Counting Catholics

Andrew Brown of the Guardian has just written a very interesting article on the problems of calculating the size of the Catholic community. The church defines this as those who have been baptised living within England and Wales (or in Scotland; these are separate church jurisdictions). A sociology colleague has mentioned however that church administrators do refer to “unbaptised Catholics” – theologically heterodox, but perhaps wise to the need to plan for school places and similar.

The main sources are the Catholic Church itself; the Government (specifically the Scottish and Northern Irish censuses, and data on marriages); and opinion polls and social surveys. The Church data for the most part depends on parishes estimating their community size, predominantly from mass attendance through an annual count, together with estimation of the size of the community which does not attend. This might be gauged from counting rites of passage (baptisms, first communions, confirmations, marriages and funerals), which are also enumerated for parish returns and sent to diocesan offices for compilation and annual report, in diocesan directories and the Catholic Directory for England and Wales.

Tony Spencer, Director of the Pastoral Research Centre, has published an assessment of data quality of various Catholic sources – diocesan directories, the Catholic Directory for England and Wales, and the Vatican-compiled Annuario Pontificio, which is difficult to access. This is discussed in some detail in A. E. C. W. Spencer (ed.), Digest of Statistics of the Catholic Community of England & Wales, 1958-2005: Volume I. A common problem is lack of clarity over how missing returns are treated.

Regarding the Catholic Directory, which is the most commonly-cited source, Spencer suggests that one issue is that data are often not supplied by some dioceses before the publication deadline. This means that the newest data are provisional but often not labelled as such, nor revised between editions. Spencer also suggests that operational definitions may vary between dioceses.

We provide here the best set of data at our disposal for England and Wales over the past decades up to 2005. The sources used have been Peter Brierley (ed.) Religious Trends 2 and Religious Trends 7 (1999, 2008); R. Currie et al. Churches and Churchgoers (Oxford, 1977); and Spencer’s Digest (2008). We also include a graph from the merged set of British Social Attitudes surveys, from 1983 to 2008. (Predicted figures for 2010 are available in Religious Trends 7, and reported data on attendance, baptisms and so on are available after 2005 from recent editions of the Catholic Directory; the latter will be added here in due course.)

For now here is a set of charts: click on the images for a high-quality version.

1. Baptisms as a percentage of live births.

The Religious Trends data originates from the Catholic Directory and may include ‘late’ baptisms. If so, it’s not really appropriate to divide this by number of live births per year. Spencer’s Pastoral Research Centre data (the red series) is restricted to infant baptisms (under 12 months) which is more acceptable.

2. Entries to Catholicism by type: infant baptisms, late baptisms, adult conversions.

This illustrates that a growing proportion of baptisms are undertaken when the child is older – partly because of falling neonatal mortality rates, and partly because families may opt for their children to be baptised at once, rather than individually soon after birth.

3. Catholics as percentage of England and Wales population.

This chart combines the Currie et al. (or CGH) estimates published in 1977, the estimates published in Religious Trends, and Spencer’s PRC estimates. The total population data are from the censuses (linearly interpolated until 1981, with ONS mid-year population estimates from 1981).

4. Percentage of five-yearly birth cohort identifying as Catholic, British Social Attitudes surveys 1983-2008.

Here we have combined all the 1983-2008 surveys and looked at the percentage reporting they are Catholic by period of birth (1900-1904, 1905-1909, etc., up to 1985-1989). Note however that the rate for each age group will be affected by differential mortality (and fertility) rates as well as tendency to retain religiosity or lapse.

The spreadsheet including the data used to create these charts will be posted here later today. There are data gaps which need filling in (by consulting Catholic Directories from the 1970s, allowing for the estimates being flawed) but this is as far as we have reached at present.

Update (9.30pm)
The spreadsheet is now available here.