Model trustworthiness and modeler responsibility in economic agent-based modeling practices: a meta-analytical approach

Establishing credibility and trustworthiness is essential in Economic Agent-Based Modeling (ABM), where clear epistemic standards cannot be defined a priori. In this paper, we first review the notions of trustworthiness and credibility in modeling. We then introduce a framework that emphasizes the modeler's epistemic responsibility to ensure coherence between modeling purposes, strategies and targets. We examine the challenges in assessing model reliability that arise from the interaction of conceptual, algorithmic and computational constituents, and we propose a meta-analytical approach to enhance model consistency by conceptualizing agent-based simulations as iterated analogies. Our analysis outlines strategies for improving model accessibility and reliability while highlighting the modeler's role in preventing mistargeting and misuse. This research provides a normative basis for justifying the credibility of both idealized and targetless models by promoting transparency and consistency between model design and intended purposes.