Simulating collective risk management from experimental data - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Simulating collective risk management from experimental data

Simuler la gestion du risque collectif à partir de données expêrimentales

Résumé

We propose a new analysis of the data from a large-scale experiment about managing collective risk (Szekely et al. 2021). We distinguish three behaviours (free riding, default cooperation, and fluctuating contribution) which were mixed in previous analyses. We show that free riding and default cooperation strongly determine the outcome of the game. Social norm features (personal normative belief, empirical expectation, normative expectation) are significant predictors of these behaviours but the decision models derived with or without these variables perform very similarly on the data. However, we also run two agent based models, one with agents not using normative features (reactive agents) the other with agents using them (normative agents). In the average result of numerous simulations, the normative agents tend to keep a significantly lower frequency of free-riding and a significantly higher frequency of default cooperation. This confirms the previous analyses concluding that the normative variables play a key role in the emergence of a norm.
Fichier principal
Vignette du fichier
ClimateChangeGame.pdf (446.85 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Licence : CC BY - Paternité

Dates et versions

hal-04211468 , version 1 (19-09-2023)

Identifiants

  • HAL Id : hal-04211468 , version 1

Citer

Omid Roozmand, Guillaume Deffuant, Giulia Andrighetto, Mario Paolucci. Simulating collective risk management from experimental data. Social Simulation Conference 2022, ESSA, Sep 2022, Milan (Italie), Italy. ⟨hal-04211468⟩
5 Consultations
3 Téléchargements

Partager

Gmail Facebook X LinkedIn More