False beliefs can bootstrap cooperative communities through social norms

Abstract

Building cooperative communities is a crucial problem for human societies. Much research suggests that cooperation is facilitated by knowing who the cooperators and defectors are, and being able to respond accordingly. As such, anonymous games are thought to hinder cooperation. Here, we show that this conclusion is altered dramatically in the presence of conditional cooperation norms and heterogeneous beliefs about others' behaviours. Specifically, we show that inaccurate beliefs about other players' behaviours can foster and stabilize cooperation via social norms. To show this, we combine a community's population dynamics with the behavioural dynamics of their members. In our model, individuals can join a community based on beliefs generated by public signals regarding the level of cooperation within, and decide to cooperate or not depending on these beliefs. These signals may overstate how much cooperation there really is. We show that even if individuals eventually learn the true level of cooperation, the initially false beliefs can trigger a dynamic that sustains high levels of cooperation. We also characterize how the rates of joining, leaving, and learning in the community affect the cooperation level and community size simultaneously. Our results illustrate how false beliefs and social norms can help build cooperative communities.

The impact of threshold decision mechanisms of collective behavior on disease spread

“…As the number of infectious individuals increases, the best response function shifts, which can shift equilibrium behavior from a state where only a few individuals adopt the NPI to a bistable system where most people will either adopt the NPI or not, depending on what they believe others do, and then to a state where almost all engage in the behavior. Such phenomena have been previously explored in the formation and dynamics of social groups ( 35 ). Fig.…”

“…We explore small positive or negative biases in these beliefs and found no large impact on the qualitative results ( SI Appendix , section 3 ); however, larger and more persistent biases driven, for example, by media or awareness campaigns can affect the dynamics more significantly. In a previous compartmental model of norm adoption, it was found that such inaccurate beliefs about population behavior could substantially impact the emergence of a social norm ( 35 ). In the context of NPIs, biases in beliefs about population behavior can arise through visibility or salience bias ( 58 ): for instance, people going out dining are more visible than people staying in, which can induce lower perception of NPI behaviors, and delay the onset of NPI adoption.…”

“…Further, adherence to norms of mask wearing can vary within a population due to group identity, which can frustrate behavioral interventions ( 29 – 31 ). Though norms can lead to coordination of human behavior ( 32 , 33 ) and the emergence of cooperative communities ( 34 , 35 ) thereby overcoming social dilemmas, they also can lead to harmful outcomes and be difficult to dislodge. It is important to incorporate these social phenomena into epidemiological theory and the design of public health control efforts.…”

Indirect reciprocity with Bayesian reasoning and biases

“…But what happens if such learning is not feasible or happens only slowly remains an open question. Under the norms where there is bistability, if a population initially has a high belief in the frequency of good individuals, that might be enough to bootstrap cooperative equilibria [ 53 ]. For error rates, individuals who intend to cooperate may know the rate at which they fail to cooperate, e 1 , through their own involuntary failures to cooperate, which could be transmitted via gossip to others.…”

Vaccination and collective action under social norms

“…The change in v ( t ) is determined by the Granovetter-Schelling social dynamic, Equation 1d. This dynamic was originally developed in the social science literature to model collective action [23–26, 30, 40, 41], and has recently been employed to model the uptake of non-pharmaceutical interventions during an epidemic [32]. f (Δ π ( I ( t ), v ( t ))) ∈ [0, 1] is a smoothed best response function of the difference between the payoffs to choosing to be vaccinated and choosing not be vaccinated, Δ π ( I ( t ), v ( t )) (which in turn is a function of the prevalence of infection and vaccination rate as discussed below).…”