The emergence of cooperation in human societies is often linked to environmental constraints, yet the specific conditions that promote cooperative behavior remain an open question. This study examines how resource unpredictability and spatial dispersion influence the evolution of cooperation using an agent-based model (ABM). Our simulations test the effects of rainfall variability and resource distribution on the survival of cooperative and non-cooperative strategies. The results show that cooperation is most likely to emerge when resources are patchy, widely spaced, and rainfall is unpredictable. In these environments, non-cooperators rapidly deplete local resources and face high mortality when forced to migrate between distant patches. In contrast, cooperators—who store and share resources—can better endure extended droughts and irregular resource availability. While rainfall stochasticity alone does not directly select for cooperation, its interaction with resource patchiness and spatial constraints creates conditions where cooperative strategies provide a survival advantage. These findings offer broader insights into how environmental uncertainty shapes social organization in resource-limited settings. By integrating ecological constraints into computational modeling, this study contributes to a deeper understanding of the conditions that drive cooperation across diverse human and animal systems.

Ferrari, S., Lammers, W., Wenmackers, S. (forthcoming) How the structure of scientific communities could impact the public uptake of uncertain science. Philosophy of Science.

A formalized implementation of Halstead and O’Shea’s Bad Year Economics. The agent population uses one of four resilience strategies in an attempt to cope with a dynamic environment of stresses and shocks.

FIsheries Simulation with Human COmplex DEcision-making (FISHCODE) is an agent-based model to depict and analyze current and future spatio-temporal dynamics of three German fishing fleets in the southern North Sea. Every agent (fishing vessel) makes daily decisions about if, what, and how long to fish. Weather, fuel and fish prices, as well as the actions of their colleagues influence agents’ decisions. To combine behavioral theories and enable agents to make dynamic decision, we implemented the Consumat approach, a framework in which agents’ decisions vary in complexity and social engagement depending on their satisfaction and uncertainty. Every agent has three satisfactions and two uncertainties representing different behavioral aspects, i.e. habitual behavior, profit maximization, competition, conformism, and planning insecurity. Availability of extensive information on fishing trips allowed us to parameterize many model parameters directly from data, while others were calibrated using pattern oriented modelling. Model validation showed that spatial and temporal aggregated ABM outputs were in realistic ranges when compared to observed data. Our ABM hence represents a tool to assess the impact of the ever growing challenges to North Sea fisheries and provides insight into fisher behavior beyond profit maximization.

The command and control policy in natural resource management, including water resources, is a longstanding established policy that has been theoretically and practically argued from the point of view of social-ecological complex systems. With the intention of making a system ecologically resilient, these days, policymakers apply the top-down policies of controlling communities through regulations. To explore how these policies may work and to understand whether the ecological goal can be achieved via command and control policy, this research uses the capacity of Agent-Based Modeling (ABM) as an experimental platform in the Urmia Lake Basin (ULB) in Iran, which is a social-ecological complex system and has gone through a drought process.

Despite the uncertainty of the restorability capacity of the lake, there has been a consensus on the possibility to artificially restore the lake through the nationally managed Urmia Lake Restoratoin Program (ULRP). To reduce water consumption in the Basin, the ULRP widely targets the agricultural sector and proposes the project of changing crop patterns from high-water-demand (HWD) to low-water-demand (LWD), which includes a component to control water consumption by establishing water-police forces.

Using a wide range of multidisciplinary studies about Urmia Lake at the Basin and sub-basins as well as qualitative information at micro-level as the main conceptual sources for the ABM, the findings under different strategies indicate that targeting crop patterns change by legally limiting farmers’ access to water could force farmers to change their crop patterns for a short period of time as long as the number of police constantly increases. However, it is not a sustainable policy for either changing the crop patterns nor restoring the lake.

Routes & Rumours is an agent-based model of (forced) human migration. We model the formation of migration routes under the assumption that migrants have limited geographical knowledge concerning the transit area and rely to a large degree on information obtained from other migrants.

This is an extension of the original RAGE model (Dressler et al. 2018), where we add learning capabilities to agents, specifically learning-by-doing and social learning (two processes central to adaptive (co-)management).

The extension module is applied to smallholder farmers’ decision-making - here, a pasture (patch) is the private property of the household (agent) placed on it and there is no movement of the households. Households observe the state of the pasture and their neighrbours to make decisions on how many livestock to place on their pasture every year. Three new behavioural types are created (which cannot be combined with the original ones): E-RO (baseline behaviour), E-LBD (learning-by-doing) and E-RO-SL1 (social learning). Similarly to the original model, these three types can be compared regarding long-term social-ecological performance. In addition, a global strategy switching option (corresponding to double-loop learning) allows users to study how behavioural strategies diffuse in a heterogeneous population of learning and non-learning agents.

An important modification of the original model is that extension agents are heterogeneous in how they deal with uncertainty. This is represented by an agent property, called the r-parameter (household-risk-att in the code). The r-parameter is catch-all for various factors that form an agent’s disposition to act in a certain way, such as: uncertainty in the sensing (partial observability of the resource system), noise in the information received, or an inherent characteristic of the agent, for instance, their risk attitude.

Consumer agents make choices which products to choose using the consumat approach. In this approach agents will make choices using deliberation, repetition, imitation or social comparison dependent on the level of need satisfaction and uncertainty.
The model is discussed in Introduction to Agent-Based Modeling by Marco Janssen. For more information see https://intro2abm.com/

Motivated by the emergence of new Peer-to-Peer insurance organizations that rethink how insurance is organized, we propose a theoretical model of decision-making in risk-sharing arrangements with risk heterogeneity and incomplete information about the risk distribution as core features. For these new, informal organisations, the available institutional solutions to heterogeneity (e.g., mandatory participation or price differentiation) are either impossible or undesirable. Hence, we need to understand the scope conditions under which individuals are motivated to participate in a bottom-up risk-sharing setting. The model puts forward participation as a utility maximizing alternative for agents with higher risk levels, who are more risk averse, are driven more by solidarity motives, and less susceptible to cost fluctuations. This basic micro-level model is used to simulate decision-making for agent populations in a dynamic, interdependent setting. Simulation results show that successful risk-sharing arrangements may work if participants are driven by motivations of solidarity or risk aversion, but this is less likely in populations more heterogeneous in risk, as the individual motivations can less often make up for the larger cost deficiencies. At the same time, more heterogeneous groups deal better with uncertainty and temporary cost fluctuations than more homogeneous populations do. In the latter, cascades following temporary peaks in support requests more often result in complete failure, while under full information about the risk distribution this would not have happened.

Both models simulate n-person prisoner dilemma in groups (left figure) where agents decide to C/D – using a stochastic threshold algorithm with reinforcement learning components. We model fixed (single group ABM) and dynamic groups (bad-barrels ABM). The purpose of the bad-barrels model is to assess the impact of information during meritocratic matching. In the bad-barrels model, we incorporated a multidimensional structure in which agents are also embedded in a social network (2-person PD). We modeled a random and homophilous network via a random spatial graph algorithm (right figure).