State neutrality is a key idea in liberal thought. It is grounded in the even more fundamental claim that as morally free and equal entities, individuals should not be coerced into ways of life that they don’t endorse, accept, or value, unless there is a particular reason that they themselves acknowledge justifying that they are coerced. This particularly applies to the state as the organization that has the monopoly for legitimately using coercion. A condition of legitimacy is precisely that the state respects the status of individuals as morally free and equal.
The idea of neutrality is fairly broad and general. Can we give it more substance? Let’s start with the following definition of the concept:
An agent A, doing a, is neutral between B and C with respect to a difference D if a doesn’t treat B and C differently on the basis of D.
Given this definition of the concept of neutrality, conceptions of neutrality will differ depending on A, B, C, and D are referring to. Because we are concerned with state neutrality, A here corresponds to the state and a to some coercive interference. A first conception interprets B and C as conceptions of the good or comprehensive doctrines and D as a difference in terms of the values and their relative importance that are singled out by these conceptions or doctrines.
John Rawls
On this conception, the state is neutral if it refrains from coercing people based on a particular comprehensive doctrine, that is for moral, religious, or epistemological reasons that are specific to a particular doctrine. This conception of neutrality is essentially the one endorsed by Rawls’s political liberalism. Rawls indeed insists that we should distinguish between political conceptions of justice and comprehensive doctrines. The former are free from any moral, religious, or epistemological commitments. Political conceptions of justice are “freestanding” in the sense that their (pro tanto) justification is independent of moral or religious beliefs. They are to be asserted uniquely based on reasons related to the fundamental claim that individuals are morally free and equal entities. Comprehensive considerations related to moral or religious beliefs are relevant only in the second justificatory step (what Rawls calls “full justification”) where the well-ordered society depends on the existence of an overlapping consensus such that the political conception of justice can be affirmed and endorsed by each individual from within their comprehensive doctrines. Comprehensive doctrines cannot by themselves justify the imposition of constitutional and justice principles to other persons, but for society to be stable we have to make sure that principles of justice are compatible with people’s (reasonable) moral and religious beliefs.
There are at least two interesting features of this Rawlsian conception of liberal neutrality to be emphasized. First, Rawls is explicit that its scope is restricted to “constitutional essentials and matters of basic justice”. It doesn’t apply in the case of laws that concerned matters outside this scope. For instance, the choice to favor a system with public or private schools seems not to fall within the scope of public reason, meaning that is permissible to bring comprehensive reasons into the debate. Second, it favors what some commentators have characterized as “epistemic abstinence” in the domain of public reason.[1] On this conception of neutrality, when constitutional essentials and matters of basic justice are at stake, people should refrain from invoking truth to defend their views. Claims of truth are indeed fundamentally related to comprehensive beliefs that cannot be used to justify coercive interferences based on principles of justice.
These two aspects make the Rawlsian conception a contentious one. The latter aspect underlines the very restrictive implications of Rawlsian neutrality. For it is not only truth that is prohibited from the domain of public reason, but also many beliefs and commitments that are constitutive of individuals’ personal and social identities. Also, it is unclear how controversial but still useful scientific results (say, theoretical and empirical results on the effects of various tax schemes) can be used to evaluate principles of justice. The former aspect seems to lessen this problem. After all, the scope of neutrality is restricted, which means that epistemic abstinence should not be widespread. But this has an unwelcome implication from a liberal perspective: it seems that outside this scope, state coercion based on the non-neutral treatment of comprehensive views is permissible and justified. This problematic implication is magnified once we realize that Rawls’s distinction between the domains of the political and the comprehensive is far from clear. The most plausible interpretation is that the domain of the political for Rawls is the one where a consensus based on shared public reasons is possible and as a matter of fact exists in society. Obviously, however, the more society is diverse, the smaller the domain of the political is. That means in diverse societies, either stability is compromised, or non-publicly (i.e., based on non-shared comprehensive reasons) justified state coercion is necessary.
Gerald Gaus
An alternative conception of liberal neutrality is obtained if B and C refer to persons, and D to a difference between B’s and C’s (political and comprehensive) beliefs. Here, we simply drop the political/comprehensive distinction. People have beliefs about what is fair, good, what is permissible or obligatory and, in diverse societies at least, their beliefs are likely to be conflictual and inconsistent. Neutrality means that the state should not treat a person differently than another one on the ground that they do not have the same beliefs. This is the conception developed in particular by Gerald Gaus in his version of public reason liberalism.[2] Neutrality is associated with a non-coercion principle according to which an agent A (either the state or an individual) should not coerce an agent B, unless A coercing B is justified by reasons RA and RB that are mutually intelligible for A and B (RA and RB can be the same, though it’s not necessary).
This conception has at least two advantages compared to the Rawlsian one. First, it dispenses with the problematic political/comprehensive distinction. Public justification does not depend on the fact that individuals agree based on shared reasons, only that they each have a non-defeated reason to accept a law. This also implies that the scope of public reason is potentially fairly wide and encompasses issues unrelated to constitutional essentials and matters of basic justice. Second, The Gausian conception doesn’t have the restrictive features of the Rawlsian conception. There is no duty of epistemic abstinence, as truth (or the belief that something is scientifically proven) is among the intelligible reason that can ground public justification.
Gaus rightly notes that this alternative conception of liberal neutrality has implications that Rawlsian may not accept. Basically, it has what I would call a bias for natural liberty. The principle of non-coercion indeed implies that if an individual has an intelligible non-defeated reason to reject a coercive law, it is immoral and so presumably illegitimate for the state to impose it on her. The default option is “blameless liberty”, a state of affairs where individuals are presumed to be free to do what they want. On many issues, every possible law will be met by a non-defeated intelligible reason to reject it. The room for state coercion is therefore severely restricted under this conception of neutrality. Gaus claims that on more fundamental issues (such as agreeing on a system of private property rights), individuals should acknowledge that they have a non-defeated reason to accept that it’s better to have a law than none, therefore limiting the risk for society to degenerate into anarchy. It is unclear however that this worry can be easily dismissed in deeply pluralistic and conflictual societies.
[1] Joseph Raz, “Facing Diversity: The Case of Epistemic Abstinence,” Philosophy & Public Affairs 19, no. 1 (1990): 3–46.
[2] Gerald Gaus, Public Reason and Diversity: Reinterpretations of Liberalism, ed. Kevin Vallier, New edition (Cambridge, United Kingdom: Cambridge University Press, 2022), Chapter 4.