This paper discusses the usefulness of metaphorical models for the study of economic and social phenomena. Metaphorical models are “low-fidelity” models of the real world: they do not reproduce it in its entirety, but focus on one or two relevant traits and embed them in a context of interactions that are plausible and realistic, but not an exact representation of the real world. The author argues that these models are useful for the study of real-world phenomena because they encourage people to make inductive inferences from the model to the real world, thus learning or understanding something more about it. As an illustration, the author discusses two models of the limits to growth in an industrialized society: a simpler one, based on the growth-and-decline pattern observed in harvest fisheries, and a more complex one called World Dynamics, developed by Jay Forrester. The first acts as a metaphorical model for the second, more complex model. Both models have the goal of representing the limits to economic growth and industrialization in our society on a global scale. The causes of such limits are understood to be “hidden” processes (such as the exhaustion of resources) which are not immediately apparent to the human actors involved in the model, but which end up curtailing growth in the long term.