While the 2021 census is the best part of three years away, it was making news last week, thanks to former Sinn Féin chair and Assembly speaker Mitchel McLaughlin. McLaughlin used his inaugural lecture as an honorary professor at Queen’s University to call for the inclusion of a question on constitutional preference in the next population count.

The census and the six counties

The relationship between the census and the constitutional status of Northern Ireland has a long history. A question on religion was first included in the British census of Ireland in 1861 – some 140 years before it was asked in England, Wales and Scotland – and data from this question was used to decide the boundaries of Northern Ireland when it was created on partition in 1921. The 1911 census had shown that the population of nine-county Ulster was only 56 per cent Protestant, and so a decision was taken to ‘sacrifice’ those living in Cavan, Donegal and Monaghan, in order to secure what was assumed to be a much safer unionist majority of roughly two Protestants to every Catholic in the remaining six counties.

While the collection of census data has gone relatively smoothly in Northern Ireland compared with other divided societies, the period of the Troubles saw calls from republicans for boycotts of the census. In 1971, faced with such a call, 9.4 per cent of respondents declined to answer the question on religion – up from only 2 per cent in 1961 – and in 1981, against the backdrop of the hunger strikes, this figure reached 18.5 per cent. 1981 also witnessed the murder of census worker Joanne Mathers by the IRA in Derry.

Partly as a result of the lack of complete data on religion from 1971 and 1981, intense political and media speculation accompanied the 1991 census. Many thought that it might demonstrate that the days of the ‘safe’ Protestant majority were over and that a Catholic majority was imminent – the implication being that a united Ireland was not far around the corner. Although the results failed to bear this out, the same speculation was repeated in 2001, only to be followed again by surprise that the expected Catholic majority had not materialised, confounding the hopes of nationalist politicians and the fears of unionists.

Sectarian headcounting

While it is true that the Catholic share of the population has steadily increased with each decennial census, there are complexities that are lost in this narrative. In 2001, NISRA, which conducts the census in Northern Ireland, responded to the growing proportion of people who were not specifying a religion by introducing a supplementary, ‘religion brought up in’ question. Where respondents fail to provide an answer to either question, NISRA ‘imputes’ answers based on the individual’s other characteristics. This provoked some criticism at the time, with Alliance Party leader David Ford arguing that it was ‘a sad example of how public agencies can inadvertently promote sectarianism’. However, because community background data from the census is used as a benchmark against which to compare employers’ fair employment monitoring returns, the practice has largely been accepted as necessary to combat discrimination – although many understandably recoil at being assigned a background against their choice.

When census results are featured in the press, the distinction between religion and religion brought up in is often lost. The Belfast Telegraph’s ‘rugby-score’ headline of ‘PROTESTANTS 53 per cent, CATHOLICS 44 per cent’, which accompanied the release of the 2001 results, and the reporting of the 2011 results as 48 per cent Protestant and 45 per cent Catholic, present data that includes those who have been imputed a religious background – much to the annoyance of some who do not identify with either ‘community’.

Another innovation introduced in the 2011 census has served to problematise the ‘two communities’ narrative, however. The 2011 questionnaire was the first to ask a question on national identity, after NISRA opted to follow the lead of the Office for National Statistics, which introduced this question in England and Wales. The results of this question revealed a complex picture, whereby no single national identity formed a majority, a significant minority identified as ‘Northern Irish’ rather than British or Irish, and many opted for more than one identity. These results arguably had the effect of undermining simplified assumptions about the relationship between religion, identity and constitutional preference that had dominated debates about previous censuses.

Brexit, the border and the 2021 census

There is a certain honesty of intent behind Mitchel McLaughlin’s suggestion to ask about attitudes to a united Ireland in the 2021 census; it would at least address the question head-on, which might be preferable to poorly informed speculation about the constitutional implications of results on religious background. However, NISRA are unlikely to agree to include such a question.

As the then Registrar General told McLaughlin at a Stormont briefing session on preparations for the last census, ‘There is a predilection for censuses to avoid attitudinal questions and to record matters of fact’. One might argue that questions about national identity (or sexual orientation, which is under consideration for 2021) are in part attitudinal, but such an obviously political question as constitutional preference is likely a step too far for any census. Moreover, there are significant technical barriers: consultations on the content of the next census are already well advanced, and NISRA prefer to keep the format of the questionnaire as similar as possible to that used in England and Wales, in order to simplify the processing of the data. There would also be questions about the reliability of the answers where one person completes the form for an entire household.

In the aftermath of the UK’s vote to leave the European Union, which has prompted renewed discussion about the possibility of a border poll, attention is turning once again to Northern Ireland’s changing demographics. The 2021 census is therefore likely to be the subject of significant debate. Perhaps inevitably, coverage will focus on the ‘sectarian headcount’, but we would do well to look beyond the headlines and reflect on the full spectrum of identities, in all their complexity, that the census reveals.