Background

The idea that democracies are more peaceful than other forms of human societies –sometimes termed democratic peace theory - has a long history, present in the works of famous 18^th century thinkers such as Thomas Paine ^[1] and Immanual Kant ^[2]. The latter summarised the logic behind this idea in his manifesto for perpetual peace: describing how autocratic leaders can “decide on war for the most trifling reasons, as if it were a kind of pleasure party”, whereas democracies “should weigh the matter well, before undertaking such a bad business”. Simply put, in democracies, where citizens more evenly share the burdens of decision-making and the costs of war, individuals should be hesitant to fight. In contrast, in autocracies, leaders can more easily monopolise the spoils of war while decoupling themselves from the costs, and so are incentivised to wage war for personal gain even if it’s bad for everyone else. Modern analyses support Kant’s original insight, with a statistical link between democracy and peace that is at least five times as robust as that between smoking and lung cancer ^[3]. As evolutionary biologists with an interest in the origins of war, we were curious to see if the idea that shared decision-making promotes peace might also hold up when applied to non-human animals, who lack any of the complex institutions like governments that are often used to explain the existence of democratic peace.

 Recent work on animal decision-making has shown that societies can vary in how power is shared, from despotic baboons ^[4] to democratic wild dogs who “vote” via sneezing ^[5]. We were particularly inspired by long-term research on banded mongooses:  highly aggressive and warlike mammals who live in packs across Africa. Research suggests that female mongooses can take control of group movement decisions, but that they abuse this power by leading their groupmates into often fatal fights while they prioritise their love life and seek mating opportunities within the rival pack ^[6]. We wanted to ask: how empowering these male mongooses to have more of a say would alter the balance of power and influence the frequency of fighting observed?

All is fair in love and war. In banded mongooses, rival packs are led into fights by females who selfishly seek new mating opportunities with rival packs at the expense of their male groupmates. This is an example of an autocratic animal society where the combined effects of unbalanced decision-making influence and biased conflict-seeking incentives can lead to more regular intergroup conflict than is expected in more democratic societies. Credit: Dave Seager, Banded Mongoose Research Project.

Approach

To investigate how differences in group decision-making influences conflict seeking we decided to build an evolutionary model using game theory. Popularised during the Cold War, game theory is an incredibly useful mathematical tool for studying strategic interactions where your best move is conditional on that taken by your partner, whether that be a rival nation or mob of mongooses. We modelled animal groups consisting of two subgroups or classes: leaders and followers. Leaders and followers each differed in the share of costs and benefits they would receive from fighting and the influence they held over the group’s decision whether to fight (termed playing ‘Hawk’) or not (playing ‘Dove’). Leaders were modelled so that they always contributed fully to the group’s decision whereas follower contribution varied depending on their influence score (Ω). This could range from an influence score of 0, no follower input at all representing an autocratic society, to 1, follower input equal to leader input representing a democratic society. We then used both analytical and simulation-based modelling techniques to find the level of conflict seeking (or ‘Hawk’ playing) that maximised the payoffs (combined benefits and costs) received by each class. By changing different parameters, such as the influence score of followers, we could then explore how changing from an autocratic to a democratic society influences the amount of conflict seeking that a group pursues (termed P), as well as the strategies played by each class (P_L and P_F for leaders and followers respectively).

Model outputs for different values of follower influence (Ω). When follower influence is low leaders can control the group decision and achieve a group strategy (P – green line) that is equal to their own preferred level of hawk playing (Ṽ_L – purple dashed line), which itself is determined by their share of costs and benefits received from fighting. As followers become more empowered and their influence increases, they can take more control and reduce the group’s aggressive conflict seeking behaviour in line with their own preferred level of hawk playing (Ṽ_F – yellow dashed line). While trying to wrestle control of the overall group-level decision in their favour, each class is incentivised to pursue an individual strategy that can lead to extreme polarisation between the classes (difference between leader strategy P_L , purple solid line, and follower strategy P_F ,  yellow solid line).

Results

We find that when followers suffer more or gain less from fighting then increasing their influence can lower the group’s level of aggression, which is consistent with Kant’s arguments and the democratic peace hypothesis. This is in a simple model based only on differences in decision-making influence and conflict-seeking incentives, and which doesn’t involve any of the human institutions that are often invoked to explain the existence of democratic peace. Interestingly, patterns of democratic war – where increasing follower influence increases group-level aggression – is also possible if the payoff structured is reversed.

Another insight from our model is that leaders and followers are often incentivised to adopt extreme positions of either all-out-attack (Hawk playing = 1), or total pacificism (Hawk playing = 0), leading to striking polarisation between different classes. This is because when a class adopts an extreme position it gives them more leverage, or a better ‘anchoring’ point, for when they combine inputs with the rival class to create the group-level strategy. The adoption of such extreme all-or-nothing preferences is also seen elsewhere in biology, such as in conflicts over gene expression ^[7]. Polarisation is also present outside biology too, for example Harvard University researchers have found a strategic advantage to electing congress representatives with extreme ideologies because they’re more effective in cross-party policy negotiations ^[8], which mirrors the results that evolve in our model.

 Future Directions

However, like all models, ours is constrained by the assumptions we are required to make. One such assumption is that individuals can’t leave or change groups. This again was inspired by banded mongooses, where the life expectancy for a lone mongoose is extremely low as they make a tempting bitesize snack for any nearby monitor lizards. If this assumption were relaxed, and individuals could freely switch groups, perhaps the followers would vote with their feet and leave when they are dragged into unfavourable fights. In consequence, perhaps a leader would be a less willing warmonger if they knew they ran the risk of desertion and fighting alone. Hopefully future work can build on our model in ways like this to expand our understanding of the role of decision-making on intergroup conflict, both in humans and other animal societies too.

References

1. Paine, T. Rights of Man, Common Sense, and Other Political Writings. (Oxford University Press, 2008).
2. Kant, I. To perpetual peace: a philosophical sketch (ed T. Humphrey) (Hackett, 2003).
3. Imai, K. & Lo, J. Robustness of empirical evidence for the democratic peace: a nonparametric sensitivity analysis. Int. Organ.75, 901–919 (2021).
4. King, A. J., Douglas, C. M. S., Huchard, E., Isaac, N. J. B. & Cowlishaw, G. Dominance and affiliation mediate despotism in a social primate. Curr. Biol.18, 1833–1838 (2008).
5. Walker, R. H., King, A. J., McNutt, J. W. & Jordan, N. R. Sneeze to leave: African wild dogs (Lycaon pictus) use variable quorum thresholds facilitated by sneezes in collective decisions. Proc. R. Soc. B: Biol. Sci. 284, 20170347 (2017).
6. Johnstone, R. A., Cant, M. A., Cram, D. & Thompson, F. J. Exploitative leaders incite intergroup warfare in a social mammal. Proc. Natl Acad. Sci.117, 29759–29766 (2020).
7. Haig, D. Parental antagonism, relatedness asymmetries, and genomic imprinting. Proc. R. Soc. Lond. B Biol. Sci.264, 1657–1662 (1997).
8. King, D. C. & Zeckhauser, R. J. Punching and counter-punching in the US Congress: why party leaders tend to be extremists. Presented at the Conference on Leadership 2002: Bridging the Gap Between Theory and Practice (The Center for Public Leadership, 2002).