This article has two main goals: to examine and classify the ways data can be used to advance normative theorizing in political theory, and to demonstrate such usages in the contested disciplinary field of religion–state relations and specifically regarding the hotly debated model of the separation of religion and state. Regarding the former, it is suggested here that the general observation that evaluation of political institutions must rely on proper understanding of such institutions and hence be data-sensitive, can be further (...) detailed or classified, via the following categories: conversion, institutional clarity, theoretical clarity and finally, theory improvement. With regard to the latter we examine how data regarding religion–state relations enables a better understanding (via the noted novel classifications or categories) of two important debates: ‘delineation’ (i.e. the argument that pure separation of religion and state does not exist) and ‘separation-establishment’ (i.e. arguing that models other than separation are permissible in religion–state relations). (shrink)