The Limits of State Neutrality
Originally posted June 9th, 2006
The principle of state neutrality enjoys great prominence in contemporary liberal thought. It is espoused, perhaps most persuasively, by the political philosopher John Rawls. There are of course many different ways in which the principle of state neutrality can be stated, so it is important to first clarify what the principle entails before I address its limitations.
The kind of neutrality that liberals like Rawls have in mind is called “neutrality of aim” or justificatory neutrality. This version of neutrality maintains that the justification for a law or policy should be neutral. It should not presuppose, for example, values particular to one conception of the good. Piety is a clear example of a value which would violate justificatory neutrality. The religious state is the archetype of the non-neutral state.
Why do liberals argue for justificatory neutrality? In Beyond Neutrality George Sher (p. 15) identifies three reasons why liberals defend state neutrality. These are:
1. because non-neutral government decisions violate the autonomy of citizens.
2. because non-neutral government decisions pose unacceptable risks of oppression, instability, or error.
3. because non-neutral government decisions rest on value-premises that cannot be rationally defended.
By requiring the justification for a law or policy be neutral, the neutrality constraint ensures that only those laws or policies that rest on value-premises reasonable citizens of a pluralist society could accept are legitimate. The neutrality constraint thus rules out many intolerant measures. For example, it rules out laws against religious heresy. The religious state is non-neutral because its laws and policies are based on values (e.g. piety) particular to one conception of the good. A non-neutral value like piety should not trump the political values of liberty and equality.
The neutrality principle may be of some use in helping us avoid the oppression of a perfectionist state but it is important to recognise that it is principle whose usefulness tends to be overemphasised by liberals who make it the cornerstone of a liberal public ethic. Consider, for example, a repressive policy like the criminalization of homosexuality. Neutralist liberals might believe that the principle of neutrality can easily dismiss such policies as they will be premised on perfectionist values (e.g. religion). But the most influential arguments for such a prohibition are not usually framed in such a way that its violation of the neutrality constraint is so obvious.
In The Enforcement of Morals Lord Devlin (p. 10) criticised the Wolfenden Report (1957) which recommended that homosexual acts between consenting adults in private be decriminalised. Devlin’s argument for such a prohibition was premised on a conception of legal moralism. He argued:
If men and women try to create a society in which there is no fundamental agreement about good and evil they will fail; if, having based it on common agreement, the agreement goes, the society will disintegrate. For society is not something that is kept together physically; it is held by the invisible bonds of common thought. If the bonds were too far relaxed the members would drift apart. A common morality is part of the bondage. The bondage is part of the price of society; and mankind, which needs society, must pay its price.
Neutralist liberals might dismiss Devlin’s argument on grounds that a prohibition on homosexuality is non-neutral. So liberals might argue that the real aim of such a measure is to promote a particular conception of the good life (i.e. heterosexuality). But such a characterisation of Devlin’s argument would be misleading. Devlin did not justify his argument on the grounds that heterosexuality is the good life or that homosexuality is merely offensive. His argument was that the decriminalisation of such behaviour would be harmful to society. His argument could thus be framed in terms more congenial with the neutrality constraint. Neutralist liberals themselves recognise that the state should not be neutral between all conceptions of the good, but only between permissible conceptions of the good. Devlin would argue that the behaviour targeted by such legislation is impermissible because it threatens the social cohesion necessary to maintain stability. By prohibiting homosexual behaviour the state is taking the steps necessary to establish and secure the minimal standards of behaviour necessary for society to be stable.
Framed in these terms, Devlin’s argument is perhaps more formidable to the neutralist liberal than one might have initially expected. But one does not have to establish that a prohibition on homosexuality violates the neutrality constraint in order to establish a persuasive case against such a provision. Emphasising this point is important for it highlights the other demands of public reason. Namely, that not only must the objectives of legislation be neutral, but the means chosen must be reasonable and demonstrably justified. Prohibitions on homosexuality fail these requirements. In order to substantiate that the means chosen in this instance are reasonable and demonstrably justified Devlin would have to support a number of claims he failed to substantiate. Firstly, he would have to provide evidence to show that decriminalising homosexuality would cause the harm he claims it would cause. Devlin’s failure to do this is more than sufficient grounds for rejecting the call for a prohibition on homosexuality. Secondly, Devlin would also have to show that less restrictive measures (e.g. education) are insufficient for establishing and securing the minimal standards of behaviour necessary for society to be stable. And thirdly, even if Devlin could substantiate his claim that such behaviour harms society, such a policy would only be reasonable if the benefits of the prohibition clearly outweighed its costs (this is the requirement of proportionality). Given the gross violation of individual liberty such a prohibition entails, the harm at issue would have to be very substantial. Devlin’s failure to establish these points reveals the different ways his proposal fails to satisfy the demands of public reason. The requirements of public reason go well beyond the requirements of the neutrality constraint.
The neutral state could pursue a number of oppressive measures in the name of promoting ‘neutral aims’. It could enforce, for example, overly stringent traffic laws or unreasonable building regulations, etc. Such laws or policies are legitimate, the neutralist liberal might argue, because they are premised on values no one could reasonably reject (e.g. public safety). But does the fact that they are neutral necessarily mean that these policies are publicly justified? There is nothing in the ideal of neutrality itself that ensures that measures be reasonable as well as premised on neutral values. The state could justify a number of repressive measures that are consistent with neutrality of aim. For example, the government could justify a prohibition on motorised vehicles on grounds of public safety. In this case we have a neutral aim but an unreasonable policy.
Two features of such a law are worth emphasising to reveal how, despite its being neutral, it is unreasonable. Firstly, while we recognise that the aim of public safety will require limitations on our freedom we also expect legislators to take a responsible approach to pursuing such an aim. Comparable levels of public safety can be secured by varying degrees of restrictive policies and we expect law-makers to opt for those measures that impair our freedom as little as possible. Opting for the most extreme form of restriction in the name of public safety is unjustified if a less restricted policy, say one that permitted motorised vehicles but imposed a number of various traffic laws, would secure a comparable level of public safety. Furthermore, even if a prohibition on motorised vehicles secured a substantially higher degree of public safety such a proposal would still qualify as unreasonable and unfair. Public reason demands that there be a proportionality between the effects of the measure and the objective in question (i.e. public safety). In the case of prohibiting motorised vehicles there is no such proportionality. The gains secured in public safety do not outweigh the burdens imposed on freedom and efficiency. People prefer to live with the higher degree of risk which comes with permitting motorised vehicles because it also brings with it benefits which outweigh these risks (e.g. greater mobility).
The coercive power of the state is not necessarily legitimate when it is consistent with the requirements of the neutrality constraint. In addition to pursuing neutral aims, the government must consider a diverse range of concerns which the neutrality constraint itself is silent about. The threat of oppression in liberal democracies often comes, not from perfectionism, by from an overzealous pursuit of neutral aims (e.g. public safety). In order to ensure that we pursue a reasonable balance between conflicting fundamental values, like liberty and safety, we need a public ethic that goes much further than simply prescribing justificatory neutrality. I say more about this topic in my paper "Neutrality, Toleration and Reasonable Agreement".