The 1998 agreement paradox
The signing of the 1998 Agreement has been identified as a seminal moment, marking the end of a conflict that had come to define Northern Ireland in terms of its conflicting communities: Catholic/Irish/Nationalist and Protestant/British/Unionist, respectively. As such, the narrative of two fundamentally distinct communities has become central to efforts at establishing peace (Taylor 2009). The all-Ireland cross-border institutions, together with those 'linking the two sovereign governments of the United Kingdom and Ireland in confederal ways' (McGarry & O'Leary 2009) represent the interconnection of communal, national and political identities across these islands. The 1998 Agreement assumes these ethno-national identities (Irish Nationalist and British Unionist) to be of greater significance than all other identities in Northern Ireland. Thus, it is implied, Unionism and Nationalism are more than political ideologies: they are political traditions, with cultural roots and social identities.
Whereas one would expect contemporary political ideologies to evolve and respond to a changing context, this is not the case in post-Agreement Northern Ireland. Tying Unionism and Nationalism to cultural identities gives them the status of 'communities', more than ideologies or political preferences. This in turn allows them to be considered as fundamentally unchangeable. Indeed, much has been made, in the implementation of the Agreement, of the need to uphold and respect the integrity of 'both communities' and 'two traditions' (The Agreement 1998, emphasis added).
Importantly, the Agreement does not assume the creation of an 'alternative' identity nor of a diverse or 'mixed' middle ground. This can be seen in the provision for a border poll, which was so important for gaining the support of Nationalist parties for the Agreement (Hayes & McAllister 2001; Mac Ginty 2003). By having the possibility of a referendum on Northern Ireland's constitutional status written into the Agreement, the presumption appears to be that there will be at some point in the future a majority in favour of Irish unity. Such an assessment assumes that the growth of the Catholic population (relative to those in the diverse Christian denominations grouped under the heading 'Protestant') would equate to a corresponding growth in the portion of population seeking Irish unity. However, there are two problems with this assumption. The first is that to be Catholic is not automatically to self-identify as a Nationalist, and it does not inevitably entail holding ambitions towards Irish unity. The second is that it disregards the fact that so many within Northern Ireland do not identify with either Unionism or Nationalism. In fact, at least 4 out of every 10 people in Northern Ireland tend to describe themselves as Neither Unionist nor Nationalist.
It is this group that provides the principal focus for the discussion that follows. The research we present here centres on data from the Northern Ireland Life and Times Survey (NILT), which has been conducted annually (with the exception of 2011) by ARK (Queen's University Belfast and Ulster University) since 1998, with funding from the Economic and Social Research Council. Most fundamentally, we seek to identify whether the decision to identify as belonging to Neither of these communities is, overall, a 'positive' sign of confidence in the post-conflict dispensation or a 'negative' sign of disengagement from it.
Do the 'two communities' really exist?
The reason that the Agreement assumes the existence of a binary divide and places it at the heart of the 'settlement' is that there is plenty of historical and sociological evidence to show that opinions among respondents who are Catholic and those who are Protestant diverge significantly on political matters relating to the Irish border and national sovereignty (Todd 2010). The status of respondents as 'Catholic' or 'Protestant' can often be a more significant variable in contemporary political or sociological analysis than age, gender or class (Evans & Tonge 2013; Hayes & McAllister 2015): opinion on Brexit is a good illustration of this (Garry 2016). Different traditions do exist in Northern Ireland, but is it accurate to claim they are contained within only two blocs?
A quite straightforward means of challenging the 'two communities' thesis is to look at the relative size of these 'blocs'. Data from the 2017 NILT (1) show that, by bringing together a diversity of Christian denominations under the heading 'Protestant', the size of the Catholic and Protestant communities in Northern Ireland is almost equal. This is close to the findings of the 2011 census, which had 41% Catholic, 42% Protestant and 16% No Religion/not stated (Northern Ireland Statistics and Research Agency (NISRA) 2012). What it also shows--to contest the 'two communities' thesis--is that almost a fifth of respondents identify as being neither Catholic nor Protestant (19%).
A similar impression is gained from a different question on political identity in Northern Ireland--one that is often assumed to replicate the Catholic/Protestant distinction --in that we see here too a third category of strong significance. Almost half of respondents identify as being neither Unionist nor Nationalist (45% in the NILT survey of 2017). The scale of this response demonstrates that the 'No Religion' category does not, on its own, explain the existence of the Neither population, but rather that some Catholics and Protestants also self-identify as Neither. This is not a new phenomenon. Data from NILT show a trend of fairly steady growth of the proportion of the population claiming to be neither Unionist nor Nationalist, although there has been some fluctuation in this since a peak of 48% in 2012. It is now quite apparent that the largest proportion of the population do not identify with either of the two predominant labels in Northern Ireland politics. Indeed, the growth of this group over time would suggest that it could be worth considering a new distinction in Northern Ireland between those who identify as Unionist and Nationalist, and those who identify as Neither (Figure 1).
This could be the manifestation of a phenomenon identified by Baka et al. (2012), namely that, rather than reflecting pre-existing 'neutrality', the design of the question itself could be producing ambiguity. In some ways, the offering of an option to be neither Unionist nor Nationalist is like offering a middle point in a Likert Scale. The decision to pick that middle point could be used to convey a lack of knowledge or a degree of uncertainty. But it could also reflect internal dilemmas on the part of the respondent (e.g. the perception of negative associations with the other options) or it could mark a decisive rejection of the assumptions behind the question itself (for instance, that there is a binary divide; Baka et al. 2012). By looking in more detail at who is ascribing themselves to this Neither category, we hope to offer some insights into whether those self-reporting as Neither can be classified as having some group features or whether it is merely an articulation of dissatisfaction, protest or rejection. We would suggest that if there are consistent patterns in those designating as 'Neither' over the course of the 20-year period, then it is more likely we can talk about a 'group', if not a community, and test this for its political significance.
Do the Neithers (2) share common identity traits?
In this section, we wish to interrogate the question of who is a 'typical' 'Neither'. In this we will look at the usual variables of sex, age and religion, but we will also note some points on which there are consistently significant differences between those who are Neither and those who are Nationalist and Unionist.
NILT data show that women have always been more likely than men to identify as neither Unionist nor Nationalist. The growth in support for Neither has occurred across both men and women (a rise of 10% among men and 13% among women) since 1998, but gender has become a more significant variable over the two decades since the Agreement: 61% of those identifying as Neither in 2017 are female. This is notable because sex is not a significant variable in support for Unionism or Nationalism (around 53% of those identifying in each category are male; NILT 2017 ). We know that nationalism itself is highly gendered and that nationalist movements are dominated by masculine interests and ideologies (Nagel 1998; Yuval-Davis 1997), and yet this does not in itself suggest a passive role for women in nationalist movements. There is much research to show that women have often been written-out of their active role in Nationalist and Unionist mobilisation and conflict--and peace building--in Northern Ireland (Gilmartin 2015; McDowell 2008). For women to constitute such a large portion of those identifying as neither Unionist nor Nationalist suggests that the rejection of the Unionist/Nationalist typology may also be associated with a rejection of the type of macho, patriarchal politics that is still quite predominant in Northern Ireland (Ashe 2007; Stapleton & Wilson 2014).
Non-religious are not predominant among those identifying as neither Unionist nor Nationalist
Because of the association between Unionism/Nationalism and Protestantism/ Catholicism in Northern Ireland, it is tempting to assume that those who don't identify with either political ideology are likely to be non-religious. In which case, we would expect to see the rise of the Neither category being in parallel with a rise in the category of those saying they have No Religion. By comparing the data on this over the 20 years that followed the 1998 Agreement, we see that the level of support for Neither (from 33% in 1998 to 41% in 2017, with highs of 46%-47%) has grown at about the...