The co-evolution of rational agency and social structure

Research output: Contribution to conferencePaper


The homo-economicus model helps us successfully predict social behaviour,
despite its oversimplifying premises. However, the model is methodologically
individualistic and thus, there is tension when describing the behaviour of
groups and societies as well as explaining their impact on individual behaviour.
I wish to suggest that human behaviour is best explained by considering
individual agency in the context of social structures as well as the interaction
between the two.
Individual behaviour determines the type of social structures that become
established through repeated interactions; at the same time, established social
structures set boundaries for individual behaviour. The argument is twofold: (1)
Rational agency is bound by a given social environment and (2) social
structures arise and become established as a result of repeated interactions. As
such, social structures are rational in that they ensure individual maximisation
and dynamic in that they are subject to change depending on individual
preferences. There are three assumptions necessary for this argument to hold:
Rational agency, iterated, non-random interactions and information availability.
In the Robinson Crusoe allegory, his strategies change when he meets Friday.
As the number and disposition of interacting agents change, rational agents'
strategies adapt to the changing environment. A utility-maximising agent has
to shift her behaviour depending on the interlocutors, the frequency, and sali-
ence of her interactions. Maximising behaviour depends on one's environment:
Crusoe's strategies are affected by his social environment and the number and
disposition of potential interlocutors. The repeated interactions between Crusoe
and Friday give rise to habits of behaviour, which replace rational analysis at
each decision node. So each interaction is not independent but part of a series
of interactions. Therefore, without abandoning rationality, Crusoe and Friday
stop behaving strategically in their set environment. In this framework, rational
agency and social structure are interdependent and co-evolve (Skyrms, 2004).
When agent A interacts with B and at the same time interacts with agent C his
experience from one interaction is transferred to the other. Therefore, A is a
link between the two interactions or among all the interactions she plays at any
given time and the strategies employed in these games. A’s behaviour is af-
fected by the outcome of each interaction and by the behaviour of B, C and so
on. Hence, their interaction is affected by their interaction history. In this con-
text, A and B never interact in a one-off interaction, as they perceive every in-
teraction part of their social life. This is a plausible account of human beha-
viour, given that we all live in societies, usually interacting in an environment
where past behaviour is to an extent known since interaction interdependence
also allows for information spreading (Young, 2001).
Humans are social animals and using exclusively individualistic models to ex-
plain and predict behaviour is misplaced and leads to errors. The proposed ac-
count is individualistic as its basis lies on individual behaviour. At the same
time however, social interactions are a fundamental premise in understanding
individual rationality.
Original languageEnglish
Publication statusPublished - Jun 2016

Fingerprint Dive into the research topics of 'The co-evolution of rational agency and social structure'. Together they form a unique fingerprint.

Cite this