Models can have pedagogical, scientific, policy, exploratory, and metaphoric uses, among many other possible uses. The uses overlap, and when thinking about the usefulness of models it is necessary to be clear about the use to which the model is being put. One can define usefulness of a model only in relations to its intended use. This paper argues that Phillips' intended use of his MONIAC was pedagogical, and used as «a» model, it can serve a useful pedagogical purpose. But it cannot serve as «the» model for framing policy since it does not capture the complexities of non-linear dynamics that can affect the aggregate economy. It argues (1) that Keynes understood this distinction, but that the distinction was lost by later economists, and that it needs to be rediscovered and (2) that the profession should train economists to specialize in developing, interpreting, teaching, and applying models to policy.