Some objections to the desire-based objection

In his stimulating paper, Bouke de Vries proposes a novel ‘desire-based objection’ to the constitutional recognition of an official religion. In developing this objection, De Vries’s stated target is a particular—and still largely theoretical—form of religious establishment, one that some political theorists believe could be compatible with liberal principles (see Laborde, 2017 for one argument to this effect). This is when recognition is entirely, or at least mostly, symbolic, as when clergy participate in formal state occasions or when religious symbols are displayed in public institutions. According to some, these symbolic manifestations of establishment might be compatible with liberalism, provided they threaten neither the freedom nor equality of all citizens, including members of minority faith communities as well as agnostics and atheists. ¹

The desire-based objection is not an all-things considered objection, either to establishment in general or even to this particular variety. Indeed, De Vries acknowledges that, in any particular case, it might be outweighed by countervailing considerations, such as concerns of social stability. Rather than criticising any actually-existing model of church–state relations, arguably De Vries’s real aim in this paper is to lay the groundwork for an original justification of multifaith or plural symbolic establishment, in which more than one religion is granted official recognition. This becomes clear towards the end of the paper, where he describes a diachronic form of multifaith symbolic establishment where ‘state religions are rotated’. Importantly, for De Vries, the case against mono-recognition is bound up with the theoretical possibility of diachronic plural recognition, since he believes it has the virtue of ‘treating citizens more fairly’ and thus ‘tends to be morally superior’ to mono-recognition.

What, then, does the ‘desire-based objection’ consist in? As I understand it, it says that it is pro tanto wrong to deny constitutional recognition to a religion if its members desire for it to be recognised (or would so desire, if they did not fear a ‘social backlash’) and provided that their religion is ‘entitled’ to recognition. To my mind at least, De Vries makes two controversial moves in developing this objection, one concerning when a religion is entitled to recognition independently of the preferences of its members, and the other concerning the relevance of those preferences. I shall discuss each in turn.

A religion is entitled to recognition, according to De Vries, provided it has a claim to recognition that is on a par with some other religion that is already recognised. De Vries identifies two main grounds for making this kind of claim: either having enough adherents, or the members having made sufficient ‘societal contributions’. ² Though he says relatively little about how we are to measure these ‘societal contributions’, or about why they matter, he discusses demographic considerations in greater detail, in particular whilst setting out his preferred diachronic model. Here, De Vries seems to assume, albeit without explicitly saying so, that each individual has an equal claim for recognition, meaning that, ceteris paribus, a per capita distribution of religious recognition would be fair. Thus, the greater the number of adherents, the more recognition a religion deserves (Patten (2014) makes a similar proposal regarding the official recognition of languages).

Since he thinks there is a limit to how many religions can be recognised, De Vries proposes a minimum threshold of 0.27% of the population, which would guarantee at least one day’s recognition per year for those who qualify. The rationale he gives for this is that having one’s religion recognised ‘for a mere hour, let alone a minute, would clearly be too short’. Personally, I am not persuaded that this is true, since I cannot see why having one day a year would be enough but one day every two years or even every decade would not be. In any case, notice that De Vries’s threshold is a function of his diachronic model, and not of multifaith establishment as such since, when recognition is mostly symbolic, there are no limits to how many religions can be recognised simultaneously.

In making recognition into a scarce resource, De Vries’s diachronic model undercuts his thesis that plural recognition is fairer than, and hence morally superior to, mono-recognition. This is because one consequence of having a threshold is that some smaller communities of faith will not qualify for recognition. For instance, in Ireland today, Buddhists, Jews, Jehovah’s Witnesses and Methodists would fall short of De Vries’s threshold, and even Hindus would only recently have become entitled to a single day. ³ Adherents of these smaller religions surely have good reasons to prefer mono-recognition over plural recognition, if their own faiths will not qualify for recognition under either scheme. Furthermore, they will surely have even stronger reasons to prefer disestablishment to anything else. This can be illustrated by considering the constitutional reforms that took place in Ireland during the 1970s. Until 1972, Ireland granted constitutional recognition to six religions, including Catholicism, whose ‘special position’ was affirmed, as well as Anglicanism, Presbyterianism, Methodism, Quakerism and Judaism. For De Vries, what would have made this arrangement objectionable was not the inequality between Catholicism and the rest, since his diachronic model explicitly allows for unequal recognition. Rather, it was that, by his standards, three of those religions (Methodism, Quakerism and Judaism) had too few members to merit recognition in the first place (Central Statistics Office (CSO), 1977). Surely, however, it was fairer to remove recognition from all six religions, as the Fifth Amendment effectively did, than to remove it only from three, as De Vries would have advised.

In any case, suppose we have a satisfactory list of religions that qualify for recognition. Even then, De Vries does not insist that each of them should be recognised, only that it would be unfair to pick one of them out for special treatment, if the members of one or more other religions also desire recognition. It is puzzling to me, however, why the preference for recognition should be given such priority, whilst the preferences that one’s own religion, someone else’s religion, or all religions not be granted recognition are discounted. People have very different preferences about the constitutional recognition of religion. Some want no religion to be recognised, some want all religions to be recognised but on identical terms (something De Vries’s diachronic proposal does not satisfy), and others want for their religion alone to be recognised, perhaps even preferring that no religion be recognised to recognition being shared. According to De Vries, some of these preferences can be discounted because they are ‘unreasonable’ and do not ‘carry moral weight’, and to illustrate this he gives the example of someone who ‘simply do[es] not want members of a specific (liberal-democracy-compatible) religion to be constitutionally recognised’. His targets here are people with preferences about the recognition of another religion, of which they are not a member. Perhaps he is right to say that these preferences should not count (though why?) but his formulation also excludes someone’s preferences about the recognition of their own religion, if they would prefer for it not to be recognised. It is not that De Vries thinks that everyone wants their own religion to be symbolically recognised, since he acknowledges that there is ‘nothing obviously irrational’ about someone who does not. The problem is that he does not consider people who prefer for their own religion to not be constitutionally recognised, as opposed to those who are merely indifferent about the possibility. Someone might have such a preference, for example, because of other beliefs they have about church–state relations, or because they are sceptical about the legitimacy or morality of the state they live in and would rather not have their religion tainted by association with it.

So, then, the only preference that seems to count for De Vries is for having one’s own religion recognised if other religions are also granted recognition. Presumably, the reason why this preference is special is that satisfying it seems to harm no one else. After all, adding Islam, Catholicism and Hinduism to the list of religions recognised in the United Kingdom would not deprive Anglicans of recognition, only require them to share it. By contrast, satisfying the preferences that one’s own religion be the only one recognised, or that no religions at all be recognised, both seem to harm others, since either outcome would deny recognition to people who desire it. However, this line of reasoning seems to be incompatible with De Vries’s claim that the desire-based objection does not assume that recognition is ‘objectively or inherently valuable’. For it seems to rely on the thesis that it is worse to not have recognition when one desires it than to have recognition when one does not. Furthermore, it is perhaps untrue that satisfying the preference for one’s own religion to be symbolically recognised harms no one else. For example, Laborde (2020) has suggested that establishment in the UK is objectionable because it aggravates the social vulnerability of sexual minorities, conveying a negative message that undermines their standing as civic equals, since the Church of England refuses to officiate at same-sex marriages (222, see also 2017: 135–136). In this case, satisfying the preference of (some) Anglicans for symbolic recognition does harm others, and this harm would surely be aggravated by shifting towards De Vries’s preferred model of multi-faith establishment.

A final problem for De Vries’s desire-based objection, I believe, is that it can be invoked in support of plural material establishment. Consider a society in which an established religion is supported with public monies, perhaps to help with the upkeep of its buildings, to pay the salaries of clergy or to fund its schools. Fellow citizens who belong to a minority faith community might reasonably desire for the same benefits to be extended to them, feeling that this would acknowledge ‘the societal contributions of their religion’s members’ and ‘help them to feel equal citizens’. It does not seem implausible that mere symbolic recognition would do neither of these things, even if the original established religion also had its status downgraded, so that it, too, was now only recognised symbolically. This is because members of minority faiths might feel deprived of the kind of support that the majority has long benefitted from. So in this scenario, no one would prefer to move towards plural symbolic recognition, and at least some would prefer to move towards plural material recognition. As a consequence, De Vries must surely conclude that plural material recognition is superior to plural symbolic recognition, and for the same reason that he thinks plural symbolic recognition is generally superior to mono-symbolic recognition (i.e. it satisfies reasonable preferences).

Perhaps De Vries might respond by saying that his argument is not intended to provide positive guidance about the best model of church–state relations. Instead, the desire-based objection applies only to the symbolic and constitutional recognition of a single religion. In which case, however, one might wonder whether the objection can tell us anything about actually existing societies. After all, the kinds of preferences that people have about recognition depend on how church–state relations are configured right now, and not on how they would be configured if only symbolic forms of recognition were permitted.