In the first chapter of AGAINST DEMOCRACY, “Hobbits and Hooligans,” Jason Brennan (JB) endorses the idea, from John Stuart Mill, that we should institute whatever form of government produces the “best results” [p. 1-2]. He lists the following among the important good results that a government might produce: (a) respecting liberal rights, (b) promoting economic growth and (c) promoting intellectual and moral virtue among the citizenry.
Because he includes [a], JB is clearly thinking of consequences or outcomes or results in a very broad way that includes adherence to deontic constraints or requirements. Though at one very general level this is all fine and well, part of the evaluative picture here is supposed to be that some form of government other than democracy might turn out to be best. This picture makes sense if we are examining results or consequences narrowly construed (such that adherence to deontic constraints does not count as an outcome). It also makes sense if we suppose that democracy is merely a way of discovering, formulating and enforcing liberal rights (say, on some Lockean conception).
But the set of liberal rights might include a right to vote! But if so then the possibility that some form of government other than democracy could beat out democracy is rather narrow and bleak. For all acceptable forms of government, on this normative scenario, are democratic. Only if there were a situation in which a government had to choose between adhering to this constraint and some more important constraint (e.g., not torturing its citizens) would some non-democratic form of government be best in some context.
All of this is obscured by JB offering what verbally appears to be a hard-nosed consequentialist standard for evaluating government – but that really is not. The effect here is that of brushing under the carpet a deontic consideration that goes something like this: in the context of collective action/activity, each participant is due a voice in how the process unfolds with regard to the important costs and benefits involved/produced.
How this requirement is best satisfied in the context of modern government is an interesting question and here I’ll provisionally agree with JB that traditional forms of universal suffrage have severe disadvantages that might, in some contexts, make them inferior to less robust forms of democracy. But we can only address these interesting issues in a realistic way with the relevant voting-rights-related constraint on collective action (specifically government action) clearly in view.
(Elsewhere, JB does address deontic justifications for democracy. In an important sense, addressing his explicit positions and arguments on this is where the action is at. My complaint here is that his verbally consequentialist standard for evaluating different forms of government is potentially confusing and obscures what is arguably the most important consideration in evaluating different forms of government.)