And back to the facts… a reply to Professor O’Leary

Colin Coulter and Peter Shirlow respond to four key points raised by Professor Brendan O’Leary in his challenge to their original paper published here on Slugger last week, along with clarification of some of the language used in both pieces. 

In his response to our recent opinion piece, much of what Professor O’Leary has to say focuses not on that essay but another publication that covers some, but not all, of the same ground. We felt it might be considered discourteous not to respond. O’Leary advances four critiques of our article, which we propose to deal with in turn.

The first criticism offered by O’Leary centres on our discussion of the possible connections between the acceleration in the growth of those identifying as Catholic in the last Census and the marked fall in the number of people not stating their religious affiliation. That thread in our paper is characterised both as ‘speculative’ and indicative of a ‘conspiratorial cast of mind’. Let’s deal with the first of those adjectives now, and return to the other later on.

That specific argument we make in the article is, of course, speculative (although based, as we shall see, on solid evidence). Why should that be a problem? Social scientists are, by definition, interested in current and future patterns of change and that often requires speculation, based at times on evidence more limited than we might wish. Accompanied by the appropriate qualifications, speculative inquiry is a thoroughly legitimate, indeed commonplace, element of academic practice.

And O’Leary himself, it is worth remembering, is no stranger to speculation. Take, for instance, his influential text Making Sense of a United Ireland. That was published literally days before the initial batch of Census data was released in September 2022 and the final draft would, presumably, have been completed only a few months earlier. Given that issues of demography are essential to the central thesis of the book, it might well have been prudent to have waited for the publication of the most recent Census figures. But that isn’t what happened.

On the opening page of the book, O’Leary assures readers that ‘we do not need to wait’ for the up-to-date demographic information as ‘the brute facts of the census of 2011 predict the future quite well.’ That is quite a presumptuous, not to mention speculative, way in which to approach such a pivotal source of information. And it helps to explain why it is that the claims that O’Leary has made about demographic trends over recent years often seem out of date.

Now for the substance of the argument. In our paper, we note that between 2011 and 2021 there was a sharp drop – from 122,252 to 30,529 – in the number of people refusing to answer the first Census question on religion. That is a remarkable development, but one rarely remarked upon. So we decided to speculate on what happened to the 90,000 or so people who, in the space of a decade, lost their reticence about filling in that part of the Census.

The most reasonable premise from which to start was that in the past most of the ‘not stateds’ had been from Catholic backgrounds. The high level of people not completing the first question on religion originated, after all, in the Census boycotts organised by the republican movement in 1971 and 1981. Given that backstory, it was reasonable to assume that most of the ‘not stateds’ were residents of working class nationalist communities. Where else would they have been from?

That presumption turned out to be correct. Given that it is impossible to ascertain the religion of those who refuse to disclose it, we used the proxy variable of spatial segregation. And what we found was that people living in neighbourhoods that are predominantly Catholic are more likely to fall into the ‘not stated’ category than those in living in predominantly Protestant neighbourhoods – see quintile 5 in the table below.

While those who refused to state their religion in 2011 were primarily Catholic, that was no longer the case a decade later. What emerges from a closer examination of the 2021 data is that almost all of those who refused to answer the first question on religion did in fact complete the second. Below we present data that was not available to us when we submitted the article under discussion:

The figures above clearly show that while Catholics may have represented a majority in the ‘not stateds’ in 2011, they were only a minority of that group in 2021. It follows then that most of the 90,000 people who changed their minds about completing the first question in the Census are most likely to be Catholics.

As we acknowledge in our paper, it is impossible to make authoritative claims about where these individuals may have moved within the Census. But one plausible suggestion would be that a meaningful number of them decided, for the first time, to tick the box that identifies them as Catholic. That would explain why it is that the rate of growth in the number of people identifying as Catholics actually accelerated between the last two Censuses in spite of the fact that in that decade the previous demographic advantages of the nationalist community went into reverse (falling birth rates, a steep decline in immigration from the Baltic states, excess deaths in poor communities, which are still predominantly Catholic, during Covid etc.).

So that is the evidence that we have to hand and the interpretation that we have chosen to place upon it. In his comments, O’Leary dismisses our arguments without any evidence to support his case. What he provides instead are some distinctly opaque comments about the ‘degree of coordination’ required for the changes we envisage to have actually happened. That’s a really curious reading of human agency.

We work on the assumption that men and women make their own history and make up their own minds. People are more than capable of reading how the wind is blowing and changing their own conduct accordingly. That doesn’t necessarily require ‘co-ordination’ and anyone who believes it does truly has a ‘conspiratorial cast of mind.’

In his second criticism of our article, O’Leary takes issue with our depiction of the rapidly expanding body of people who specify in their Census forms that they have no religion. O’Leary appears to suggest that we have claimed that ‘all or a very high proportion’ of the ‘nones’ hail from Protestant backgrounds.

We make no such claim in our article. What we do argue, however, is that those who have no religious affiliation are simply more likely to have a Protestant cultural lineage than a Catholic one. And we can back that claim up quite comfortably with two forms of empirical evidence.

The first of these derives from the supplementary Census question on religion. If we consider those who have no religion but were raised in one – see the graphic below – we see that such individuals are almost twice as likely to have grown up in Protestant (102,764) households as Catholic ones (52,320).

After the second question on religion, there are 177,360 people who insist they simply have no religion, regardless of the number of times they are asked (the number includes the 3,662 who failed to answer the first question but nonetheless responded ‘none’ to the second). Given that those individuals have made it amply clear they have no religious affiliation, we need to find an alternative means of ascertaining their cultural orientation. The most appropriate alternative variable is that of national identity. The Census data on the nationality of the ‘nones’ is set out below.

The two tables that follow provide us with our second body of evidence that those who have no religion in Northern Ireland come mainly from unionist backgrounds. We can see in the first table that the ‘nones’ primarily see themselves as ‘British’ (47%) – an identity associated overwhelmingly with the unionist tradition – and ‘Northern Irish’ (38%) – an identity that Protestants are, now at least, more likely to subscribe to than Catholics. In contrast, a mere 12% of Northern Ireland’s atheists and agnostics see themselves as ‘Irish’ – the identity most closely associated with the nationalist tradition.

The respondents that feature in the table above contain a relatively large number of people from England, Scotland, and Wales and that might be said to be skewing the numbers. So let’s remove everyone born outside Northern Ireland and look solely at the region’s homegrown atheists and agnostics.

As before, a clear pattern emerges here. The evidence from the Census reveals that the ‘nones’ predominantly choose nationalities associated with the unionist tradition. And that means that the only reasonable response to the inquiry O’Leary makes in his post – ‘aren’t there high numbers of cultural Catholics among the Nones?’ – is an emphatic ‘no’.

This inability to appreciate the nature of the ‘nones’ is echoed in how he approaches another constituency that exists beyond the traditional nationalist/unionist binary. On other occasions, O’Leary has suggested that the ‘don’t knows’ are ‘persuadable’ (see 09.12 here) and he echoes that assertion here in his claim here that there is a ‘feasible’ scenario in which they might ‘line up in favour of unification’.

While certainly possible, that is not the most likely development. Faced with the prospect of serious political upheaval, those who do not have strong convictions either way may well plump for the relative safety of the constitutional status quo. That common-sense interpretation is confirmed in the literature on referenda held in other countries. For instance, in their deliberations, the Working group on unification referendums on the island of Ireland noted international research indicating that during referendum campaigns popular opinion tends, on average, to swing 3.9% in favour of the status quo. O’Leary would be familiar with those findings, of course, having been a member of the panel that collated them.

In his third critique of our paper, O’Leary takes issue with the absence of any mention of gaelscoileanna in our analysis. The relevance of that particular comment isn’t readily apparent. Our paper was concerned with the Census, not with schools or languages. It is difficult to see, then, how we might reasonably have been expected to have discussed Irish-medium education.

The fourth and final criticism that O’Leary advances returns to the subject of Northern Ireland’s schools. O’Leary concedes that ‘Catholics might not be a majority in any one specific age group’ before going on to cite 2021-22 data showing that Catholic pupils represent more than half of school enrolments.

That is a significant, strategic shift in focus, one frequently made by commentators sympathetic to a united Ireland when faced with data from the Census inconvenient to their worldview (see several of the comments under our original opinion piece).

It signals an implicit acknowledgment that some of the demographic claims circulating in the public domain of late are erroneous. And that, in turn, begs the following question: if Catholics aren’t a majority in any Census cohort, then why would anyone make the claim that they are?

We might add also that the data presented by O’Leary are already out of date. The most recent information, for 2022-23, indicate that the proportion of Catholic pupils in Northern Irish schools has now fallen below 50% (see the data on ‘pupil religion’ here). In view of the declining birth rate within the Catholic community – down 22% over the last Census period – that is the likely direction of travel for the foreseeable future.

So, that’s O’Leary’s four critiques dealt with. Now a few words about words. At the start of his response, O’Leary takes issue with how we described his contribution to the recent Ireland’s Future gathering. This is a simple misunderstanding, on his part. What we said was that he had ‘read the script’ at the event, not that he had ‘read from a script’ provided by someone else.

The phrase to ‘read the script’ is a metaphor akin to ‘read the room’, suggesting that someone has a facility to recognise what is politic in a specific context. And that figure of speech seemed appropriate given that O’Leary’s comments clearly chimed with the opinions held by most of those gathered in the SSE Arena.

O’Leary also takes issue with the word ‘millenarian’ which he assumes, incorrectly, was used in relation to him. It is crystal clear in our opinion piece that it was, in fact, directed at the target date mentioned in the most recent document issued by Ireland’s Future. While the term may well have religious origins, it can also be used in a secular sense.

Most socialists are secular, and many are atheists, but that does not mean they cannot be millenarians. Those in the revolutionary tradition are precisely that. So the term ‘millenarian’ is used here in its secular sense. It’s a little baffling that it would be taken as otherwise.

Finally, we really should draw attention to one of the more striking assertions made by O’Leary in the course of his intervention. Responding to criticism of the absence of any mention of the ARINS survey data in his presentation to Ireland’s Future, he offers this explanation:

I don’t believe I can say much that is social scientifically credible about public opinion in 2030 based on current surveys, and nor can anyone else.

That is quite a remarkable claim. Demographic trends are, needless to say, crucial to the discussion of constitutional change. But ethnoreligious status is no absolute guarantor of constitutional inclination. Electoral patterns are also critical to that debate. But party political preference has no definitive connection to constitutional choice either.

Surveys of opinion have their shortcomings, needless to say, as well. But their very particular value to debates on constitutional change is they deal specifically in the variables of constitutional change, not merely in its proxies. In plain terms, such polls ask people directly about where they might wish to live in the future, not just who they plan to vote for or which foot they happen to kick with. And that has a genuine value, all the conventional caveats notwithstanding.

What makes O’Leary’s claim all the more remarkable is that it challenges the very rationale of every survey about (Northern) Ireland’s future. If an opinion poll really cannot, as O’Leary seems to suggest, provide accurate insights into how the balance of constitutional forces might lie in just six years’ time, then why bother with them at all? And where, more specifically, does that leave the surveys in which he personally is involved?

Over the last couple of years, O’Leary has given multiple presentations about what the data from the ARINS project might have to tell us about the potential future(s) of this island. Are we simply to conclude now that all those fascinating numbers have, ultimately, really very little reliable to tell us about what lies ahead?

Discover more from Slugger O'Toole

Subscribe to get the latest posts sent to your email.

We are reader supported. Donate to keep Slugger lit!

For over 20 years, Slugger has been an independent place for debate and new ideas. We have published over 40,000 posts and over one and a half million comments on the site. Each month we have over 70,000 readers. All this we have accomplished with only volunteers we have never had any paid staff.

Slugger does not receive any funding, and we respect our readers, so we will never run intrusive ads or sponsored posts. Instead, we are reader-supported. Help us keep Slugger independent by becoming a friend of Slugger. While we run a tight ship and no one gets paid to write, we need money to help us cover our costs.

If you like what we do, we are asking you to consider giving a monthly donation of any amount, or you can give a one-off donation. Any amount is appreciated.