aalto1 untyped-item.component.html

A Bayesian causal analysis of the effect of moralizing supernatural punishment on cocietal resilience to collapse risk

Loading...
Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

School of Science | Master's thesis

Authors

Department

Mcode

Language

en

Pages

44

Series

Abstract

This thesis investigates the causal effect of moralizing supernatural punishment on societal resilience to crisis using systematic historical data from the Seshat Global History Databank. The "Big Gods" hypothesis suggests that moralizing religions enhance social cooperation and should improve crisis outcomes, but empirical testing of this theory has been limited by methodological challenges in historical analysis. We employ Directed Acyclic Graphs to formalize competing causal hypotheses—whether social complexity confounds or mediates the ideology-resilience relationship—and implement a principled Bayesian workflow to develop robust multilevel regression models accounting for geographical clustering and overdispersion in crisis severity data. Using a causal mediation analysis of 169 crisis events in 150 unique polities across ten world regions spanning four millennia, we find that Moralizing Supernatural Punishment(MSP) operate through competing mechanisms that largely offset each other, resulting in near-zero net effects on crisis outcomes while revealing both protective and vulnerability-enhancing pathways. In contrast, structural factors including social complexity and military technology show robust positive associations with crisis severity, while substantial geographical variation indicates that regional context dominates ideological factors in determining societal responses to existential challenges. While these null findings suggest that structural capacity and geographical context may be more fundamental determinants of societal resilience than moralizing religions, they confirm the value of systematic quantitative methods for testing historical hypotheses.

Description

Supervisor

Vehtari, Aki

Thesis advisor

Albarracin, Mahault
Friedman, Daniel Ari

Other note

Citation

Endorsement

Review

Supplemented By

Referenced By