Epic Statistical Fail

The Studying Your Congregation and Community tables have been updated for 2009, and the next phase of departures can now be assessed (at least for the patient and dogged) for the effect on church numbers.

Except that they can't.

If you pull up the chart for the Diocese of San Joaquin, yes, it is still almost entirely gone. Pittsburgh also has dropped precipitously. The charts for Fort Worth and Quincy, however, don't look that different: Ft. Worth ASA shows a maybe 30% ASA decline, while Quincy shows a tiny increase. OK, what's going on here?

Well, the answer is that there is apparently a large scale pattern of misrepresentation if not outright fraud. Let's work with Ft. Worth, which in 2008 had 55 parishes. Well, looking through the charts, only 17 parishes show changes in reported numbers between 2008 and 2009, and there are only a couple more which show changes in reporting between 2007 and 2009. They are simply repeating the old values rather than acknowledging that the other parishes no longer count themselves as ECUSA congregations.

Of the seventeen that do show changes, only one does not show a loss from 2008 to 2009, and all show losses from 2007. Only three do not appear in ACNA Ft. Worth's list of parishes, and these three have relatively small losses (13%, 25%, and a gain of 11%). There are three other apparently contested parishes with relatively small losses (9%, 20%, and 35%). Every other parish lost at least 60% of its attendance, with the worst case going from 94 to 2 ASA, a loss of 98%.

Taken together these parishes have a combined ASA of 970, down from 2218 in 2008 for a loss of 56%. If one assumes that all the other parishes haven't reported numbers because they are no longer part of the ECUSA diocese, then given a reported 2008 ASA of 6945 (and this number isn't really accurate, because few parishes reported a change in ASA between 2007 and 2008) the drop in ASA was not around 30%, but in excess of 85%. It's a good thing that Ft. Worth is a relatively small diocese, because the loss I compute is 0.8% of total domestic ASA.

One hopes that there is a good (read, canonical and longstanding) reason for this kind of reportage. I'm told that the three phantom parishes in LA are still on the books, and one is led to suspect that there are others that haven't been checked. And of course, there's also Quincy, which at an ASA of 935 in 2008 isn't going to be sending statistical shockwaves out when its reporting is rectified, but still, it all adds up.

UPDATE
The diocesan numbers have now been released, and at least one of the peculiarities has been remedied: numbers for Ft. Worth now show departures. The Quincy numbers have also been updated, but they show little change from the previous year. I will have more analysis in a separate post.