Model Transfer, the Challenge of Tractability Standards and Methodological Embedding

Abstract Domain-specific epistemic priorities determine which procedures for solving computational models are accepted in a given domain. Tractability assessments thus depend on how disciplines resolve epistemic trade-offs, and ignoring this domain-specificity renders the notion of tractability underspecified. I argue that recognizing domain-specific tractability standards reveals an overlooked challenge to model transfer: skepticism about a model’s suitability as a computational device in the new domain. Based on the case of a hybrid agent-based model in macroeconomics, I demonstrate that this challenge can sometimes be mitigated by reformulating the model’s representational mode to align with the new domain’s tractability standard.