Climate adaptation decisions involve heterogeneous stakeholders, partial information, and institutional constraints. This paper presents a role-based Multi-Agent Computer Model (MACM) to simulate climate adaptation deliberation across Living Labs. The main contribution is a configurable architecture that keeps behavioural mechanisms fixed while externalising context-specific inputs. Each stakeholder is represented as one agent that may play one or more roles: Expert Evaluator, Disseminator, Positioning Agent, and Decision-Maker. The simulation proceeds through four phases—initialization, information exchange, positioning and influence, and final decision-making. We argue that this role-based design improves both interpretability and cross-case reusability in social simulation of climate adaptation governance.

see attached

Agent-based models (ABMs) are widely used to study complex adaptive systems, yet the behavioral mechanisms encoded in agents’ decision rules often remain difficult to identify and compare across models. We propose Influence–Response Functions (IRFs) as a method for extracting, analyzing, and comparing these mechanisms. An IRF describes how variation in an influence variable maps onto an agent’s behavioral response under a given decision rule. We illustrate the approach using three related fraud ABMs that differ in how social influence shapes agents’ decisions. Using Monte Carlo simulation, we derive IRFs for each model and show that they reveal distinct response patterns associated with linear, threshold-based, and utility-based decision logics. For one model, we additionally show that the IRF can be recovered from ABM simulation data using a design-of-experiments approach. Overall, IRFs can improve transparency, support comparison of micro-level assumptions, aid behavioral verification and calibration, and complement existing documentation standards.

We developed relation-based modelling (RBM) as an approach for studying social-ecological systems (SES) from process-relational perspectives (PRP). It is an open and reflexive modelling practice to think with the emergence of a SES in ways that allow us to (i) question established categories of the social and the ecological, and ii) be explicit about how our relations with the system shape and are shaped by the emerging model and understanding. In this contribution, we describe a prototype relation-based model that builds on key process-relational ideas, such as the primacy of relations over entities, the concept of assemblages and possibility spaces. It was inspired by a case of self-governance in small-scale fisheries in Mexico. Compared to an agent-based model, taking a PRP changes what elements in the model are and how they can act, which has consequences for simulating the emergence (or lack thereof) of the self-organized fishery. We discuss differences between a relation-based and an agent-based model and how RBM fosters thinking differently about what constitutes social-ecological systems and how this affects their dynamics. These explorations have revealed potentials of the approach as a practice for process-relational analysis of SES but also challenges and limitations. We present some key reflections here.

This paper presents a case study of how ABM and Process

Tracing can be used together to analyse social phenomena. We use the

UK housing market crashes in 1989 and 2008 to tell ‘justified stories’, or

empirically grounded causal stories, of how these housing market crashes

came about. We conclude that combining Process Tracing with ABM

allows for greater depth in the empirical grounding of ABM and a possi-

bility to generalise and test causal hypotheses entertained and evidenced

by Process Tracing.