Intergroup helping during the COVID-19 crisis: A moderated mediation with effects of ingroup identification, ingroup blame, and perceived global common fate

Strong identification with the national ingroup encourages acts of ingroup solidarity and helping of ingroup members, but it is less well understood how ingroup identification affects willingness to help outgroup members in need. This was tested in the context of the COVID-19 crisis, asking British nationals about their willingness to donate money to help those suffering from the coronavirus crisis in China. It was hypothesized that ingroup identification would have an indirect negative effect on willingness to help outgroup members, mediated by ingroup blame, i.e. preparedness to attribute blame for the worldwide problems caused by the coronavirus crisis to the British national ingroup. Identification was expected to be negatively associated with ingroup blame, and ingroup blame was expected to be positively associated with outgroup helping. Moreover, it was hypothesized that the link between ingroup blame and outgroup helping would be attenuated if perceived global common fate in terms of managing the pandemic was high. Support for these predictions was found in a survey of British participants (N = 210).


a b s t r a c t
Strong identification with the national ingroup encourages acts of ingroup solidarity and helping of ingroup members, but it is less well understood how ingroup identification affects willingness to help outgroup members in need. This was tested in the context of the COVID-19 crisis, asking British nationals about their willingness to donate money to help those suffering from the coronavirus crisis in China. It was hypothesized that ingroup identification would have an indirect negative effect on willingness to help outgroup members, mediated by ingroup blame, i.e. preparedness to attribute blame for the worldwide problems caused by the coronavirus crisis to the British national ingroup. Identification was expected to be negatively associated with ingroup blame, and ingroup blame was expected to be positively associated with outgroup helping. Moreover, it was hypothesized that the link between ingroup blame and outgroup helping would be attenuated if perceived global common fate in terms of managing the pandemic was high. Support for these predictions was found in a survey of British participants (N = 210).
The COVID-19 crisis has been a vivid reminder that humans across the globe are more connected than they are typically aware of. Our fates are intertwined in multiple ways, ranging from the effects of pandemics to the effects of global warming. Many health experts and politicians have emphasised that teamwork is needed to solve global problems such as the COVID-19 pandemic, and they have spoken out against vaccine nationalism where the world's poorest risk being 'trampled' in a vaccine stampede, as put by WHO (e.g., BBC, 2020 ;the Economist, 2020 ). Hence, exploring predictors of help offered not only to ingroup members but to national outgroup members in this context is not only of theoretical but also of practical relevance. This paper will investigate factors that encourage monetary donations across national borders to people negatively impacted by the coronavirus crisis. Specifically, the effects of national ingroup identification, ingroup blame and perceived global common fate across national borders will be examined.
Ingroup identification, and ingroup and outgroup helping . Many studies have demonstrated that, by and large, people prefer to help ingroup members over outgroup members ( Dovidio et al., 2010 ;Levine et al., 2005 ;Levine & Manning, 2013 ). Social and group identities affect prosocial behaviour ( Stuermer & Snyder, 2010 ;van Leeuwen & Zagefka, 2017 ) and they can be assumed to do so most strongly for those who are highly identified with their social groups. After all, social group membership can provide a positive social identity and self-esteem, and ✩ Author note: the link to data/materials is provided in the manuscript.

E-mail address:
Hanna.Zagefka@rhul.ac.uk it therefore strongly affects thoughts, emotions, and behaviour ( Tajfel & Turner, 1986 ). The importance of social group membership has also been demonstrated in the context of the COVID-19 pandemic, through evidence that those who are strongly identified with their country are more likely to engage in behaviours designed to help fellow national ingroup members, such as observing social distancing or making financial sacrifices to help others who are struggling ( Van Bavel et al., 2020. Hence, group identification can be safely assumed to be positively related to proclivity to help ingroup members suffering due to the COVID-19 crisis. In sum, there is a well-documented link between ingroup affiliation and the subjective importance attached to group memberships and prosocial behaviour towards other members of the social ingroup category. Yet, it is also the case that people do sometimes, albeit more reluctantly, help members of social outgroups ( Sierksma et al., 2015 ;Stürmer et al., 2006 ). For example, Batson et al. (1997) found that empathy can induce helping of members of stigmatised outgroups; clearly showing that outgroup helping can be encouraged. The ingroup helping bias is contingent on other factors such as the presence of perceived threat ( Flippen et al., 1996 ), and whether the help is empowering in nature and whether group status relations are stable ( Cunningham & Platow, 2007 ). The bias might also be more strongly evident in some groups compared to others (Piliavin et al., 1969 What is not well understood to date is the effect of ingroup identification on help offered to outgroup (rather than ingroup) members. The present study tested a potential indirect effect, via ingroup blame attributions for the problem which gives rise to the need for help. The hypothesis was that strongly identified group members would be more reluctant to accept or aportion blame for an existing problem to their ingroup, and that this reduction in ingroup blame would, in turn, reduce willingness to help members of outgroups suffering due to the problem. It is perhaps surprising that the effect of ingroup identification on willingness to offer help to outgroup members has not previously received more attention, because it has certainly been tested as a predictor of intragroup help ( Van Bavel et al., 2020 ), and also of willingness to receive (rather than offer) help across group boundaries: for example, Nadler and Halabi (2006) found that strong identification attenuates willingness to accept help from higher status outgroups. What emerges is a picture where it is well understood how identification affects some aspects of intergroup helping, but not others, and the present contribution aimed to redress this.
Identification as a predictor of ingroup blame . There are a myriad of factors to which one can hypothetically attribute the coronavirus crisis, all of which have featured in the British media over the last year or so. Some commentators have blamed the Chinese government for not acting swiftly enough in response to the emergent health crisis or even trying to cover it up, while others emphasise the ineptitude of the response by the British government for handling the crisis, the behaviour of certain segments of the population (e.g., young people) for failing to observe social distancing rules, etc. There is also a growing understanding that failures in one country to manage the crisis effectively can negatively impact other countries, as COVID-19 cases get reimported across nation borders to countries which had temporarily sucessfully managed to reduce infection rates. The attribution most relevant in the present context is that to the national ingroup: the extent to which British nationals believe that the British response to the coronavirus crisis is causing problems for people in other countries, and in particular China. On the basis of the social psychological literature, the link between national identification and ingroup blame in this context can be expected to be negative, as will be outlined below.
Because group members draw a sense of positive self-esteem from their group memberships, they are motivated to see the ingroup as positively distinct from relevant outgroups ( Tajfel & Turner, 1986 ). The desire to see the ingroup in a positive light, which should cause reluctance to acknowledge any culpability of the ingroup for the creation of problems, will be especially strong for those who care most deeply about the group. In other words, in the face of adversity or challenges ingroup identification will be positively related to a desire to blame outgroups for the issues at hand, and negatively related to a willingness to acknowledge ingroup culpability. This is indeed what Zagefka (2021a) found for British nationals and their attributions for problems caused by the coronavirus crisis. Strongly identified Brits were more likely to emphasise things the Chinese had done to exaccerbate the crisis, and less likely to think that the British themselves were to blame for the problems because they had mishandled the crisis (see also Zagefka & Sun, 2021 ).
Ingroup blame as a predictor of intergroup helping . If identification reduces ingroup blame, blame attributions can in turn be expected to affect decisions to help outgroup members. Attributions of problems to a range of different possible causes have clear consequences for behavioural responses ( Weiner, 1980 ;1993 ). This has been demonstrated, for example, in the domain of victim blame: victims who are blamed for their plight are helped less ( Schmidt & Weiner, 1988 ). Another example, from the context of donations to disaster victims, is that the perceived root cause of a disaster event (i.e., whether it is seen to be naturally or humanly caused) impacts on decisions on whether or not to donate money to help the victims ( Zagefka et al., 2011 ).
The type of attribution relevant here is whether or not people believe that their ingroup is responsible for exacerbating a certain problem, in this case the coronavirus crisis. Believing that the ingroup is to blame for a wrong causes feelings of guilt ( Wohl et al., 2006 ), as well as an increase in perceived moral obligation and perceived responsibility to intervene and fix the problem. For example, James and Zagefka (2017) found that when an accident that caused casualties was perceived to be caused by a national ingroup rather than a national outgroup member, perceived responsibility for the event was increased, and this in turn led to a greater willingness to donate money to help the surviving victims of the event. Others too have found that perceived ingroup culpability and associated emotions of guilt increase support for reparative actions to benefit those harmed ( Kardos et al., 2019 ;Rothschild et al., 2013 ). The flip side of this, of course, is that an absence of ingroup culpability is likely to decrease perceived responsibility to make good any damages, and a lack of perceived culpability of the ingroup is therefore likely to reduce willingness to assist outgroup members that have been harmed. Taken together, then, it is proposed that identification with the national ingroup will decrease blame attributions to the ingroup for existing problems (in this case, the coronavirus crisis), and that this in turn will reduce willingness to help national outgroup members negatively impacted by the crisis. In other words, there is a hypothesized indirect negative effect of ingroup identification on outgroup helping, mediated by ingroup blame.
Moderation by perceived global common fate . One question driving the present research was whether the effect of ingroup blame on willingness to help outgroup members might be contingent on additional factors. When theorizing about potential moderators of this effect, it seems that perceived global common fate might play an important role. Common fate is one of several factors that have been shown to make it more likely that individuals get perceptually categorised as belonging to the same social entity ( Campbell, 1958 ). Common fate and other factors related to perceived group entitativity has also been shown to be related to identification with that social category ( Castano et al., 2003 ). Common fate is assumed to be psychologically relevant in the context of Covid-19. Reinfection rates can rise due to new, internationally imported cases in countries that had temporarily done well in managing the spread, and health experts emphasise our true global interdependence in the fight against Covid-19 (e.g., Farrar, 2020 ). However, there are differences in the extent to which people appreciate and understand this goal interdependence between nations and regions in the world in terms of overcoming the pandemic (Zagefka, 2021). The real-life importance of global common fate in managing this global problem makes it especially important to study the effects psychological beliefs about it, with a view to potentially developing interventions.
Within the psychological literature, a perception of shared common fate has been investigated in the context of identification not only with small social categories but also with the most global, all-comprising categor of 'all of humanity ' ( McFarland et al., 2012 ). A perception of shared common fate has also been shown to have important behavioural consequences in the context of a flood disaster: Ntontis and colleagues (2020) showed that perceived common fate was a very powerful predictor of the provision of social support. The strong motivating effect of perceived common group identity and common fate for solidarity and helping has also been demonstrated in other contexts, ranging from lab studies ( Dovidio & Morris, 1975 ) to rescuing Jewish people under threat in Nazi-occupied Europe ( Reykowski, 2002 ).
If perceived common fate is such a potent predictor of helping, then it stands to reason that the effects of other considerations, like the extent to which the ingroup is to blame for the problem, pale in situations in which perceived common fate is high. In a sense, perceived common fate might 'trump' and dominate over other considerations, to the extent that it wipes out or at least diminishes the effects of other psychological predictors. In other words, it was predicted that the effect of ingroup blame on outgroup helping would be attenuated by perceived common fate, in such a manner that it would be less strong if perceived common fate is particularly high.
There are two reasons for why such a pattern might be expected. First, if global common fate is salient then solving an outgroup's problem becomes self-serving: by helping them, we are helping me/us. Given the suggestion in the literature that prosocial behaviour can be motivated by self-interest (e.g. Cialidini et al., 1997), and the reasonable expectation that people are concerned with self-preservation, providing a frame which turns outgroup helping into self-serving acts should lead to high endorsement of helping, no matter what other psychological considerations (such as blame) are salient. Second, as mentioned above, a perception of global common fate can turn former outgroup members into ingroup members, or at least people who share category membership at a superordinate level; and such shared category memberships potently affect behaviour (see e.g. González & Brown, 2003 ;Sherif, 1966 ). Given the demonstrated preference for helping ingroup members, turning outgroup members into ingroup members by making shared common fate salient should lead to high endorsement of helping, and 'trump' the effects of other psychological considerations such as blame.
Overview of the research . In sum, while many factors affecting intergroup helping are well understood (e.g., Nadler & Halabi, 2006 ;van Leeuwen & Zagefka, 2017), the present study aims to make a novel contribution by 1) answering how ingroup identification affects offers of intergroup help (not, as has been explored previously, offers of intragroup help or acceptance of intergroup help), by 2) exploring how intergroup helping is affected by blame attributions, and by 3) exploring how perceived global common fate might attenuate some of these relationships.
Taken together, then, it was hypothesized that identification would reduce blame for the problems caused by the coronavirus crisis attributed to the ingroup. Moreover, it was hypothesized that ingroup blame for outgroup problems would be positively associated with willingness to help the outgroup. Last but not least, the effect of ingroup blame on outgroup helping was expected to be moderated by perceived global common fate, so that it would be particularly pronounced if perceived global common fate is low. The hypothesized processes are summarised in Figure 1 . Hypotheses were tested with a sample of British nationals, asking participants about the degree of identification with their national ingroup, blame attributions for problems caused by the coronavirus crisis for people in China, perceived global common fate in relation to defeating the crisis, and their willingness to help people negatively impacted by the coronavirus crisis in China.

Participants
Two hundred ten participants completed the study on the Prolific oneline platform in June 2020 (mean age = 33.43 years; 149 females, 60 males, 1 differently identified). Participants were offered monetary compensation for answering a short survey; the rate of pay was £7/per hour. Only participants who had identified themselves as having British nationality participated. All participants completed all measures, there were no missing data. Sample size was determined with G * Power, computing the required sample size given an alpha of .05 and a power of .80. Estimates were based on moderate correlations of blame and outgroup helping found in previous work in the context of the coronavirus crisis ( Zagefka, 2021a ).

Measures
All measures were assessed with Likert scales (1 = "strongly disagree " to 5 = "strongly agree "). British ingroup identification was measured by averaging three items based on Brown et al. (1986) : 'It is important to me to be British'; 'I see myself as British'; and 'I am glad to be British', = .90 (higher scores indicate higher identification).
Ingroup blame was measured in relation to China, by asking 'In your view, whose fault is it that the pandemic is causing problems for people in China?' Participants indicated their agreement with the statement 'The British government is to blame' (higher scores indicate more blame).
Perceived global common fate was measured by averaging across two items: 'As long as coronavirus is somewhere in the world, we are all in danger', and 'We must address coronavirus together as a global community', = .56 (higher scores indicate more perceived common fate).
To measure outgroup helping , participants were asked 'Would you be willing to donate money to help people suffering from the coronavirus crisis?', and they rated their agreement with the statement 'I would donate to help Chinese people in trouble' (higher scores indicate more helping). 1 This study received ethical approval from the author's institution, and all aspects of the research were in line with the ethical guidelines outlined by the British Psychological Society. The data is available here: https://osf.io/vtpgn/?view_only = d85a7de9a84247ebaeb16dad0a480c54 .

Results
Descriptives and bivariate correlations are displayed in Table 1 . The pattern of bivariate correlations was in line with predictions, in that identification manifested a negative association with ingroup blame for outgroup problems, and blame manifested a positive bivariate correlation with outgroup helping. The proposed mediator of global common fate was not significantly associated with either identification or ingroup blame, and the direct association between identification and helping was not significant.
Moreover, in line with predictions adding the interaction between perceived global common fate and blame to the model significantly increased the variance explained, ΔR 2 = .02, F (1, 205) = 4.16, p < .05. To interpret the interaction, it was plotted at three different levels and is shown in Fig. 2 . As can be seen there, in line with the predictions the effect of ingroup blame on intergroup helping is particularly strong when perceived global common fate is low, and it is significantly attenuated when global common fate is perceived to be high. 2

Discussion
Overall, results clearly supported the predictions: there was a negative indirect effect of ingroup identification on outgroup helping, mediated by ingroup blame, and this latter effect was further moderated by perceived global common fate, in the sense that it was attenuated when perceived global common fate was high.
Hence, while ingroup identification has been shown to have beneficial effects on ingroup solidarity and willingness to assist ingroup members, as recently also demonstrated in the context of the coronavirus crisis Van Bavel et al., 2020, at the same time ingroup identification might reduce willingness to help people who are seen to be outside the boundaries of the social ingroup. Thus, group identification can be a double edged sword: it encourages desirable behaviours towards the ingroup, but at the potential cost of discouraging desirable behaviours towards the outgroup. As such, the effect demonstrated here speaks to the classic question raised by Allport: does fondness of the ingroup necessarily imply negativity towards the outgroup ( Dovidio et al., 2005, Brown & Zagefka, 2005? Or, put in more absolutist terms, are ingroup love and outgroup hate the same thing ( Brewer, 1999 )? The present results suggest that indeed, at least in the context of helping, solidarity with ingroup members motivated by ingroup affiliation does come at the cost of lesser solidarity with those outside the social ingroup boundaries.
The results also point to the importance of attributions of social problems for informing behavioural responses to those events. Perplexingly, only some types of attributions have been researched extensively, whilst others have received less attention or have been ignored almost completely. Well-researched are effects of blaming victims for their plight (e.g., Lerner, 1980 ). Considerably fewer studies have tapped into potential effects of blaming the social ingroup ( James & Zagefka, 2017 ). However, a host of other attributions are possible for many issues that have been almost entirely ignored. For example, many events might not be due to what the victims did, or what the ingroup did, but they might be due to what a third group did. Moving away from simplistic ingroup-outgroup dichotomies ( Dixon et al., 2020 ) might go a long way in drawing more attention to more complex attributional patterns and their potential consequences.
The results also confirm the importance of a perception that the fate of the ingroup is shared with others around the world. As seen in other contexts (e.g., Ntontis et al., 2020 ;Zagefka, 2021b ), a sense of shared adversity or shared destiny can have powerful consequences for behavioural choices, especially in the face of disaster events. In the context of the current coronavirus crisis, indeed social scientists have highlighted the importance of shared identity and social cohesion for generating compliance with measures that might be inconvenient for individals but that are essential for the management of the pandemic for society as a whole (e.g., Miller, 2020 ). The present findings, however, suggest that maybe the best level of categorisation for generating a sense of commonality, solidarity and shared fate is the global level. Caring about the national ingroup might only go so far: in order to truly beat the virus, global cooperation and an understanding of global goal interdependence is needed.
The present research is, of course, not without limitations. Given the desire to rapidly respond to the current pandemic, the research was conducted with easily accessible online samples. Although there is good evidence that data acquired via this route is generally of high quality ( Peer et al., 2017 ), of course caution is advised when trying to generalize from the present sample to the general population. Future research could improve on this by seeking larger, truly representative samples. Also, because hypotheses were not pre-registered the present findings must be interpreted as exploratory only, and future research should aim to confirm the pattern following pre-registration. Moreover, although the hypotheses clearly propose a causal direction of the effects, causality cannot be infered from the present cross-correlational data. Future reseach could utilise experimental paradigms to get a better handle on causality of existing effects.
More excitingly though, future research could not only improve on these methodological limitations, but also follow up on some substantial questions thrown up by the present findings. Most of all, it would be exciting to test whether the processes at play here in the context of the coronavirus pandemic hold true also across other issues of global reach. Two in particular spring to mind. First, climate change is clearly an issue where objectively there is interdependence between people in different nations, because global climate change affects all, even though it might do so unequally because it has stronger negative effects in some parts of the world. One could study, for example, ingroup blame by citizens in rich, gass-guzzling nations, as well as beliefs about global common fate in terms of global warming, and the effects of those variables on willingness to help, for example, victims of climate-change induced disasters such as flooding, drought and wildfires.
Second, one could study psychological processes that drive (or discourage) intergroup helping by members of rich nations for refugees from poorer parts of the world, for example those trying to cross the Mediterranean from Africa in a quest to reach Europe. People might differ in the extent to which they believe their nations are responsible for poverty in the global South (e.g., due to a history of colonialization, unfair trade agreements, etc.), and in the extent to which they perceive interdependence between nations in achieving individual prosperity as well as prosperity for the world economy at large. The effects of such variables on desire to help migrants and refugees could be tested, in an attempt to test whether the psychological processes concerning identifi-cation, blame, and perceived common fate found here in relation to the coronavirus crisis will generalize to other global challenges.
Third, two rationales for why global common fate will moderate effects were outlined in the introduction, one related to self-interests, and the other related to common fate affecting social categorisation, turning outgroup members into ingroup members. However, neither of these potentially explanations were directly assessed in the present research. One reason the focus was on perceived global common fate was that a goal of the study was to find factors that could be useful in designing interventions, and it seem possible to devise plausible easier to devise messages that persuade of common fate ('We are in this together!'). However, clearly more work could be done to explore exactly which psychological processes are tapped into by inducing a sense of shared common fate, which then lead to behavioural outcomes relevant for helping decisions.
As highlighted from the outset, the aim of this research was to generate knowledge about basic psychological processes driving intergroup helping decisions, but also to hopefully generate insights that can be of practical relevance. Any conclusions should be considered tentative though, given the exploratory nature of the work and the relatively small effects. When reflecting on the applied implications of the data patterns found here, two variables drive greater intergroup prosociality: the interaction showed that willingness to help across national group boundaries was elevated when global common fate was perceived to be present, When perceived global common fate was low, helping was still favoured when blame attributions to the ingroup were made. Taken together, one might tentitatively suggest that perceived global common fate is a promising leaver for encouraging behaviours to combant the pandemic. A perception of joint adversity shared with all people across the globe, including those in other countries and even continents, might encourage solidarity that cuts across narrowly defined social divides. Public support for global measures and a global solution to the problem is of pivotal importance when trying to combat 'vaccine nationalism', whereby rich states monopolise resources urgently needed in poorer regions of the world (e.g., Farrar, 2020 ), and perceived global common fate seems an important driver of a willingness to share resources.

Footnotes
1 Some other items were included in the study which were related to a totally unrelated question/study, nothing to do with Covid or China. Data collection for the two studies was combined for reason of economy. There was, however, just one further item related to the present topic, and that was about blaming China for problems in China. When predicting willingness to help from both ingroup blame and outgroup blame simultaneously, the ingroup blame effect remains virtually unaffected, and outgroup blame did not have any significant effect, suggesting that this construct does not significantly affect the processes described in this paper.
2 It was also explored, using step-wise regressions, whether perceived common fate would moderate the effect of identification on blame attributions, and whether it would moderate the effect of identification on willingness to help. Neither of the interactions was significant ( = -.26, ns when predicting blame, and .28, ns when predicting helping).