Here is a quick argument for non-voluntary (and hence non-consent-based) normative authority. No doubt this needs some tightening-up or is otherwise flawed. And I have to do a lot more reading about the various “fair play” approaches to political authority (and authority generally). But right now, something like this seems pretty compelling to me.
(1) When one is participating in a collective endeavor or institution of value, one is obligated to contribute one’s fair share (irrespective of whether this maximizes or optimizes utility relative to the other ends in view – in this sense, the obligation is thought of as deontological).
(2) Not all such participation is voluntary: in some important cases, one simply finds oneself participating and cannot, not without unacceptably high cost to oneself, escape or opt out (e.g., participation in systems of language and communication).
(3) So, in these non-voluntary-participation cases, one in this sense cannot get out of the obligation to contribute one’s fair share.
(4) In some further subset of these non-voluntary-participation cases, all or part of one’s fair share is personal compliance with the social rules.
(5) Sometimes these rules include those specifying that one is to obey the (relevant-enough, competent-enough but not necessarily that great) orders of qualified persons who occupy certain roles or offices.
(6) In these cases, the only way to realize the non-voluntary obligation to fairly contribute is by obeying authorities (people in relevant roles or offices) in ways that constitute normative authority.
Unlike Estlund’s argument, this argument does not require that the collective task be urgent, important, super-valuable, etc. Also unlike Estlund, I don’t worry that everyone will be obligated to provide fair contribution to a valuable task that only all of us around here are engaged in. For it is the particular value to the participants, not any strictly agent-neutral value of the endeavor or its aims, that I take to do the work. I also don’t have a “normative consent” condition here (requiring that, in order to get authority, it needs to be the case that you are obligated to agree to obey).
However, by adding qualifications to this most-basic sort of case, it is easy to move in a broadly liberal direction. It seems likely that, especially in social conditions adequately propitious to liberal social arrangements, the value of and obligations to respect individual autonomy – and as well considerations of fair play in choosing between the various adequate options in schemes of social rules that we enforce – might amount to additional conditions on authority. For example, schemes of social rules might, for small or questionable social gain, (a) impose high autonomy-costs on individuals or (b) impose differential compliance costs across different types of individuals (arguably, something that is unavoidable) but without providing adequate space for public objection, persuasion and social or political organization. If either of these things are true, it is plausible that we are not obligated to comply with at least some of the social rules that we would otherwise be obligated to comply with. I suspect that there are other broadly liberal qualifications for authority (or political authority) that would plausibly appropriately qualify the story of fundamentally non-voluntary fair-contribution-based obligation here.