Abstract for: Places to intervene in a model: Using Meadows' leverage points to assess model evolution in participatory system dynamics

Meadows’ Leverage Points have been widely appropriated in the field of sustainability science as a powerful boundary object in which scholars and stakeholders from multiple backgrounds can find a common language to discuss system transformation. However, their use remains largely absent in the system modelling literature. Here, we propose that the leverage points can be used as an analytic tool to understand model evolution across an SD participatory modelling cycle. We propose a structured protocol to connect stakeholder input with model changes classified according to the 12 leverage points proposed by Meadows. By taking a modelling cycle perspective, we were able to evidence that model changes evolve from deep to shallow over time. We illustrate how stakeholder feedback in early stages of the modelling cycle lays ‘deep’ foundations that determine the model capabilities of later stages—converging to ‘shallow’ refinements. Our approach responds to previous literature calls for more transparent and reflective modelling practice in SD. Findings point out that careful parameterisation and even advanced feedback analysis remain ‘shallow’ compared to the ‘deep’ foundations of a model. Further empirical research can build on our methods to structurally explore model evolution across a broad set of SD participatory modelling settings.