An Ideal of Ideal-Free Democracy

Schumpeter and Przeworski (2003) put forth a – at first sight – very convincing argument for a procedural conception of democracy. Especially Przeworski’s main interpretation of procedural legitimacy as a tool of waging proxy-wars through the ballot and, thereby, avoiding real conflict, is an appealing prospect. One would, indeed, hope any convincing democratic institution to be a stand-in for tribal-instinctual animosity. Whenever violence does erupt after a democratic procedure – such as, e.g., in many cases of the human rights movements in the U.S. – one is poised to assess this as a failure of the democratic method.

However, the Schumpeterian view of procedural democracy, upon which both argumentations rest, cannot overcome many of the difficulties of democratic legitimacy that require a more substantive measure of legitimate outcomes. The most grappling of these is the fact that the procedural view, in the strictly Schumpeterian sense, cannot necessarily ensure the longevity of its democracy. If one restricts democratic legitimacy solely to procedure, nothing can stop an electorate from voting the democracy into an autocracy. Granted, Cohen (2009) wages that the procedural view does allow for some minor substantive measures of legitimacy, corresponding precisely to these political ‘liberties of the ancients’. However, this claim was only sidelined and more convincing argumentation would have to be given to remedy this major defect in the procedural view.

Furthermore, both Schumpeter and Przeworski correctly conclude that if one were to follow a non-procedural approach to democratic legitimacy, one would necessarily have to impose normative ideals upon the ensuing democratic institution. However, their striving towards an ideal-free democracy is, in itself, an ideal. Therefore, their procedural view cannot eliminate the necessity of setting democratic ideals. It is simply the choice of a minimal, non-restrictive normative ideal that they deem desirable. To this, one may put forth the argument – as many behavioural economists and social choice theorists do under the nomenclature of Thaler and Sunstein’s (2003, 2009) “libertarian paternalism” – that if an institution necessitates the setting of an ideal, one may as well choose the default as one that nudges subjects towards a desirable behaviour, whatever that behaviour may be. This “paternalism” seems to be closer to reality, as many prominent democratic institutions, such as the E.U. or the U.S., do set non-minimal normative ideals as parts of their own conceptions.

As such, Schumpeter’s procedural conception is, still, an idealistic theory of democratic legitimacy, which seems to rarely be realised in practice. Adding to the previous example, Schumpeter’s view that the ‘primary purpose of an electorate is to produce a government’ seems to stand in stark contrast to the operationalisation of democracy in many of its direct forms – ancient Athens or modern Switzerland, to name two examples. One needs, then, to ask themselves: is the procedural conception merely a philosophical construct, or can it be realised as a viable guide to democratic institutionalisation? Both Schumpeter and Przeworski fail to give a convincing positive answer to the second question.