Bridging Ideal and Non-Ideal Theory. An Incrementalist Approach to Normative Theorising

In this paper I propose and defend a framework for pursuing inquiries in political philosophy, which advances the idea that normative models built at the level of institutional design should be tested and selected by examining the implications which incremental changes in their inputs (set of assumptions) have on the outputs (set of normative principles) generated. In the first part I present the structure of a normative model, drawing on Stemplowska (2008), and I argue that proper evaluations of such models should take into account only the outputs, i.e. the combination of desirability and feasibility of normative principles, generated by the model, but that the fact-sensitivity of inputs cannot be itself used as a direct evaluative criterion, as Farrelly (2007) suggests. In the second part I offer an outline of the ideal/non-ideal theory debate, with a particular focus on theories of justice, and I argue that the various dimensions of this debate can be reduced to idealizations in inputs (with models being more or less significantly fact-insensitive) and to idealizations in outputs (with models weighing desirability more or less heavier than feasibility), illustrating some of the ways in which idealizations in inputs can affect the outputs generated by the model. In the third part I develop an incrementalist framework for normative theorising at the level of institutional design, drawing on the concept of derivational robustness analysis from the philosophy of economics. The framework proposed implies that normative inquiries of this type should start from parent-models (characterized by a particular set of inputs and outputs) and incrementally move in the direction of both ideal theory (by making the model less fact-sensitive) and non-ideal theory (by making the model more fact-sensitive) through offspring models, which differ from parent-models through singular modifications in the set of inputs and corresponding changes in the set of outputs generated. The value of this framework is threefold. First, it offers a method of systematizing knowledge in political philosophy. Second, it defuses the tension between ideal and non-ideal theory, by showing that neither of the two have a privileged status in normative inquiries and both are necessary not only for epistemological or evaluative pursuits, but also for practical or prescriptive ones. Third, it constitutes a guideline for what Simmons (2010) has called the transition from non-ideal to ideal theory, since it is able to offer a map of the various paths that might be taken from the present state of the world to a perfectly just state of the world, allowing us to select the path which would supposedly yield the best overall combination of desirability and feasibility.