Governments have always liked to count heads. The enumeration of people not only describes power relations but helps to create them. The account in the Gospel of Luke of the census called by Caesar Augustus is engrained in our culture as part of the nativity story. Whatever the historical accuracy of Luke’s account it is the case that the Roman empire regularly enumerated the Israelites, partly in order to raise taxes and partly to maintain control over them. Today the state of Israel controls its own census, and the enumeration of its people now carries an existential charge. For the first time the Arab population living between the Mediterranean Sea and the Jordan River has risen slightly above the Jewish population, and that presents a dilemma for Israel. Writing in The Washington Post (23/07/22) the president of the World Jewish Congress, Ronald Lauder, put it this way: “If [Israel] gives Palestinians full citizenship ‑ and therefore full rights ‑ it will no longer be Jewish. If it doesn’t do so, it will no longer be democratic. Either way, Israel, as a Jewish democratic state, will cease to exist.”
The counting of numbers has been crucial to the state-building project in Northern Ireland. The border that was drawn around six Irish counties was decided upon, after detailed scrutiny of the 1911 census, in order to provide the greatest amount of security to unionists in the northeast. A census conducted on the centenary of the state was bound to focus attention on the numbers. Inevitably, within Northern Ireland this was going to be an “usuns and themuns” exercise. There is an argument that the census process not only reports on communal distinctions but helps to reproduce them. In England, for example, Protestants and Catholics are not census categories; instead, they are batched together as Christians. It is in all probability the case that if that were to happen in Northern Ireland it would increase the sense of shared identity, but the realpolitik is that any attempt to allocate public resources in a divided society requires a proper understanding of societal patterns.
Besides, the ten-yearly census exercises a particular fascination for the people of Northern Ireland. The publication of the figures for the new ratios of Catholics to Protestants are often compared to the football results: indeed, when the 2001 figures were published in the Belfast Telegraph the front page of the newspaper resembled the back page in the way the block type announced the tally: Protestants 53%, Catholics 43%. Actually, to compare them to the football results is to understate their impact. They are like the FA Cup final, the World Cup and the Six Nations all rolled into one.
This year’s results delivered their share of sensation. The SDLP leader, Colum Eastwood, used the word “seminal”. Professor Colin Harvey from the Ireland’s Future group chose “seismic”. Sinn Féin MP John Finucane said the results showed that “historic change is happening”. They were reacting to the headline results, which showed that Protestant numbers were lower than Catholic numbers, and that those who self-identified as British Only were down to 31.9%, or less than a third of the population. The irony of this reversal was supreme. The jubilation of nationalist politicians was in exact proportion to the dismay of unionist politicians. The leader of the hardline Traditional Unionist Voice, Jim Allister, was able to give the bitterness of that disappointment a voice: “The fact that the Catholic population in Northern Ireland has flourished presents a telling contrast with the near extinction of the Protestant population in the Republic.”
Many of the press releases that followed publication of the results were written before the figures were known. Once attention moved beyond the headlines it became apparent that this was no 10-0 win. It was all so much more complex. The big winners were not unionism or nationalism, but those who wished to slip free of those moorings altogether, the people sometimes referred to as the neither/nors or, using the language of the Belfast Agreement, the Others. It was as if, towards the end of the match, another team came on the pitch and started kicking the ball around. And with three teams now at play no one knows how this game will end.
If there were no clear winners in these results, there was one clear loser. Unionism is on an ice floe that is melting underneath it. The figures are pitiless. The census form asks two questions about religion. The first is to do with your current religion, the second asks what religion you were brought up in. Thus, someone brought up as a Catholic would record themselves as being brought up in that community, but they may have become an atheist and so in the “current religion” category would self-designate as someone of “no religion”. The two identities are recorded separately. On both registers Protestant numbers are in decline. Ten years ago, 41.6% of the population identified as being in one or other Protestant denomination (there is of course no such religion as Protestant: it is the generic term for the aggregation of denominations). The new census puts the total at only 37.4%. And even that provides too optimistic an assessment if one is gauging the health of traditional Ulster Protestantism. The full title of the category is Protestant and Other Christians, and the latter phrase denotes an extremely heterogenous collection, made up of Mormons, Quakers, Evangelicals and Romanian Orthodox – not the sort of people you would expect to encounter on an Orange procession. Between them they account for almost 7%, leaving the total for traditional Protestantism at 31%.
When you turn to national identity the situation is, if anything, even more stark. In 2011 the total who identified as British Only was 40%. All the trends suggested that this was not yet the bottom, that British identity would dip further. Talking with friends about it before the results were published, we agreed that a fall to 35% would be experienced by unionists as a bad result. A drop to 33%, or one-third of the population, would be a disaster. On the day, the figure came in at 31.9%, a notch below disaster. Some encouragement could be taken from another table. In the spirit of the Good Friday Agreement, it is possible for people in Northern Ireland to identify as having more than one national identity. Thus, one could be British and Irish or Irish and Northern Irish or even Irish, British and Northern Irish. When taken with a dash of another identity, the British identity, or rather, the British Plus identity rises to 42.8%. Cold comfort really, as it’s still almost six percentage points below its total ten years ago, when it claimed a 48.4% share.
What has happened? Where did all the Protestants go? Over the past ten years the overall population of Northern Ireland has increased by 5%. In that same period the Protestant population has dropped by 40,000, or 2%. Some of this is due to simple demographics. More Protestants are dying than are being born: the Northern Ireland Statistics and Research Agency estimates the natural change (the difference between births and deaths) as a net 30,000 loss in the past ten years. The ballast for the Protestant demographic has always been in the upper age bands ‑ but time is the enemy. The younger cohorts with their Catholic majorities are moving up the escalator, and the movement is unstoppable. For the last ten years the annual School Census has been showing Catholic school enrolments at just over a half of the total, with Protestant enrolments at one-third. There is only one way this can go.
There are two other forces that are combining to break up the concrete of the old unionist hegemony. One is emigration. More and more Protestants are leaving the country to go to other parts of the United Kingdom. These are often people who take their British identity seriously, in the sense that they feel at home on “the mainland”. In the case of younger people, particularly those who are choosing university places, they are trying to put them-selves at a distance from a unionist identity which they see as toxic and choosing to leave it behind for a more socially liberal environment.
The census figures also reveal another type of migration, the statistical migration of people from one category to another – in this case, from a British identity to a Northern Ireland identity. The motivation for this is the opposite of that just cited: it is the desire of other Protestants, usually older, to distance themselves from what they see as a liberalising English culture. At the height of the flags protest in 2013 the then first minister, Peter Robinson, rounded on the protesters who were demanding a return to direct rule from Westminster: “Are they content to have Westminster impose same-sex marriages and abortion on demand on our community? Such folly.” His warnings were prescient. The subsequent suspension of Stormont allowed for Westminster legislation introducing abortion and same sex marriages in Northern Ireland, in 2019 and 2020 respectively. It was, if you like, the imposition of British values in this corner of the United Kingdom.
Another small, but telling, skirmish took place in the week leading up to the release of the census results. The Boys Brigade in Northern Ireland broke away from the UK Boys Brigade, citing “doctrinal differences”, widely believed to be related to LGBT issues. For historical reasons Protestants who feel themselves to be at a cultural distance from modern multicultural England have found comfort in a kith-and-kin relationship with Scotland, but a particularly wounding split opened up in recent years between the Presbyterian Church in Ireland and its mother church, the Presbyterian Church of Scotland. Once again, the issue was sexuality. In 2018 the Presbyterian Church in Ireland decided with regret to break off relations with the Scottish church over the issue of same sex marriages. These retreats into an “ourselves alone” bunker can find expression in a choice to opt for a Northern Ireland identity, or a British and NI identity, a category favoured by 8%. It is one which allows an alignment with UK citizenship while maintaining a distance from modern British mores. As it happened, the person most admired and cherished by Northern unionists, Queen Elizabeth, died on September 8th and her funeral caused the publication of the census to be delayed. When it was released, it was reporting on an era that had just passed.
Those who claim a British identity are not, of course, all Protestants. There are many social, political, cultural and economic reasons why Catholics might want to stay within a UK framework; this is, after all, the deal that they signed up to with the Good Friday Agreement. But for Catholics who choose a pro-union stance the relationship with unionism is strained, and the rise of English nationalism has strained it further. The post-Brexit chill factor is undoubtedly why we see a boost in the Irish Only identity, up from 25% ten years ago to 29%. But, as we shall see, that figure is not in itself enough to provide comfort for nationalism.
There were expectations that this census would deliver the results that would leave nationalism a knight’s move away from a border poll. It hasn’t quite worked out that way. The percentage choosing an exclusively Irish identity is 29.1%, still trailing the exclusively British figure of 31.9%. When the additional identities are added (British plus and Irish plus) the lead widens further. The British identity extends to include 42.8%, while the Irish only stretches to 33.3 % ‑ exactly one-third of the population. An optimistic nationalist might point to the fact that both these figures are four or five percentage points in advance of the 2001 referendum, and it is undoubtedly the case that they show incremental gains for those who support national unity. Incremental, but not transformative. The hope had been that Brexit would be shown to represent a rupture of a much more significant kind. At this point I should confess that I had expected something of that order. After all, Brexit has been the greatest earthquake in British politics in a generation, with tremors have felt in every part of these islands, and indeed across all of Europe.
Where do those tremors show in the census data? The only statistics that show a step change beyond what demographics might deliver are those for Irish passports. UK passports now only account for 53%, as against 60% in the last census, while the Irish share has moved up more than twelve percentage points, from 21% to 32.3%. That’s still a long way short of the figure that would compel a British secretary of state to call a referendum on Irish unity. The wording in the Good Friday Agreement is that he (sic) will call a border poll “[i]f it appears likely that a majority of voters in Northern Ireland would back a united Ireland”.
Of course, the census figures for national identity should not be taken to be definitive on their own. Other indicators would include election results, opinion polls, and the figures for religious affiliation. A quick sweep through them is enough to reinforce the point that religious identities, constitutional preferences and electoral behaviour should not be confused. There is no direct readout from one to the other. It is more than likely, in fact it is a certainty, that among the third of the population who feel themselves to be Irish there are those whose constitutional preference is to remain inside the UK. Ever since the Good Friday Agreement the combined nationalist vote in elections has hovered around 40%, almost as if it is floating free of the growth of the Catholic population or the numbers of those who claim an Irish identity. Demography is not destiny. The most common mistake of the demographic determinists in the run-up to this census was to assume that if Catholics were to become the largest population group a border poll would follow along with the same predictability as night follows day. The picture now emerging is much more complex, and also much more interesting.
The big story of this census is that the two big, frozen identity blocs are melting, and we are seeing fluidity in the middle. The exclusive identities, those who identify only as either British or Irish, make up just 59% of the population. The rest is made up of people who don’t have that attachment to an anchor identity, people whose identities are more complex and more contingent. This was of course the promise of the Good Friday Agreement, that people could escape their birth community label and choose “to identify as Irish or British or both, and to continue to hold both Irish and British citizenship if they so choose”. The great conceptual leap was to allow people to move from either/or to both/and. Northern Ireland citizens can hold both British and Irish passports, and the census form was even more accommodating. You could go beyond dual identity and tick the boxes for Irish, British and Northern Irish (full disclosure: I picked all three, which put me in the same category as 1.5% of my fellow citizens).
At the buffet table of national identity there was a smorgasbord of choice: English, Welsh, Scottish, Irish, British, and Northern Irish were all on offer, waiting to be mixed and mingled. Oddly, the one that was initially touted as the illustrative example of dual identity at the time of the Good Friday Agreement, the British and Irish combination, got one of the lowest take-ups: only 0.6% chose to cojoin the parent identities. The others were a jumble of hybrids which in the round accounted for 19.4% of the total population. The most popular partner for link-ups was the Northern Ireland identity, which was chosen to link with another identity by 11.7% of the population. On its own as a single identity, it accounted for 19.8%; taken together with a hyphenated partner as NI Plus it accounted for 31.5%, just less than two points behind the Irish Plus identity.
What does that mean in terms of political change? In numerical terms perhaps not a lot. The 19.8% as a single identity is in fact marginally lower than the 2011 total, of 20.9%. The last census was the first time the NI identity had been offered as a category, and much was made of it representing a new, moderate constituency. Surveys showed that those who identified with this label tended to be more inclusive, more tolerant of the other tradition. A political party with the name NI21 was launched in 2013 and had two MLAs elected to the NI Assembly. It turned out to be a rocket that only briefly lit up the night sky. Within two years it had burnt and crashed, a casualty of internal disputes. If the hope of a new political movement had been based on the census figures it was mistaken. While it was the first time NI had been offered as a category option in the census, opinion polls for decades – notably, the NI Life and Times survey – had been tracking a swathe of opinion of similar size that didn’t self-categorise as unionist or nationalist but ticked boxes for Ulster, Anglo-Irish and other such labels. That “none of the abov”’ constituency has existed for decades, but its political colours have always been hard to pin down.
As a shape-shifting phenomenon it naturally attracts attention. The fact that its total numbers have hardly changed in the past ten years may suggest that it has become a static category. Underneath the new total though it looks like there is a lot of churn. Leakage from the British identity is making its way into the Northern Ireland identity. At the same time, it appears that the boost in Irish identity is due, at least in part, to Catholics in the post-Brexit era moving from a Northern Ireland identity to a more uncompromising Irish Only identity. In short, Protestants are moving into this category as Catholics are moving out. So what looks on the surface like a stable identity may in fact be a work in progress. As stated earlier, it has become a form of ethnic affirmation for unionists who no longer feel cultural affinity with permissive English values, but it still offers shelter for soft unionists and soft nationalists. The only generalisation that holds true for such a diverse category is that you cannot generalise about them. It will take further studies for the cultural cross currents to be tracked.
If we can’t know how these people might vote in any future border poll, we know even less about the new communities. The figure that most suggests the scale of transformation in Northern Ireland is that for the number of people who were born outside its borders: 13.5%. Half of these, 124,000, were born outside the UK and Ireland, and 57,000 outside the EU. The percentage of non-white people is 3.4%, a small number compared to Dublin, London or Manchester but almost twice the size of the minority ethnic community ten years ago. That 3.4% figure, incidentally, does not include most Eastern Europeans, who are largely classified as ethnically white. A different measure of the size of the newcomer community can be found in the number who have passports that are not either British or Irish, and that total is 5.5%.
Taken in the round, the 39% of the population who do not fall within the traditional categories of unionist and nationalist are, in theory, the people who might hold the balance in a future border poll. If a border poll is to be decided by fifty per cent plus one, then as a demographer once put it to me, that one could well be a Chinese grandmother in Finaghy. For people trying to guess how a border poll might play out the problem is that the floating mass of hybrids and new arrivals are not tethered to either national identity pole. Moreover, they very often belong to the part of the electorate that does not bother to vote. This distancing from electoral politics becomes more pronounced as you move down the age cohorts: research by Professor Katy Hayward at Queen’s University shows that more than half (52.6%) of those aged 18 to 29 did not vote in the 2019 UK general election. More socially liberal than their parents, they fight shy of party politics and when ticking national identity boxes are more likely to go for hybrid identities or to pick Northern Ireland in the same uncertain way that one might pick the mid-point in a Likert scale.
There are then two great ironies to be observed in the 2021 census. The first is that a state deliberately carved out to ensure that Catholics would be a permanent minority has ended up in its centenary year with Catholics as the largest population group. The second irony is that if there is a border poll it will be those who care least about national identities who will make the momentous decision on the future for both nationalism and unionism.
Exclusive attention on Catholics, Protestants and others can tend to obscure some of the social developments that will also have long-term importance. Secularisation is one of the big stories of this census. The number with no current religion is a whopping 17.4%. This is in line with international trends. The 2021 Census in Australia shows 40% with no religion. The latest British Social Attitudes Survey, conducted in April 2021 gives the share with no religion as 53%. In the Irish census of 2016 one in ten say they have no religion, making the “no religion” category the second largest category after Catholic (78%). Quite what it means to be a Catholic country following referendums which have voted in huge nu-bers for divorce, same sex marriage and abortion is a question for another day, but it does suggest that the old identity categories are leaky.
Finally, it should be remembered that the census is primarily the framework for public planning. It will help when we plan hospitals, cope with pandemics and make provision for schools and care homes. In this regard the headline story has to be the growth in the overall population, up to 1,903,175, the highest total ever for the area now known as Northern Ireland. By far the biggest increase has been in the population aged sixty-five and over: from 101,800 in 1926 to 326,500 today. This is very definitely an ageing society. Indeed, it is expected that within the next ten years there will be more people aged sixty-five and over than children under fourteen. Long term, this is a dependency crisis, and in two ways. Firstly, the old age dependency ratio (the population aged sixty-five+, compared with the population aged fifteen to sixty-four) has increased to 27%, meaning that there are fewer prime age workers to support a growing population of older persons. Who pays? The tax yield will be too low to support this population profile, so the question has to be asked in a different way. Who pays the subsidy, Britain or Ireland? Whichever government takes responsibility would be wise to look at these census figures in detail.
Dr Paul Nolan is an independent researcher based in Belfast. He writes on conflict societies and the Northern Ireland peace process.