Brewer, 1984). In accordance with these ideas, current opinion polls internationally indicate very high motivation to comply with the most common protective measures, such as avoiding crowded public places or wearing a face mask in public places (YouGov, 2020). However, there are some significant exceptions, for example in the US, where wearing or not wearing a mask became a highly charged symbol of liberal versus conservative political identity in the latter part of 2020 (van der Linden et al., 2020).
Subjective group dynamics theory (Marques et al., 1998) highlights that people are especially sensitive to deviant behavior within their ingroup, particularly if that behavior veers towards the norms of an outgroup (Abrams et al., 2000). Consequently, the theory predicts that people would want to exert strong controls over fellow ingroup members who do not respect the current rules. This "policing" of deviant members sustains the subjective validity of ingroup norms and identity (Marques et al., 2001), often by demonstrating the group's embodiment of higher order values such as honesty or open-mindedness (Pinto et al., 2015). The desire to reduce uncertainty may also motivate people to go beyond their leader's directives. Such extremization happened in the United Kingdom during March 2020, when a petition urging the government to implement a lockdown collected more than 400,000 signatures in just a few hours. 1 However, implied in these processes is the presence of a contrasting group (or set of people) against which the group defines its norms. Anecdotally, this appears to have happened during the pandemic with multiple occurrences of "blaming, naming, and shaming" those who failed to keep social distance (Tait, 2020), and scholars warning that group polarization was likely to increase between "distancers" and "non-distancers" as the pandemic progressed (Prosser et al., 2020).
In summary, in the early phases of a crisis such as the COVID-19 pandemic, facing highly uncertain and frightening circumstances, people are likely to embrace a superordinate (e.g., national) level of identity and turn to their superordinate leader for support and guidance. The self, leader, and group are perceived as more homogeneous, promoting adherence to the superordinate group's norms and legitimizing strong responses to group members who deviate. However, as noted earlier, this situation can change as the crisis persists and grows in scale or severity (see also Packer et al., 2021, for further discussion of conformity and deviance during the time of COVID-19).

Social Identity and Intergroup Relations When the Crisis Persists
As the depth and complexities of the crisis unfold, identification with and perception of a superordinate homogeneous ingroup will tend to fade or at least to fluctuate. People's expectation that the initial response would have restored normality begins to appear unrealistic as life is increasingly pervaded by uncertainty. In the current pandemic, the somewhat romanticized characterization of common fate became problematic as huge numbers of people were forced into social and physical isolation through various forms of lockdown. The practicalities of personal survival and of dependency on and by specific others also militated against fusion to a superordinate identity.
Therefore, as the pandemic progressed, and comparisons between subordinate groups or cross-cutting categories have become increasingly salient (e.g., growing awareness of national, regional, ethnic, and age differences in infection rates), people have started questioning the superordinate identity, leadership, rules, and restrictions. They also have demanded justifications for why different groups have been subjected to or have complied with different rules. A darker side of crisis-induced uncertainty has also been exposed with the emergence of extremist groups and identities, populist ideologies, and in which autocratic leaders hold greater attraction to some people (Hogg, 2014(Hogg, , 2020. Presently, where inclusive framing of identity weakens, some groups (often those that are more disadvantaged) are becoming targets of blame or stigmatization associated with their perceived risky behavior or their vulnerability to the virus, or both. Conditions in which uncertainty is high and intergroup comparisons are salient are highly conducive to a dangerous emergence of "rivalrous cohesion" (Abrams, 2010). In periods of rivalrous cohesion, group identities furnish a strong and meaningful sense of identity through the group's competition or conflict with other groups. The ensuing highly cohesive subgroup identity also sustains subjective well-being (see Abrams et al., 2019).
As part of this process, citizens are likely to scrutinize the actions taken by different leaders to assess whether those actions are justified and appropriate. Research on deviance credit (Abrams et al., 2018) would predict that when new norms are required, leaders of superordinate groups may feel particularly emboldened by people's willingness to give them free reign. Specifically, citizens may be more willing to tolerate transgressive ingroup leadership (Abrams et al., 2013), Donald Trump's repeated refusal to wear a face mask during the U.S. 2020 election campaign perhaps being one case in point. However, when subordinate groups become more salient again, superordinate leaders will lose the advantage of universally perceived prototypicality and conferral of their right to lead. A strong backlash against superordinate leadership is also likely if those leaders overuse their deviance credits but fail to sustain the group's goals (a status-liability effect; Wiggins et al., 1965). Such discontent seems to be emerging around the globe, with citizens engaging in lawsuits against their governments for poor management of the crisis (for example in France; BBC News, 2020b).
These fractures create conditions under which groups may polarize and become more extreme. Indeed, once superordinate group leaders become perceived by some as representing an outgroup, there is likely to be a schism in trust and compliance. People's willingness to accept either innovation or transgression alters dramatically when they focus on differences between ingroups and outgroups (Abrams et al., in press). In the context of the pandemic, an initial spirit of national unity is likely to be fractured when different sections of the population (e.g., regions, ethnic groups) become more aware that they are disadvantaged or are ill-served by the superordinate group's agenda. We may then see a move towards polarized norms of, and trust in, leaders of different groups representing different objectives and priorities.
There are also important implications for prejudice and the treatment of minority groups, because other-blaming is one route through which groups may defend and maintain positive ingroup social identity while reducing uncertainty about the causes of the situation. Prejudice towards certain minority groups tends to increase when economic conditions decline and uncertainty increases (Abrams & Vasiljevic, 2014). In this pandemic, the transition from common ingroup to intergroup definitions of identity is reflected in transitions in public discourse. Emphasis shifts from depicting particular groups sympathetically because of their high personal vulnerability to infection (most notably, ethnic minorities vs. White people in European countries) to scrutinizing those groups' behaviors in the quest to assign blame for not respecting the social distancing guidelines (BBC News, 2020a). Tensions have also emerged involving newly salient social groups and categories such as people working from home versus those returning to the workplace, people wearing face masks versus those who do not, or people under local lockdown versus those who are not.
As the pandemic proceeds, the battle is not just to defeat a virus but to assert and defend different groups' rights and relative authority through laws, norms, and allocation of resources. There is also a battle surrounding who is best placed to interpret the uncertainties we all face. Individuals find that their jobs, roles, and relationships are all thrown into a different gear, often involving unwelcome curtailment, change, or obligations. This means that identity itself is destabilized, creating space for new social identities, new norms, new intergroup comparisons, and new bases for shared and differentiated group memberships.
In the absence of counteractive forces, a dystopian implication can be read from the layering of massive uncertainty surrounding COVID-19 on top of preexisting uncertainties surrounding climate change, globalization, employment, migration, and threats to democracy. Society may become fragmented into populist identity silos and/or autocratic leadership that gravitate to and polarize existing socioeconomic, political, ethnic, racial, and religious divisions (Hogg, 2014(Hogg, , 2020.

Future Directions
Understanding how social identity can be mobilized for the common good is a huge challenge for national and global leadership. Constructive innovators may have to compete with extremists for the hearts and minds of communities (see Dupuis et al., 2016). The struggle is likely to become even more relevant in shaping our futures. Yet there is relatively little research on exactly how this can be done. Optimistically, it is possible, in principle, to work towards a more positive future by reducing COVID-related uncertainty through consistent evidence-based information and clear regulations. These must be consistently delivered by trusted local, national, and global leadership that is not viewed as partisan or self-interested.
We can also go with the grain of people's desire for clear and prototypical leadership, and their willingness to exert implicit or explicit social controls over one another's behavior, to achieve positive rather than negative outcomes. Inspiring positive and inclusive leadership can succeed but, again, more research is needed to fully understand how far or fast it can transform norms and values, or transcend narrow ingroup interests, without squandering trust. More research is also needed to understand how groups can protect themselves against unscrupulous leadership that seeks to trade short-term advantages and to capitalize on group members' tolerance of unethical or self-serving practices to further an agenda of rivalrous cohesion (Morais et al., 2020). This difficult challenge involves our leaders working towards collaborative and complementary social identities, and building conditions conducive to harmonious, rather than rivalrous, cohesion (Abrams, 2010). Finally, even with the advent of technical solutions such as vaccines and treatments, the COVID-19 pandemic has revealed that societies must adapt to survive. Given what we know about groups' responses to deviance, resistance to value-challenging perspectives and evidence, and motivation for positive identity, research must continue to address the question of how to foster and spread constructive innovation. As academics pursue the many interesting and significant opportunities to develop new theory and research, we believe a key focus should be the potential fluidity and intensity of intergroup dynamics. To address the COVID-19 pandemic and beyond, understanding the role of intergroup relations should be an integral part of policy and practice.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Part of this work was supported by a grant to the University of Kent from the Nuffield Foundation (grant number WEL/FR-000022582). The views expressed are those of the authors and not necessarily the Foundation.