Collective decision making by rational agents with differing preferences

Abstract

<jats:p>Collective decisions can emerge from individual-level interactions between members of a group. These interactions are often seen as social feedback rules, whereby individuals copy the decisions they observe others making, creating a coherent group decision. The benefit of these behavioural rules to the individual agent can be understood as a transfer of information, whereby a focal individual learns about the world by gaining access to the information possessed by others. Previous studies have analysed this exchange of information by assuming that all agents share common goals. While differences in information and differences in preferences have often been conflated, little is known about how differences between agents’ underlying preferences affects the use and efficacy of social information. In this paper I develop a model of social information use by rational agents with differing preferences, and demonstrate that the resulting collective behaviour is strongly dependent on the structure of preference sharing within the group, as well as the quality of information in the environment. In particular, I show that strong social responses are expected by individuals that are habituated to noisy, uncertain environments where private information about the world is relatively weak. Furthermore, by investigating heterogeneous group structures I demonstrate a potential influence of cryptic minority subgroups that may illuminate the empirical link between personality and leadership.</jats:p>

Optimal use of simplified social information in sequential decision-making

“…For each of these two scenarios I calculate how an agent should rationally respond to the social information they are presented with in order to maximise their expected utility, making the assumption that agents are aware that decisions have been made sequentially (although they cannot observe the full sequence), and that other agents are subject to the same limitation as themselves. I focus on identical agents (those sharing the same utility function and cognitive processes) as previous research has shown that agent preferences must be strongly aligned to produce significant social information use [31], and this assumption also permits a more straightforward mathematical treatment. These calculations (see Methods) result in a single number associated with each possible distinct observable social information, which denotes a critical threshold for the agent's private information: if their private information exceeds this threshold they will choose one option, otherwise they will choose the other.…”

“…In common with previous developments of this model [30,31], I evaluate the probability of an agent choosing A or B by considering the position of an external observer, for whom the private information of all agents is unknown. In this paper I assume that all observations are made under natural conditions, meaning that the experimental noise level is the same as the habitual noise level (see ref [30] for a discussion on the role of experimental noise).…”

“…Recent theoretical work on sequential decision-making identified the optimal response of an agent to observing a sequence of prior decisions [30,31]. These studies showed that optimal use of that information involves the full ordered sequence of decisions.…”

Optimal use of simplified social information in sequential decision-making

“…Recent theoretical work on sequential decision-making identified the optimal response of an agent to observing a sequence of prior decisions [30,31]. These studies showed that rational use of information from previous decisions makes use of full sequence information.…”