This paper contributes to the debate—traceable at least to Condorcet (1785)—on designing collective decision rules for multiple-alternative voting. This phenomenon manifests recurrently in high-stakes institutional contexts, including constitutional conventions, electoral system reforms, and multi-option referenda. We focus on three influential rationalization frameworks: epistemic (accuracy-seeking), distance-based (consensus-oriented), and strategic (game-theoretical robustness). While each offers valuable guidance, they can generate conflicting prescriptions and raise significant computational complexity issues (of particular importance for electronic democracy). Drawing on recent literature in philosophy, economics, and computational social choice, and inspired by Sunstein’s notion of “incomplete theoretical agreements,” we propose a lexicographic meta-rule that hierarchically orders those criteria to achieve decisions that are normatively balanced and computationally tractable. The approach is particularly relevant to constitutional settings, where institutional design must reconcile diverse normative commitments under practical constraints.
