Wednesday 10 December 2008

The 2007 Annual Report of the Registrar General

Today NISRA published the 86th Annual Report of the Registrar General, which provides a wealth of information on the demographic make-up of Northern Ireland in 2007.

Not all of the information is entirely new. Information on Births in 2007 was published in March 2008; information on Deaths in 2007 was published in April 2008, and information on Marriages, Divorces and Civil Partnerships was published in June 2008. The Annual Report is, therefore, more of a compilation than a new source of information, though it does present the information in more depth and with more commentary. It is well worth a read.

Nonetheless, it is useful to use today's publication to reiterate a number of points.

Firstly, concerning births; while the religion of the parents of the babies is not given, it is possible to estimate the community of many of them from the District Council area in which the birth was recorded. Thus, for example, we can assume that most births in Derry or Newry and Mourne are to Catholic parents, while most of those in Carrickfergus or North Down are to Protestant parents. What this, admittedly rough, estimate shows us is that the birth rate remains higher in Catholic areas than in Protestant areas.

Appendix 2 of the Annual Report shows, for example, that the birth rates in majority Catholic areas are generally higher than those in majority Protestant areas. The table below shows the birth rate in the 26 District Council areas sorted according to the proportion of their 'child-bearing cohort' (i.e. all those aged 20 to 39) that is Catholic. It should be noted, however, that the figures used for the religious breakdown of the district populations are those of the 2001 census, while the birth rates are those of 2007. So while we are not exactly comparing apples and oranges, we are perhaps comparing Golden Delicious with Cox’s Orange. The loss of accuracy is probably not great, however.

Eleven of the districts have a Catholic majority in the age group 20-39, and in seven of these the birth rate exceeds the Northern Irish average (13.9 live births per thousand in the population). Of the 12 areas with a majority-Protestant child-bearing cohort, only 3 had birth rates that exceeded the average. All three were in outer-Belfast commuter areas, where young families are likely to set up home. Two of the three areas with religiously balanced child-bearing cohorts had birth rates that exceeded the average:

The average birth rate in the 11 Catholic-majority areas in the table above is 14.5, while the average of the 12 Protestant-majority areas is 13.1. The three balanced areas (including, of course, the biggest by far, Belfast) had an average birth rate of 14.4. If the figure for the Catholic-majority can be used as a proxy for the 'Catholic birth rate', then it is likely that many of the births in these three balanced areas were actually Catholic births.

Looked at in this way, though, the figures are misleading. The birth rate is the number of live births per 1000 in the whole population. Thus, it compares births in areas with a high proportion of old people with areas with a high population of young adults. This is not a valid comparison, as it ‘rewards’ areas like Dungannon for having relatively few old people compared with North Down, for example.

If the figures for live births are shown as a proportion of the child-bearing population, aged 20-39, a completely different picture emerges. Dungannon, with its record birth rate of 16.1 per thousand of the population drops to the third lowest – a birth rate of 37.0 per thousand of those aged 20-39. The same table as shown above, now looks like this:

The three areas with the lowest number of births per thousand aged 20-39 are all nationalist: Magherafelt, Cookstown and Dungannon. It seems that Mid Ulster is dying out! However, of the eleven majority-nationalist areas, seven have birth rates above the average according to this method of calculation. Of the 12 Protestant-majority areas, six are above average – but of the eight most-Protestant areas only two are above average, while fully 5 of the eight most-Catholic areas are above average.

Secondly, the proportion of marriages that take place in Catholic churches remains stable, while the proportions that take place in Protestant churches is tumbling. The proportion of religious marriages that are in Catholic churches has been over 50% for several years now, and appears to be increasing. Of course the real winner in the marriage statistics is the registry office, and we know nothing much about the religions of the people concerned. [As a small titbit of information, in 1887 in the six counties, the proportion of marriages celebrated in Protestant churches was 69.5%. It is now 33.2%.]

Deaths, unlike births, happen to people we already know. So there is little doubt about who is dying. We know the proportions of most ages according to their religions, and we know the age specific mortality rates. Putting the two together, we can make a fairly good estimate of the community breakdown of the deaths. The average age of death is 75 (71 for men, 79 for women); so if we assume that the religious proportion of the deaths is that of the 75 year-olds in the population (known from the religious proportions at age 69 in the 2001 census), we can estimate that 65% of the deaths (9522 people) were Protestant, and 35% (5127) were Catholic.

(When the words 'Catholic' and 'Protestant' are used above they refer to people who are of Catholic or Protestant community origin. The people themselves may have no active religious belief or interest, and they may not vote according to the norms for their communities, but most evidence points to an enduring link with the religious, cultural and political identities of the group into which people are born.)

2007, therefore, saw no great change to the patterns that we have seen over the past generation. Catholics continue to have a slightly higher birth rate, so although they are lightly less numerous than Protestants in the age group 20-39 the number of Catholic children being born continues to exceed that of Protestant children. Two factors will act on the birth statistics over the next few years ­– the Catholic fertility rate may drop to close to that of Protestants, but the proportion of the child-bearing cohort that is Catholic will increase, and start to form the majority. So these two factors will cancel each other out, leaving, in all likelihood, a continuing Catholic majority amongst the babies. The deaths are mainly amongst Protestants, and will continue to be for at least another generation. So the net gain to Protestants will be lower that that to Catholics, leading to a continuing ‘greening’ of Northern Ireland’s population.


Mack said...

That's an interesting analysis (births per thousand in the cohort). The average age for a Mother is now 30.

I was wondering about mid-Ulster. If you look at the graph you did (link below) there is an almighty spike in the under 19's (20% leap).
mid ulster age distribution 2001

These would be around 24-27 now, which I imagine is one of the lowest 'producing' age groups, but yet they're adding greatly to the numbers in the cohort.

Other factors to consider would be age distribution factors esp. teenage pregancies (probably lower in rural areas). I'd imagine that the bulk of 'older' mothers are distributed in around the late 20's and early 30's, but that there is also probably a clump of teenage mothers followed by a 'barren' stretch of late- teens & early 20's with a gradual reduction in numbers throughout the 30's. (But this is pure speculation).

Horseman said...


That is a good point. The spike has probably not yet started to have children in significant numbers. They should start to be visible in a year or two though. I wonder will we see a relatively rapid increase in the birth rate soon?

Mack said...

Mid-Ulster has also benefited from high levels of immigration (>8% of total population in 6 years), the immigrants tend to heavily concentrated within the child bearing cohort (both stats from the report - whether or not they'll impact your figures depends on whether your using 2001 census figures or 2007 estimates). It's possible (likely - given low current rates) in mid-Ulster that they also haven't settled down and started having babies yet.

So either there'll be a baby boom there in a couple of years or there's something else entirely going on. Time to research opening a pram shop in south Derry...


Also take a look at figure 1.20 in the report (page 18) - they do something similar but define the child bearing cohort as 15-44. Produces markedly different results (mid-Ulster shoots to the top of the pile - so it may well be that spike that pushes your numbers down).

Horseman said...

I didn't take immigration into account, since it is currently impossible to know the precise numbers of immigrants in the population. The figures I used were from the 2001 census (were the immigrants there then? I don't think so, at least not in such numbers), so the immigrants would have boosted the 2007 numbers in the child-bearing cohort beyond what was recorded in 2001, but seem to have had very little effect on the birth rate. Maybe they are not having kids in NI, but are waiting till they go home. In any case, since I used the 2001 census figues, the effect of the immigrants would be to push down the age-specific birth rate even further.

The reason I took 20-39 as the child-bearing cohort is because, basically, it is. There are very few births outside those ages (see figure 1.18 on page 17 of the report).

I'd hold off on that pram shop for a little while, but if there isn't an increase in demand in a few years then we'll be looking for some new and unexplained phenomenon.