Churches in Australia, Part 2

Some indication of how the church groups in the previous post differ can be gained by looking at their core belief statements posted on the web. The Lutheran, Salvation Army, and Assemblies of God websites are overtly religious, for example, while some of the other churches reflect primarily on their history. The "success" score from the previous post correlates weakly (r = 0.7) with the number of times the Holy Spirit is mentioned.

Summarising the data:

Group	Total Word Count	Jesus/Christ mentioned	Holy Spirit mentioned
Anglican	212	2	0
Baptist	168	2	1
Churches of Christ	15	0	0
Lutheran	190	8	1
Salvation Army	328	7	1
Uniting	94	0	0
Pentecostal (AoG)	309	5	4
Presbyterian	594	1	0

For comparison, the evangelical organisation FOCUS has numbers 235, 9, 4.

The picture below positions churches as in the previous post ("size" vertically, and "success" horizontally). However, this time the area of the circles gives the word count of the belief statement, and the colour (red, yellow, green) gives the number of mentions of the Holy Spirit (0, 1, 4):

Church Attendance in Australia

I've been meaning to summarise the 2001 National Church Life Survey data ever since I found it on the web, although obviously more recent data would be better.

Group	Att (2001)	% Ch	Census (2001)	% Att	Congr. (2001)	% Ch	Census /congr.	Att /congr.
Anglican	177,700	-2	3,881,162	4.6	3,128	-5	1,240.8	56.8
Baptist	112,200	8	309,205	36.3	927	7	333.6	121.0
Churches of Christ	45,100	7	61,335	73.5	448	-2	136.9	100.7
Lutheran	40,500	-8	250,365	16.2	605	-3	413.8	66.9
Salvation Army	27,900	-7	71,423	39.1	359	-3	198.9	77.7
Uniting	126,600	-11	1,248,674	10.1	2,373	-15	526.2	53.4
Pentecostal	141,700	18.6	194,592	72.8	1,207	11.5	161.2	117.4
Pres/Reformed	42,100	-2.7	637,530	6.6	745	-4.9	855.7	56.5

In the table, "Att" is attendance, and "Congr." is number of congregations. Percentage changes (% Ch) are from 1996, while "% Att" is the percentage of census-tickers attending church. The last two columns are census-tickers per congregation and average attendance per congregation.

In the picture below, size of the circles indicates number of congregations (in 2001), while colour indicates average attendance per congregation.

Now there's too many numbers there to digest, but many of those numbers are highly correlated, and statistically most of them break down to two main factors (as identified by Principle Components Analysis). I can get an overall size factor by combining the census data, total attendance, and number of congregations (for statistical reasons, I multiply them and take the cube root). This size factor forms the vertical axis of the picture.

I can get an overall success factor by averaging the percentage change in attendance, the percentage change in number of congregations, the percentage of census-tickers attending, and the average attendance per congregation (multiplying the latter two by 0.5 to avoid having them dominate the result). This success factor forms the horizontal axis of the picture. The Pentecostal churches are doing best, because of their rapid growth, the combined Presbyterian/Reformed group is very slowly shrinking, and the Uniting church is in collapse.

The correlations seem to indicate two fairly simple facts: (1) growing congregations tend to be larger, and (2) if you can't attract your own nominal members, you don't have much chance of attracting complete outsiders either.