The cultural evolution and ecology of institutions

Abstract

Human societies are structured by what we refer to as 'institutions', which are socially created and culturally inherited proscriptions on behaviour that define roles and set expectations about social interactions. The study of institutions in several social science fields has provided many important insights that have not been fully appreciated in the evolutionary human sciences. However, such research has often lacked a shared understanding of general processes of change that shape institutional diversity across space and time. We argue that evolutionary theory can provide a useful framework for synthesizing information from different disciplines to address issues such as how and why institutions change over time, how institutional rules co-evolve with other culturally inherited traits, and the role that ecological factors might play in shaping institutional diversity. We argue that we can gain important insights by applying cultural evolutionary thinking to the study of institutions, but that we also need to expand and adapt our approaches to better handle the ways that institutions work, and how they might change over time. In this paper, we illustrate our approach by describing macro-scale empirical comparative analyses that demonstrate how evolutionary theory can be used to generate and test hypotheses about the processes that have shaped some of the major patterns we see in institutional diversity over time and across the world today. We then go on to discuss how we might usefully develop micro-scale models of institutional change by adapting concepts from game theory and agent-based modelling. We end by considering current challenges and areas for future research, and the potential implications for other areas of study and real-world applications. This article is part of the theme issue 'Foundations of cultural evolution'.

Keywords: barrier effects; cultural evolution; driven trend; institutions; political complexity.

Figure 1.

Population size, governmental roles and hierarchical levels of authority show correlated increases over time, i.e. as societies get bigger they also tend to get more complex in terms governmental and hierarchical institutions. The rows refer to the 1000-year time period prior to the dates indicated on the left. Frequency distributions are calculated based on cases drawn from a global, historical database [39]. x-axis values refer to log₁₀ polity population size (left), governmental roles index (proportional measure, 0: no such roles, 1: maximum possible recorded in sample), mean number of hierarchical levels (see the electronic supplementary material and [39] for details of dataset and measurements). The red arrow highlights the shift in the mode of these distributions towards greater scale and complexity which is indicative of a driven macro-evolutionary trend mechanism, suggesting there is some evolutionary force favouring larger, more complex societies [43] (see text). Note also the reduction in the relative frequency of smallest, least complex societies. (Online version in colour.)

Figure 2.

Countries that are more culturally distant from the earliest modern democratic country (USA) have tended to adopt open and competitive election institutions later than culturally similar countries (data taken from [52]). Cultural distance is based on time in years since the most recent common ancestor (MRCA) measured on a phylogenetic tree of Indo-European languages [54] (right). Where countries have adopted and abandoned elections we use the last calendar year that democratic election institutions were adopted (y-axis). Coloured (non-black) branches on the phylogenetic tree represent groups that have the same MRCA with the USA, which relate to the points on the left that share the same values on the x-axis. (Online version in colour.)

Figure 3.

Output from an agent-based simulation implementing an institutional form of punishment in a public goods game. A schematic (top left) shows how the model works. Agents have a fixed strategy of either being cooperators or free-riders in the economic game, and vote to decide on a level of punishment for free-riders (the political game). Pay-offs from the economic game determine agent fitness and thus feedback to affect the proportion of different strategies in the next time step of the simulation. The median cost value (black line) over 1000 generations of 100 simulation iterations is shown (top right), with an example from one iteration (grey line) to illustrate the kinds of dynamics seen in the model. The punishment cost value affects the pay-offs to agents in the economic game, giving cooperators (blue) higher pay-offs than free-riders (red) (bottom right), leading to changes in the proportion of cooperators such that they become the most common strategy (bottom left). Agents vote for punishment values that although they lead to lower pay-offs in the short term, in the longer term lead to pay-offs that are better for cooperators and any free-riders in the population (bottom right). Values shown in bottom graphs are medians from 100 simulation iterations.