10.02: Contagious cooperation: A closer look at collective self-organization

Aim of the project

The main aim is to examine evidence for spontaneous prosocial/cooperative behaviour under conditions of acute and chronic adversity (threat: external shocks). Evidence already suggest that there is prosocial/cooperative default under conditions of acute threat (e.g., disasters, evacuations). We also assess whether this already prosocial default can be sustained over time (sustainable cooperation) notably when the external threat is more chronic in nature.

Theoretical background

A central SCOOP premise is that cooperative responses to external shocks are one thing but to keep cooperation going is another matter as cooperation breaks down or decays over time (e.g., the “tragedy of the commons”). Reasons include self-interest, a lack of surveillance inter alia. But there are also well-documented exceptions to this general rule. Under acute collective threat (e.g., disaster, external attacks), even though individual self-interest and personal survival are on the line, people can act in surprisingly prosocial and selfless ways. A large body of empirical research shows that in emergency situations such as mass bombing or natural disasters, panic and selfishness are rare, whilst spontaneous cooperative behaviour is common (Auf der Heide, 2004; Drury, 2018; Fritz & Williams, 1957; Janis, 1951; Mawson, 2005). Consistent with this, evolutionary scholars see humans as an ultrasocial species, for reasons of group/multilevel selection (e.g., Dunbar, Barrett & Lycett, 2009) and to cultural evolution (Henrich & Muthukrishna, 2021; Richerson et al., 2016). Sustained cooperation may also stem from social institutions invoking group identities and subcultures, fostering cooperative social norms and evolved collaborative practices. At the same time, one should not naively assume that cooperation reigns: it can be eroded among others by alternative culturally grounded understandings of individual self-interest or free market capitalism. In other words: cooperation might arise spontaneously and be subsequently promoted (or threatened) by theories of human nature.

In this proposal we focus on the processes underpinning “flash cooperation”. Elinor Ostrom’s research has shown much compelling evidence for self-organization underlying cooperation that was also prefigured by Hayek’s concept of “spontaneous order”, the insight that complex systems (like the free market) have ways of organising themselves efficiently, bottom up, rather than depending (solely) on top-down rules or principles. However, these mechanisms typically require some initial communication and coordination not least to establish that reciprocity, for example, is working and mutual. A good example is provided by Axelrod (1984) who cites the example of troops in the WWI trenches who avoided destructive lethal action that could escalate to mutual disadvantage. These analyses remain compatible however with an individual self-definition and analysis focusing on interdependence and reciprocity between individuals (see Iacoviello & Spears, 2022; Spears, 2021), and remain more agnostic about whether there is a “prosocial default”, a norm associated with the collective level. However more recently, research on team reasoning (Bacharach, 2006; Hindriks, 2012) has highlighted the collective level of being and behaviour consistent with the social identity tradition, that seems to transcend the individual focus on self-interest characterising homo economicus.

Nevertheless it is much less clear that this cooperation can occur spontaneously, even automatically without conscious communication or coordination. However, this is precisely what recent evidence suggests might be possible, namely, that in collective situations in which capacity for deliberation and communication is absent and which require an immediate response, a high degree of collective order can emerge seemingly spontaneously (Sieben & Postmes, 2023). This research demonstrated experimentally that people who are highly motivated to gain entrance to a concert venue race each other (which might suggest competition, selfishness and a breakdown of order) to get a good place in the queue (a form of social order which suggests the opposite). Numerous other examples of rapidly emergent cooperative structures have been described in historical studies of food ‘riots’, collective religious violence and mobilization for self-defence (Davis, 1973; Lefebvre, 1932; Thompson, 1971). Similar phenomena can be observed in improvised search-and-rescue operations after disasters (Auf der Heide, 2004) or survival strategies in war (Kindsvatter, 2003). Interestingly, the recent experimental evidence suggests that such “flash cooperation” can occur in situations of acute collective threat, but also of acute collective opportunity, when people perceive that change is required to secure a sudden benefit or windfall (Sieben & Postmes, 2023). In large groups, then, processes of self-organization can occur very fast, and even seemingly without much social deliberation or overt coordination (cf. Turner & Killian, 1987). In short this suggest a process of “contagion” whereby people, seem to be spontaneously organized and influenced by others in context where there is uncertainty and insufficient time for an organized, or communicated, response.

For SCOOP, such rapidly emergent “contagious cooperation” is clearly of interest, especially because it has been described in many different contexts but the processes involved remain speculative. Indeed, the concept of contagion has a dubious history harking back to Le Bon (1896). His argument was that the crowd was a source of atavistic, antisocial and even animalistic processes: the crowd’s cooperation he ascribed to the ‘group mind’ that steered it.^[1] Such notions have been strongly disputed in sociology and psychology (McPhail, 1991; Reicher et al., 1995). In psychology, Le Bon’s ideas led to the development of deindividuation theory in the 1950s, reformulated in modern (cognitive) social psychological terms, invoking loss of self-awareness (see Postmes & Spears, 1998, for a critique and review). The critique of this idea evolved into a social identity model of deindividuation effects (SIDE) arguing that the self is not lost or submerged in the mass, but that there is a switch of identity, shifting the focus from individual to group level concerns and the social norm associated with group identities (Postmes & Spears, 1998; Reicher et al., 1995; Spears & Lea, 1994). As well as being more socially grounded, an indeed as a consequence, this recasting of contagion in more social terms suggests that it far from unconditional or “kneejerk” but is highly sensitive to context, as well as being contingent on collective identity. In this sense it requires an analysis of the social context and the specific situation, of which the external threat is one prominent example.

Until now, the SIDE model focused on cognitive and strategic processes associated with immersion in the group or crowd, explaining the effects in terms of depersonalization (the idea that immersion and anonymity could render group identity more salient) and strategic processes (i.e.  taking account of social support and the cover of anonymity to pursue group goals that might be sanctioned or thwarted by powerful authorities or outgroups). In short, according to SIDE the crowd’s cooperation is in part due to cognitive and deliberative processes, that make the possibility of spontaneous, even unconscious action, seem less viable.

In the current proposal, we therefore propose that other processes seem necessary to explain contagious cooperation, because this can occur in contexts where deliberation or communication is quite difficult if not impossible. We suggest that contagion-like cooperation appears to be more grounded in processes of collective emotion (Kuppens & Yzerbyt, 2012; Spears et al., 2011) and relatively automatic behavioural imitation (Spears et al., 2004). Also relevant, we believe, are processes described in the philosophical literature on collective intentionality (Searle, 1995; Zahavi, 2021). The current research therefore aims to examine processes contributing to more spontaneous cooperative behaviours in collective situations. If this account is valid, what factors can cause these newly formed cooperative structures, not just to emerge as collective coping devices, but also to thrive and endure? We distinguish three: context, culture and connectedness.

Contextually, it is known that unusually high levels of cooperation and even self-sacrifice can sustain for a long time, when conditions of extreme external threat continue, as in war (Janis, 1951; Kindsvatter, 2003) or in the corona pandemic, especially in the first wave (Mao et al., 2021). In psychological terms, if common fate endures, so will the cooperation that it invites (Drury, 2018). This factor is external to the community affected.

Cultural and associated normative factors may promote cooperation and cause it to endure but also to break down. Cultural norms are often intrinsic to the community or wider society. For example, queueing behaviour may be quite sensitive to specific national norms or stereotypes. In particular culturally ingrained assumptions about self-interest, (the ‘homo-economicus’ prevalent in western/capitalist society), may erode newly established cooperation and stimulate relational models such as market pricing in social relations among strangers (Fiske, 1992; see also goal framing theory). What is less clear is whether there is a more generic prosocial default and how sensitive this is to local norms in the (sub)culture. It remains to be seen whether culturally ingrained normative factors impinge on defaults (especially where economic factors are not cued or primed) or only emerge later in interaction with others, and how easily such prosocial defaults can be restored, to revitalize cooperation when contradicted. For example, in new groups and settings do people return to pro-social defaults or are they adversely affected by being burnt in the past (e.g., by free-riders, defectors). There is legion evidence from the social dilemma literature and the minimal group paradigm suggesting that people often begin with prosocial or fair choices until (ingroup) norms, or behaviour of others deviate (e.g., Iacoviello & Spears, 2021).  It is a methodological challenge to distinguish cultural and more spontaneous behavioural factors, but we think this distinction is conceptually valid and largely neglected in research. Establishing the basis of such behaviour may then provide the key to understanding how cooperation (both deliberative and spontaneous) can be sustained and when and why it breaks down.

Finally, a third reason why cooperation may endure is related to the connectedness and camaraderie generated by (successful) cooperation. This is a factor that is intrinsic and emergent. For example, in Dutch health care institutions during the first year of the corona pandemic, job satisfaction and experienced work pressure were uncharacteristically positive, despite elevated risks and objective pressure (CBS, 2022). We hypothesize that this was caused by the intrinsic experience of reward of mobilizing a collective disaster response during the first wave in particular. The processes responsible for this positive feedback loop are likely to include a sense of empowerment or efficacy (Drury & Reicher, 2009), an emergent sense of solidarity (Koudenburg et al., 2015), or the simple fact that participation in new cooperative action can be intrinsically fun (van Mourik Broekman et al., 2018). Crucially, the role of such collective emotions in sustaining cooperation has never been directly studied.

To summarize: the general aims of the proposed research is to examine evidence for spontaneous and ‘contagious’ cooperation in situations of high uncertainty, flux or threat, and to examine evidence for the emotional and behavioural basis of emergent cooperation in addition to more deliberate cognitive and communicative processes examined in previous research (e.g., using SIT and SIDE). A second key aim is to investigate the sustainability of ‘contagious cooperation’, by examining the influence of context, culture and connectedness.

Research design

We propose the PhD to conduct two research packages. The first will take a closer look at the contagious cooperation. Lab studies and a field study examine the conditions for cooperation, the process of self-organization and its psychological correlates. The second strand examines the conditions under which cooperation sustains.

Package 1: contagious cooperation after acute threat and opportunity

This compares conditions and cases of high self-organised cooperation with control conditions in which this cooperation does not emerge, focusing on processes involved. Study 1a is a real-life field study of the induced earthquakes in Groningen, using existing data. Field work in 6 communities found that only some mounted a collective response to acute external threats (widespread damage, home demolition) whereas many others did not (Stroebe et al., 2019). The data suggests this was not due to socioeconomic factors, but rather to collective awareness of the threat being acute. In communities without this awareness, cooperation was minimal. This field work data was published only in Dutch-language reports. The emotions involved were hitherto not analysed: the PhD would re-analyse existing data and publish internationally. The fieldwork is complemented with a quantitative analysis of longitudinal community resilience data N>1600 (Postmes et al., 2018). In this analysis, community resilience can be related to perceived risk of earthquakes, actual exposure, and various emotions.

In the lab studies 1b, we emulate and extend the fieldwork. The lab studies introduce control by minimising the opportunity for deliberation and so ensure that we are indeed studying ‘contagious cooperation’. The lab studies extend the fieldwork by exploring the question whether acute and destabilising threats (a crowd fleeing an undesired outcome) has the same effect as acute and destabilizing opportunity (a crowd approaching a desired outcome, cf. Sieben & Postmes, 2023). There are three conditions (acute threat, acute opportunity, plus a neutral control condition). Our prediction is that acute collective threat and opportunity lead to rapid and spontaneous self-organization and co-ordinated action, compared to control. Process variables are measured. We model these studies on successful paradigms used before (see Sieben & Postmes, 2023). The lab studies allow us to conduct a more precise analysis of the collective emotions and possible other processes involved in this rapid spread of cooperation. Counterbalancing threat and opportunity is of course especially relevant with respect to assessing these emotions: strong evidence of our hypotheses would be to find positive emotions in the threat condition. Our ambition is to go beyond current understandings of ‘contagion’.

Package 2: sustained cooperation

Inducing self-organised cooperation is one thing, getting it to stick quite another. As explained above, we propose that cooperation may sustain because of external factors, a priori– and emergent community characteristics: context, culture and connectedness. In order to examine each of these we want to use paradigms developed in prior research on small group collaboration: we established groups and norms in a phase 1 task, and studied cooperative outcomes in an unrelated phase 2 task (for example see Postmes et al., 2001; van Mourik Broekman et al., 2019). The challenge for the group is therefore to successfully self-organize a second time in order to execute a complex task. The degree of successful cooperation during the phase 2 task is the key outcome. The conditions for sustained cooperation are manipulated in Phase 1 as follows:

Study 2a studies the role of context, and manipulates whether the external threat is acute only during Phase 1 and then dissipates, or chronic continues throughout Phase 2. So concretely, the experimental groups in the acute condition would face a threat only at the beginning of Phase 1, whereas the chronic condition faces a threat that continues to the end of Phase 2. The prediction is that cooperation sustains more into Phase 2 if threat persists.

Study 2b examines the role of culture and local norms. Using the social relations model of Fiske (1992) we aim to prime different forms of social relations (e.g., “equality matching” vs “market pricing” associated with more collectivistic vs capitalist cultures) in a multi-session lab study to investigate how these forms of social relations affect both the spontaneous and enduring levels of cooperation over time. Our previous research also shows that political ideology (right vs. left) also stimulates ingroup level cooperation vs. fairness at the supraordinate level (i.e. also including outgroups) respectively (Iacoviello & Spears, 2018, 2022), so this will be an important moderator. 

Study 2c examines the role of connectedness. We will impose the same collective threat on all in Phase 1, and will manipulate whether, within the group, it is possible to connect during the response or not. This can be achieved, by varying the extent to which people have the ability to co-ordinate their behavioural response to the threat or not using a previously developed method by Van Mourik et al. (2018). They compared groups carrying out a collective task whilst visible to each other vs. being separated by screens during the task execution. This separation by screens does not affect the task execution itself but does affect the collective emotions reported afterwards (including having fun and experiencing solidarity). Our prediction is that when collective emotions are experienced more strongly, it is more likely that cooperation will sustain.

^[1] Such notions are not restricted to Le Bon’s analysis of crowds. Keynes for example noted that markets operate by “animal spirits”. Interestingly, economists Akerlof and Shiller (Akerlof & Shiller, 2010) adopt a social identity approach in their analysis of such processes.