Abstract

Religion is well-known to correlate with cooperative behavior, particularly within religious groups. In light of a sudden decline of religion in Western societies and rising indicators of loneliness and social fragmentation, it becomes pressing to understand not only why this connection occurs, but also whether its underlying mechanisms can be sustained under secular conditions. The precise mechanisms driving this effect, however, remain a matter of ongoing debate in the scientific community. There is no shortage of models aiming to explain the phenomenon, and the offered explanations are as numerous and diverse as are the researchers promoting them. This paper argues two main conceptual points from a philosophical point of view: First, many existing models oversimplify the underlying mechanisms by emphasizing one or a few stable explanatory factors, leading to shortcomings in both retrospective explanation and prospective prediction. Second, a more adequate approach requires an actor-oriented model that accounts for fluctuating relationships both within individuals and between individuals and their circumstances. Such models, rooted in recent advancements in causal inference for social networks, recognize that certain factors within and across religious groups are inherently unstable but still secure cooperation through multiple, dynamic pathways. This proposed model promises a more nuanced and scientifically robust reconstruction of the relationship between religion and cooperation.