Demarcating normative political theory from empirical political science - and why it methodologically matters

Recent debates in the methodology of normative political theory (NPT) have suggested different approaches regarding the function, location and relevance of empirical political science (EPS) vis-a-vis NPT. Such approaches can be classified into three rough categories: ‘pure normative’, ‘hybrid normative-empirical’ and ‘pure empirical’. In ‘pure normative’ EPS plays no part at the level of forming principles and theories in NPT (represented by Estlund, Walsh and McDermott); in ‘hybrid normative-empirical’, EPS plays a part in forming theories in NPT (represented by Dowding, Perez, Blau, de-Shalit); in ‘pure empirical’ EPS supersedes other considerations (represented by Jonathan Floyd and earlier positivists). The debate between these three options has analytical-methodological aspects - attempting to clarify the structure of NPT; and also prescriptive aspects - as each option would point to a different kind of research in NPT that is desirable with important implications vis-a-vis future research in NPT, and also the kind of training that future political theorists will receive. In response to these three options, the current article suggests clarifying the function and location of EPS in NPT. In order to do so it begins with suggesting a simple demarcation principle, as follows: for a study to count as a normative political theory it must include an evaluative step, in which evaluation is grounded in some principle of justice broadly conceived. If this demarcation is adopted, it would mean that empirical political science and normative political theory would share many attributes, as examining the research design of both, will point to numerous similar characteristics, such as description and analysis of political institutions, rules of inference, prescriptive parts, the usage of models and more. The reason is that the evaluative part in NPT typically takes up only a specific, particular part of NPT, while other parts are not unique to NPT. If this demarcation principle is adopted, it arguably means that the ‘pure normative’ category is under-representative, as it excludes what is central to many studies in political theory - the evaluation of concrete political institutions; similarly, the ‘pure empirical’ category would also be under-representative, excluding a major part of seminal studies in contemporary political theory - the grounding of the evaluation of political institutions in some principle of justice. Normative-empirical hybridity, while perhaps less elegant, is the most accurate description of the structure and methodology of normative political theory research as recognized by a survey of several of the field's contemporary well known studies. An understanding of this demarcation principle would fulfill three goals. The first is analytical; it aims to suggest a better understanding of the differences and overlaps between NPT and EPS. The second goal is to better comprehend the role that EPS will fulfill within NPT. Arguably, most methodological approaches to NPT would include substantial parts that utilize the methods and research outcomes of EPS. A third and practical goal concerns the methodological implications of the demarcation principle; arguing that the methodology of NPT will be pluralistic, containing both evaluative techniques and numerous empirical and inferential techniques.