My recent presentation at the NASSP annual conference, “Justifying-To As Moral Address,” I distinguished between “access to reasons” and “moral address” views of a general sort of necessary condition for state legitimacy (or permission to coerce). In the presentation, I interpreted Bernard Williams’ view of state legitimacy as a version of the second view. And also argued that there are reasons to think that the moral-address-focused approach, while distinctive, needs to appeal to a reasons-access condition in order to get its permission-generating work done.
I find the following striking in characterizing the justifying-to approach (definitive, supposedly, of public reason theorists) as simply a condition of reasons (or access to reasons):
“Some political philosophers and theorists place a requirement of public justification on the permissible use of state coercion or political power. According to these theorists the recognition of citizens as free and equal moral persons requires that coercion be justified for or to others by their own lights, or with reasons that they could recognize as valid. On this view, a public justification is achieved when members of the relevant public have adequate or sufficient reason to endorse a particular coercive proposal, law or policy. [from section 1 of SEP entry, “Public Justification,” Vallier & Zalta]
… The family of public reason liberal political theories… can be categorized in accord with a master principle, the Public Justification Principle (PJP)…
…A coercive law L is justified in a public P if and only if each member i of P has sufficient reason(s) Ri to endorse L.
… The justification is “public” because it is a justification that encompasses a public, though it does so severally and individually. It is not necessarily public in another important sense, i.e., the kind of publicity associated with common knowledge. So an individual’s reason is Ri need not be commonly known to others who are members of the same public P…” [from section 2 of the same SEP entry]
Just a little commentary and interpretation, then two substantive points.
[interpretation] I take the relevant conclusion here to be the practical (or practical-attitude) conclusion of the coercee endorsing L (including its coercive imposition on some group of people that includes oneself). However, to make my points below, I’ll switch from endorsement to acceptance — the reasons for this will become obvious. On any version of PJP, it would be an additional thing to add that the reasons to accept L need to be presented by the state (or anyone). What matters is simply that each person under the power of the state has sufficient reason to accept L. So PJP interprets all public reason approaches as having the necessary condition to be a condition (of or in the coercee) of reasons or access to reasons. I’m not that concerned about what views count as public reason views (if one rules out a substantial amount of belief and ignorance as unreasonable — and hence not in need of address for achieving reasonable acceptance — what we get might not look much like a public reason view of state legitimacy; as Williams’ view of state legitimacy does not look much like Rawls’). I’m interested in the narrow issue of what a broadly justify-to type of necessary condition on state legitimacy might be.
 The intuition behind PJP (or more generically a justify-to necessary condition on state legitimacy) seems to be something like this: obtaining actual (and reasonable-enough) consent (to coerce) being impractical, but certain collective ends being achievable only via state coercion, the way, or main way, for the state to respect the agency, reasoning and interests of those under its coercive power is to restrict its coercive power to those coercive elements that each under its power has sufficient reason to accept. (This condition strikes me as more necessary and sufficient than simply necessary — or at least as fully specifying one way of generating state legitimacy — but we can consider it simply as a necessary condition.) By contrast, one might focus on the condition of a coercer (e.g., the state) engaging in reason-giving behavior that exhibits and communicates adequate moral address or benign intent (so that the coercee has sufficient reason to interpret the coercion as something other than domination or some other form of abuse or exploitation). Arguably, something like this is Bernard Williams’ focus in his essay “Realism and Moralism” (my jumping-off point for the presentation). Such a condition is most plausibly construed as a minimal necessary condition for state coercion to be permissible (or for one mechanism of permissibility generation to do its work). Minimal in that quite a lot more would be required to specify the mechanism of permission-generation (and hence the sufficiency). (As Williams’ essay makes evident, it does not seem that the essential content of adequate moral address needs to involve the regarding of all people or all citizens as “free and equal” in anything like the sense that one would associate with liberalism historically (or with any more specific sense of this phrase associated with the public reason or political liberal approach to liberalism).)
 On either sort of view, I suspect that the right way to generate permissible coercion runs through the acceptability of the coercion to each relevant coercee. The basic meta-ethics-y, conceptual point here is this: plausibly, X’s coercing Y via PHI-ing being acceptable to Y (i.e., it being something that Y has sufficient reason to accept) is sufficient for X’s PHI-ing (with respect to Y) being permissible. (That might not be quite right. But something in this ballpark. And to get a general permission to PHI, we would need to specify the relevant group and relevant reasons for accepting the coercion of self or others such that the several acceptability is unified or non-accidental in some way.) On this kind of picture, (adequate, genuine) moral address by the coercer might be necessary, in an empirical real-world sense, in order for the coercion to be permissible because (a) without solid knowledge that an act of coercion is not mere domination or other form of abuse, the coercee would not have sufficient reason to accept the coercion and (b) the most reliable way to obtain this information is via relevant reasons-giving behavior of the coercer. Such moral address might be communicated via the sincere presentation of any number of reasons favoring the coercee accepting the coercion (“look, this is necessary to achieve something really important for everyone concerned, including you,” “look, this is something you should already want to do or contribute toward,” “look, the likely cost to you here is just not so high,” etc.; nothing sophisticated is required and neither is anything like an airtight case favoring acceptance). However, the basic mechanism for generating permissibility would be via the action being acceptable for each relevant party (relative to her background beliefs and values, at least if they are reasonable enough — something that public reason theorists and their critics get quite a bit of mileage out of). In this way and for these reasons, I think theorists who are focused on (adequate, genuine) moral address as the important justify-to type of necessary condition for state legitimacy should see moral address as operating through the more general reasons-access condition implicit in acceptability or reasonable acceptance.
Does my distinction in different possible “justify-to” necessary conditions for state legitimacy or permission to coerce seem clear and on-point (abstracting from what exactly is definitive of being a public reason liberal)? Does it seem plausible (due to an acceptability-to-permissibility connection or for other reasons) that a moral address necessary condition would operate through a reasons-access-related mechanism of generating normative permission? (Bonus question: does the sort of connection between acceptability and permissibility outlined seem like a promising way of generating permissibility via something like hypothetical consent or reasons that would support this (noting that no reference to the normative powers of actual consent, agreement, asking-then-getting-permission, etc. is made)?)