Thursday 26 February 2009

Schools Census 2008-2009

Today the Department of Education released the detailed results of the Schools Census for this academic year (2008-2009). For this blog the particular interest lies in the tables dealing with the religions of pupils, insofar as it provides a preview of the community breakdown of the future Northern Ireland.

The results reinforce the pattern that is visible elsewhere – the proportion of the pupils in Northern Ireland's schools who are Protestant continues to decline, while the proportion who are Catholic remains over 50% and continues to rise.

The proportion of schoolchildren in Northern Ireland who declare themselves to be Protestant or 'other Christian' (a category including the following denominations: Jehovah's Witness, Church of Jesus Christ of Latter Day saints, or Other Christian) now stands at 40.7%.

The proportion of schoolchildren in Northern Ireland who declare themselves to be Catholic stands at 50.9%.

The graph below shows the historical evolution of these raw Catholic/Protestant proportions taken from the Schools Censuses from the last 10 years:

The remaining 8.4% are made up of non-Christians (0.4%) or 'Other/no religion/not known' (8.0%). In the northern Irish context the non-Christians are largely outside the political debate, and are primarily Chinese, Muslims and Hindus. The 'Other/no religion/not known' are more interesting because they account for one pupil in twelve and their actual community backgrounds will make a considerable difference to the shape of Northern Ireland in the future.

Some people believe that atheists or agnostics are more likely to come from a Protestant community background, on the basis that Protestantism has traditionally been more liberal and that the proportion of people who state 'no religion' both in the Schools Census and the decennial census is higher in majority Protestant areas. Another possibility, though, is that Catholics in Protestant areas (or schools) may be more likely to not want to draw attention to themselves, and thus fail to declare themselves as Catholics. The argument based upon the relative liberalism of Protestantism may have had some merit in the past, but is less likely to be the case currently as the strength of all the churches on their flocks is increasingly weak. It is likely that a significant part of the 'Other/no religion/not known' group is actually from a Catholic community background.

The two graphs below show two possible interpretations of the 'Other/no religion/not known' proportion. In the first, a strictly even breakdown is made (though this, of course, slightly benefits the Protestant score, as it is only four-fifths of the Catholic score in the 'raw' figures. The second graph makes some concession towards those who believe that most of the 'Other/no religion/not known' are from a Protestant community background, and allocated them as 70% Protestant, and only 30% Catholic.

In neither case does the allocation of the 'Other/no religion/not known' significantly alter the inescapable outcome. The proportion of the schoolchildren who are Catholic, or from a Catholic community background, remains considerably higher that that who are Protestant or from a Protestant community background.

And the gap is widening over time.

When you consider that the majority of every year's Schools Census was included in the previous year's exercise – only the previous year's leavers and this year's P1 class are different – the changes visible represent only the net differences in two years out of the complete 14 year cycle. In a nutshell, if the proportion of Protestants has dropped, this means that far fewer of the P1 intake are Protestant than was the case of the previous year's leavers.

The significance of these figures is enormous. It shows that, barring asymmetric migration, in the future Protestants will be a minority of Northern Ireland's population, and Catholics will be a majority. The figures reinforce the decline in the Protestant proportion of the population as a whole seen in the decennial census, and the linked decline in the unionist vote seen in the elections over the last generation. If the two communities continue to vote along religious/community lines, it means that the unionist political project is doomed.

Even if, as some would argue, the 8% 'Other/no religion/not known' represents a centrist non-sectarian group, by definition it cannot come to the aid of the unionist cause. If its adult counterpart votes for the Alliance Party, the Greens, or various independent candidates, then they form part of neither community block, and will take a neutral position on the constitutional future of Northern Ireland. If unionism must rely on the votes of Protestants, as it currently does, its future looks increasingly bleak; if it tries to break out of its community ghetto (as the UUP increasingly claims to be trying to do) then it will have to make far more serious efforts than it is currently making. If nationalism retains its support from the Catholic population it will succeed in outvoting unionism within a generation.


Anonymous said...

Fascinating stuff as always - and the result of hard work. I'm not sure I agree on a few points

(1) Even if, as some would argue, the 8% 'Other/no religion/not known' represents a centrist non-sectarian group, by definition it cannot come to the aid of the unionist cause.

Well, they can indeed - by voting to stay in the Union in a Border poll. Remember that this, and not a Nationalist majority in Stormont, is the mechanism by which this will be sorted.

(2) In the northern Irish context the non-Christians are largely outside the political debate, and are primarily Chinese, Muslims and Hindus.

Again, they may choose the 'no change' option in a referendum.

Be careful with the 'Catholic kids' and 'Protestant kids' labels. Richard Dawkins might picket your house.

All the best and keep up the good work.

Horseman said...

Hi Anonymous,

Thanks for the comments - I agree with them to a great extent; I didn't want to turn the blog into a thesis so I cut some corners, obviously.

When I meant that the 'centrist non-sectarians' wouldn't come to the unionists aid, I meant that their parties would probably not espouse a unionist position (the largest, Alliance, has repeatedly said that it will be neutral in a border poll and will respect the outcome). Obviously their voters would vote, though, and this is why it is interesting to think about their underlying community identifications.

The positions of the non-Christians may be very status quo, as you say, but they are very few in number and are largely politically uninterested.

Yeah, I was a bit too short-hand in describing the kids - but I get tired of saying 'kids of Protestant parents', etc. I hope most readers know what I meant. To be honest, it is also a bit inaccurate to describe most people in NI by religion, since very few of them display even basic Christianity. But if I used other words to describe the two tribes someone would get upset!

Anonymous said...

Sammy Mc Nally Said

Horseman are these figues released anually? Based on these figures the trend has been there for some time - but I dont remember much talk this time last year.

The only thing that may have impacted which is a fairly recent development is our good friends the polish plumbers and their female companions - they may of course not be Nationalist.

Horseman said...

Hi Sammy,

Yes, this is an annual thing. I've been following the figures for years, though nowadays DENI give you the whole series every time (as you'll see if you download the Excel tables). I really don't think that the trend is due to the Poles or Lithuanians - there just aren't enough of them (their kids, I mean). It is simply another example of the general tipping of the demographic balance. There are more Catholics than Protestants at all ages up to 30 or so, and Catholic parents are still having more kids. It all provides momentum to the demographic shift - unless Protestants start having more kids than Catholics (unlikely) the Protestant majority will soon be over. And not a day too soon!

Anonymous said...


what puzzles me is the lack of attention to these figures particulalry as they have been in the public domain year on year. I have seen a debate on Slugger where someone was suggesting that there were more Unionist background (I am uncomforatable with the religious tag) - and the figures were not quoted.

Also have you seen on your Euro elction thread I have put the link to the Euro turnout by Westminster consituency - which would appear to show much lower turnouts in Unionist areas?

Anonymous said...


Regarding last year's census has the protestant percentage not increase in the current school census.

Keep up the good work.

Horseman said...

Hi anonymous (26 February 2009 23:30),

Denial has been very strong in unionist circles for years - the results of the 2001 decennial census were interpreted to 'show' that the proportion of Catholic kids had dropped below 50% (Garret Fitzgerald was particularly gleeful about this). The problem is, it was a mistaken interpretation, as the schools census shows! The fact that a large number of people did not put down a religion for their kids did not mean that they were not going to raise them in a religion. For example, of the babies below one year old in the census, a full 26% had no 'declared' religion. Yet, eight years later, when all those babies are now in school, the proportion with no declared religion is around 8.6%. So there were clearly other reasons why people failed to declare a religion for their kids. And since only 42% of the babies were declared as Catholic in 2001, while 51% of today's primary school children are declared as Catholic, the census figures cannot be relied on. Those, like foolish Fitzgerald, who greeted the census with such glee, will have to eat some humble pie.

I saw your table on the EP turnouts per constituency - thanks for that. It tends to show that unionists were much less motivated to vote (in 2004 at least). I wonder will the TUV factor motivate them this time? Maybe it will just turn them off even more? Especially since none of the parties appear to have any actual interest in the EP as such, or in its role or importance.

Horseman said...

To anonymous (27 February 2009 01:53),

No, the Protestant percentage has decreased vis-a-vis last year. In 2007-2008 it was 41.3%, and this year it fell to 40.7%. I include the 'other Christians' in the Protestant numbers, as they are likely to comprise either evangelicals (Mormons, etc), or are likely to be ex-mainstream Protestants. I know this is an assumption that not all might agree with, but I feel that on balance it is correct.

Anonymous said...

What's with all this 'kids' stuff?
I thought I was in Ireland not the U.S.A. lol.

Anonymous said...

What's with all this 'kids' stuff?
I thought I was in Ireland not the U.S.A. lol.

blogmenai said...


Do these figures indicate that the census office was incorrect to distribute the 'no religions' to the Protestant side of things by a factor of 7:4 when they were determining so called 'community background'?

Faha said...

The school census contains some other interesting data. There are 7,000 students who speak a language other than English. However, I could find no breakdown of the languages. Many would be of East European background but also Chinese, Indian, Muslim and African. It is unclear what perecentage are Catholic, but it would definitely be higher than 50% The None group of 8% is higher than for the same age groups in the census. I would estimate that 5% are None and the other 3% did not answer the question. If the former would eventually vote 60% unionist and the latter are the same proportion as all those in the school census, then that 8% would break down to approximately 4.5% unionist and 3.5%nationallist. Eventually, the whole school age cohort would be roughly 55% nationalist and 45% unionist when they reach voting age

Anonymous said...

V. Good as always

Horseman said...


To be honest, I do think that they were wrong in 2001. I do not know the precise methodology that they used, but the outcome, as seen in the schools censs, shows that they seriously underestimated the proportion that were Catholic. I think that they often assumed that people who declared 'no religion' in a Protestant area were generally Protestant, whereas they may have been disproportionately Catholic - people keeping their heads down.

Horseman said...


The non-English speakers are, s you say, a small mystery. But whether they will still be in NI in a few years is doubtful. If the economy crashes they may move elsewere. In any case, they are no more likely to adopt a unionist identity than a nationalist one. It would probably depend on where they live and who their friends are.

I agree more or less with your breakdown of the 8%, but even if it is 5:3 or 6:2 it provides little comfort for unionism.

Anonymous said...


The confusion i noticed was due to the including of the other christians in the total protestant pool.I have a couple of queries regarding the census. Is it fair to include the nursery schools as the number going are not representative of the number of children born each year. about 25000per year from memory.

Are the regional colleges included the old techs? Just on what other bloggers have said regarding foerign national children. I would suggest that in the coming years you will see a steep rise in the catholic share. My reasoning being that 10percent of births are now to foerign mothers who tend to be coming from catholic countries. In dungannon i believe it is 18 percent. It is a pity the department does'nt incluge religion tables by years ex. P1,P2 etc.

Thanks again for the good work.

Horseman said...

Hi Anonymous (at 03 March 13:00),

I exclude the nursery schools, nursery classes and recepotion from my figures, as school is ot compulsory at that age and thus the figures are not reliable. I include only the years of compulsory schooling (which include, of course those few years at the end when a few may have left).

I am not certain about the regional colleges - if they are counted as secondary schools then they're included. I have to assume DENI don't exclude them.

You're right about the 'foreign mother' thing, but I guess we'll have to wait until the next census to see the impact of them - if they're still here by then.

Faha said...

I would like to add a few more insights into the 2001 census. I believe that the Catholic percent of the populatiion was 1% higher (44.75%) than the actual number in the census ( 43.75%). There were 3 factors that contributed to the 2001 census underestimate.
#1 British military members and their dependents were included in the census as well as foreign students ( see OA birth file ). When these are excluded the Catholic percent increases by 0.3%.
#2 There was an undercount in the rural wards in the west and south. If you look at the voter registration rates in 2003 ( after the stricter registration requirements were in force) there are many wards with registration rates of 105% to 110%. These are mainly wards with Catholic majorities. This accounts for another estimated 0.3 to 0.4 % undrestimate.
#3 In the 2001 census 3.5% of the population did not answer the religion question. Another 0.6% checked off that they had a religion but did not identify it and another 0.6% checked None but did not indicate if they were raised in a religion. The census office assigned a religion to these 4.7% based on 3 factors:
#1 The religion of others in the household.
#2 Age
#3 Knowledge of the Irish language
I suspect that most were assigned based on knowledge of the Irish language since 22% of Catholics had some knowledge of Irish and only 1% of Protestants did so this would be a strong correlating factor for religion. The underestimate of Catholics occurred because knowledge of Irish is least common among Catholics in areas such as North Down, Carrickfergus, Castlereagh which are also the areas that had the largest percentages of those who did not answer the religion question. Thus, if only 11% of the
Catholics in North Down had knowledge of Irish ( as opposed to the 22% number that was used to make the calculation) then this would lead to an underestimate of Catholics in North Down. Of course there would be an overestimate in areas where Catholics have a higher than 22% percent knowledge if Irish but these areas had a much lower percentage of those who did not answer the religion quesstion. This factor accounts for approximately another 0.3% to 0.4%. The census actually assigned 40% of the 4.7% non responders to the Catholic population. My revised estimate would have increased it to 48%, which seems realistic.

Anonymous said...

163950/323264x100=50.7% not 50.9% or am I missing something? That's would be an increase from 50.6% in 2001.
Still a majority but not much change since 2001. Do the Unionists who want an end to power sharing know about these figures? Turkeys voting for Christmas?

Horseman said...

Hi Anonymous (08 March 2009 18:15),

I excluded the kids younger than compulsory school age, i.e. those in nursery classes and the like. I don't know whether there might be a bias in those figures, but to be certain I left them out, and took only those ages where schooling is compulsory and thus includes everyone. I realise there may be some early school leavers so the figures for secondary schools don't include every kid in the age group, but there is no way to know how many there are, or what their community background is.

Anonymous said...

Do you see the Unionist one seat majority in Stormont going at the next election and if so where will they lose the seat? When do you think Unionists will be a minority in Stormont?

Anonymous said...

this is fenian propaganda use should all be ashamed!

Arnold said...

I'm not so sure if you can project such a trend too far into the future given the massive changes that have happened in NI in recent years.

For one thing, there was virtually no immigration to NI until just a few years ago. Certainly not on any significant scale which is indicated by the low non-English speakers in the schools. Clearly that has changed a lot and very quickly too. When I went to school there was one black child, now it's unusual to see a school that doesn't have an ethnic mix.

Also worth noting is that during the worst of the troubles the number of unionists leaving for GB would have been significantly higher than republicans leaving. Now that peace is, hopefully, here to stay many of those are returning and skewing the figures somewhat. Given the time that they would have left I don't think that they would be returning with primary school age children thus they'd skew your projections even further.

As regards the non-Christians, my guess is that they would largely consider themselves British-Moslems etc and would therefore be likely to vote to stay in the UK should a referendum situation arise at some point. I would have thought that this would also apply to other immigrant groups too. Thus labeling someone from, say, Manchester as catholic really says nothing about how they might vote and that they are British-catholic is more significant ie one could assume that they would vote for the status quo. Don't forget either that those coming to NI from GB would not normally be entitled to Irish citizenship so would be more likely to vote for the status quo if only for that reason. Having said that, these immigrants would probably not count as "Northern Ireland people" under the definition in the Good Friday agreement so might not be entitled to vote in any referendum on the border issue.

One good side-effect of this is that maybe we are on the way to being able to separate the religion from the politics as that binding together is very unlikely to apply to immigrants.

Pedro said...

According to NISRA stats the country (for want of a better term) that contributes most to immigration to NI is none other than the ROI. Most of the rest are from Catholics countries like Poland and Latvia who, when the become integrated, will tend to absorb the general politocultural ethos of indigenous NI Catholics - at 2nd generation at least..