This is a few months back, but what the heck:

Now that the 2007 Red Book numbers are out, it’s time to apply the analysis I performed for the 2006 data. This year I’ve decided to make a longitudinal comparison for 2004-2007, the years for which data is available on-line.

What is most striking to me is the utter consistency of the numbers. The rates of activities per member (baptisms, marriages, etc.) vary but slightly from year to year, showing a slight decline overall in every category of about 10% over the period.

The average rates are as follows:

Baptisms per member: 1.92%
      Child: 1.69%
      Adult: 0.22%
Receptions: 0.31%
Confirmations: 1.21%
      Child: 0.55%
      Adult: 0.65%
Marriages: 0.70%
Burials: 1.50%

As before, the rates are in proportion to each other. Burials are a bit more than twice marriages, and the latter is slightly less than child baptisms. The disturbing number, as before, is the departure rate. Baptisms plus receptions together are 50% greater than burials, and adding adult confirmations just makes it worse. Somewhere in excess of six thousand people appear to leave the Episcopal Church every year, or about 2.6% of the total membership; this contrasts with average net losses each year of about 1.9% of the membership.

So who is leaving? The conventional wisdom is that it happens soon after teens leave home. puzzling this out of the data is difficult. It’s reasonable to assume that child baptisms roughly represent births to Episcopalian parents, and these happen at over the replacement rate at about 2.4 baptisms per marriage. (Note that there is an error possibility here, because of course not all Episcopalians marry within the church. I’m assuming for the moment that marriages that take people out of the church are balanced by marriages that bring people into the church; I’ll account for that assumption in a moment.) Now, according to the CDC, about 10% of the population who survives to age 15 never marries. This is surprisingly consistent with the marriage to burial rate, although accounting for successive marriages would lead to a lower expected rate of burials to marriages. Another factor here is that people who are dying now were generally married a long time ago, mostly when the church was quite a bit larger. There actually should be a substantial excess of burials to marriages. Therefore there does seem to be a large outflow of people who have had children and then left. Probably the larger outflow is those that leave before marrying, but at the moment I haven’t figured a way to puzzle this out of the data. One of the contributors to this rate is people who marry out of the church (e.g., to Catholics—we generally would lose these marriages and the subsequent child baptisms to the Catholic church).

At any rate, the evidence is clear: poor retention is what is causing the church to decline.

See the original post here.

Thanks, Charles.  Any thoughts about what the margin of error might be, particularly given the notorious unreliability of the data (e.g. preparers of parochial reports padding the numbers or people leaving for other denominations and not being counted as transfers)?  Do statistical tests suggest that things like the sample size limit the influence of any given parish’s (or diocese’s) data quality?  Is there a tendencey not to enforce strict classification (e.g. “member” vs. “communicant”) which would skew the result?

The main factor that inflates the numbers is that clean-out of inactive members (as opposed to those who transfer or die) tends to happen at intervals rather than continuously. A lot of people make a big deal out of this but my guess is that, over the long run, it’s not that important, and it probably has little effect over long term trends. A more present issue is that dioceses tend to report old numbers for parishes which they do not actually have (due to schisms). Assuming they regain control, this has the perverse effect of reporting that regaining as a loss rather than a gain, because now the parish is reporting real data again, shows the real extent of the losses. I’m going to guess that the recent court wins in LA are going to result in a several percentage point loss in diocesan membership.

From a statistical point of view it’s really impossible to talk about the margin of error for membership because it’s something of a subjective notion anyway; the numbers probably run a bit high, but that excess is directly connected to the question of determining whether or not a person is an active member. It’s more reasonable in the case of ASA because counting heads is pretty objective. The error of course (besides bad handwriting and arithmetic) is in the counting: we use mechanical counters in our parish so I imagine our count is quite accurate. Absolute numerical accuracy is largely important when comparing with other churches, and in all but a few cases the numbers for the latter are wild-assed guesses. The Presbyterians, for example, do keep good tallies, maybe better than ours, because the higher levels of the hierarchy are funded by a head tax from each congregation. OTOH the only serious study of Eastern Orthodox numbers found that official estimates of membership were several times what the likely membership actually was.

The most important thing within the church is that people keep counting the same way. Membership numbers before and after 1985 for the whole church aren’t comparable, for example, because after 1985 the foreign dioceses were broken out as a separate category; therefore there is a large but spurious drop recorded at that time. The communicants number (used in this page from Louie Crew’s site is problematic because changes in the definition (again starting in 1986) have caused the number to converge on the active membership tally, to the point where many congregations (acto Kirk Hadaway) report the same number for both. The changes seem to have settled themselves out, as the ratios between the various aggregates (membership, communicants, others, ASA, and church school pupils) have been nearly constant since 2005 at least.

Very helpful, Charles.  This also underscores the value of your methodology in demonstrating consistency of the data over time which then supports your analysis of the dynamics.  Now, if retention is the critical issue, we then need to look at reasons for leaving, and while I believe that theology is a critical element, I wonder if many folks don’t simply cite other sorts of things which then get picked up by factions which have a theological ax to grind (please understand, I am NOT saying that theology is inconsequential; some theological positions are indeed much more in keeping with Anglicanism than others).

For example, I know of parishes where general disaffection with the rector (perhaps over issues of music or staff changes) led to what one might call a predisposition to leave.  Then, when the national church took some particular action (revised the BCP or approved the consecration of Gene Robinson), folks cited THAT as the reason for leaving the parish and the Episcopal Church.  It may well have been an important factor, but if folks had felt warmly connected to the rector and the congregation they might well have remained in the parish and on the books even while being quite vocal in their disagreement with the Church’s action at a national level.

On the other side, I am familiar with a parish in which the choir director was asked to resign (because of a professional disagreement which could not be resolved, not because of any issues of moral turpitude).  Many choir members left (and encouraged their non-choir friends to do likewise) in part around an accusation that the rector was obviously homophobic, even though there was absolutely no substantiating evidence.  This showed up on the books as a loss of membership, even though the people leaving were proabbly more supportive of recent actions of the national church.

In both these cases, one could well argue that any true understanding of the “one holy catholic and apostolic Church” is woefully lacking, and that the decisions to leave are more rooted in personal feelings about the rector (in contravention of Article XXVI) or in whether folks simply get their way or not.

Perhaps more to the point, lots of folks who may not be Episcopalians come to check out a local parish and, though they may experience a warm initial welcome, they are not drawn into the life of the congregation.  They flounder around the edges a bit trying to find an opportunity to exercise their own ministry (which they might not understand very well, either), but just never seem to make a real connection.  Thus, they go on to another nearby congregation - perhaps Presbyterian or Methodist - where they and/or their children expereince a better fit.  This group never shows up in the Parochial Report, but with some awareness and a real commitment on the part of the Episcopal parish, they might well have become a net ADD.  Unless one is heavily into litmus-testing, my experience is that folks tend to stay if treated well and a place is made for them, and if it feels “safe” to disagree with others, almost without respect to the theology evinced by the national church.

Understanding that this represents a bias on my part, I still tend to read the trends as evidence of a more generalized disaffection rather than as the result of applying a well-considered theology and ecclesiology (though some certainly do just that).  The good news is that while the trend will erode the body over time, it is not so dramatic that it will take a decade or more just to reverse the direction.  Perhaps I’m just a hopeless optimist?

I am constantly struck by the similarities in US reports, as in this discussion of membership rolls, and the Australian experience which broadly parallels earlier comments. Like other contributors, I find personal experience problematic but unavoidable given the lack of reliable research. However, I would direct anyone interested to the Australian Church Life Survey (online) which provides perhaps the best glimpse of everyday church life in the English-speaking world.

Our situation in Canberra has recently changed as it appears new privacy legislation prohibits inclusion of people on parish rolls unless they give written permission annually.

I once undertook research into parish annual returns covering a ten-year period in the late 1980s-90s. The returns were filed by the rector and kept in filing cabinets in the diocesan registry. I was assured that no-one took much attention of the returns and most laity were unaware that they existed and my clear impression was that the bishops and archdeacons relied more on personal contacts.

Another recent study of the same diocese confirmed that attendances had been declining steady for no apparent reason other than cultural shift affecting all mainstream Australian denominations. Churchmanship, broadly speaking, seemed irrelevant in statistical terms although one or two parishes had core congregations of very strong values. A common feature was the aging of all congregations with most regular worshippers nearer sixty years of age than the fifty years reported twenty years earlier. The virtual disappearance of Sunday Schools and a weak profile of ministry to teenagers and younger adults was also characteristic.

I noticed the reference to Presbyterian roll-keeping. In my experience, this is attended to at each monthly session meeting (ruling elders with teaching elder). Of course, theory and practice vary. I recall one Presbyterian minister who insisted that a 1960s figure for Sunday School attendance was still relevant in the 1990s.

My longer Anglican experience has been that ‘purging’ the roll is rare and depends very much on the attitudes of the rector. As a churchwarden I found that people often remained listed long after they had died, departed the parish, etc. I do not recall a serious discussion on membership rolls in several decades of vestry membership in different places. It was unusual for new arrivals to be questioned, let alone asked for a reference from a former rector or any paper evidence of church affiliation.

My experience is that theology is not a major strand in member arrivals or departures. Obviously an extremist of any persuasion is more likely than not to seek to join a church broadly sympathetic to their views, although I am familiar with a very low church cum evangelical parish that counts quite a number of avowedly Anglo-Catholic to Broad Church members and not a few Roman Catholics to leaven the lump. I also know an Anglo-Catholic parish that is comfortable with low church members.  Most rectors seem pretty charitable about such issues.

I think ministerial personality is the critical factor for many people. Changes of rector often seem to lead to departures when parishioners find themselves no longer emotionally close to their clergy leaders and can lead, over time, to a final separation from the denomination and more generally, from churchgoing. Somewhere in this, I suspect, is something to do with issues of authority and status, coupled with education.

The general Australian rule seems to be that clergy must move on after ten years and this may be reflected statistically as people fail to identify, for whatever reasons, with a new leader. I know of no detailed studies on this dimension.

One pattern that has emerged very clearly in Australia in recent years is that restrictive rules on infant baptism and insisting that both partners in marriages be regular worshippers has accelerated separation from the churches. Infant baptism has almost disappeared in Australia and considerably less than half of all Australian marriages are conducted by clergy. A result of the baptism decline has, of course, been a parallel decline in confirmations and hence in communion membership.

Ian Welch, Canberra

Most interesting, Ian.  Any early indications of how parishes are dealing with the administrative change requiring “written permission” to be listed on the roll?  I would assume that we could certainly count signed pledge cards, though that would cut the roll by about half… 

Pledge cards are less and less common in Canberra Anglican churches, and indeed, I think are rare generally in Australia. In short, Australians have a very different attitude to managing church finance than is my impression of the US situation.

Tithing has never caught on, in part for theological reasons but encouraged by Australia’s very high (by US comparisons) overall tax arrangements.  Church and evangelistic (as in the case of missions) giving in Australia is not deductible from income tax. Most Australians cannot deduct individual payments as the government allows a non-taxable amount of initial income. After that, it is open slather for the taxman. The secret to tax avoidance (and evasion of course) is to minimise ‘official’ income, something limited to those with smart legal and accounting advisers.

Pledged giving of the ‘Wells’ kind of the 1950s faded several decades back, although the idea of envelopes had endured and many parishes still use numbered envelopes and keep a record of the amount given against the number, and this aggravates some people who regard this as an intrusion on their privacy. Giving is not part of the parish roll process.

As I understand the new Canberra situation re parish rolls, it is necessary for the member to specifically authorize, in writing, the inclusion of their name and address on the roll each year. In my parish, this is done by filling in a card and placing it in the offering plate. The same card is used for requesting prayers, or making other constructive comments, so it is not an innovation for regular worshippers. I did note, however, that our printed parish roll, issued annually, seems down in size over last year. The parish recently ‘planted’ a new congregation in a totally new suburb at the other end of town and this may have affected the numbers somewhat. However, Australians have a rooted aversion to any kind of form filling so I suspect that the ‘real’ game is that you stay on the roll unless you specifically ask to be removed, at least for the time being.

Our parish membership is the largest in the diocese and among the largest in Anglican Australia, at around 500 folk. Our general Anglican pattern seems similar to American Episcopalians, with c200 adult members being considered good. Many of our inner city churches, and many in rural areas, have less than fifty adult members as is, I gather, also the case in the US. One result is that the normal situation in Australian Anglicanism is one parish, one clergyman and that is reinforced by parish financing.

Ian Welch, Canberra

You might want to read FACTs on Episcopal Church Growth. It’s the result of a 2005 survey. Now, it must be read carefully, because it is chock-full of statistical booby traps: there are hidden correlations all over the place. For example: look at Figure 1, which shows that growth is most common in parishes in a “newer suburb of city”. OK, well, there’s an obvious reason for this: these are areas that are growing, generally rapidly, so it’s not surprising that parishes in such areas grow. But there are also some other correlations. For example, it’s a safe bet that these areas skew towards families with youngish kids and away from retirees. So we move on to some of the other charts. See Figure 2? Well, there’s a hidden correlation there: newly planted congregations are going to be put in areas that are growing. Moving on to figures 5 and 6, it seems obvious that parishes with lots of old people are going to fade, but again, there’s that correlation with regional growth: areas which are growing may have more kids than usual, so one would expect growing parishes to have more kids than usual. (To throw an old but probably not irrelevant data point in here: the development I grew up in was a monoculture of young families, all moving in over a period of a very few years. It naturally caused, along with its neighbor developments, a huge spike in church growth. Fifty years later, the neighborhood is full of retirees: not much potential for church growth there.)

Let’s move on to a different problem correlation. Look at figure 13, which shows that congregations that have more services grow more. Now this is a fallacy, because parishes that have more services have grown and thus become large enough to require multiple services; there is no promise of further growth. That, by the way, is an issue with the whole report: the assumption is that past growth is an indication of future growth. But as my anecdotal case shows, it doesn’t work that way. There are several cases in the report where being “young” in some sense gives a parish growth. Of course, one does not stay young: new suburbs become old, and families with young kids become empty nests, and newly arrived rectors become old and established.

I want to come back to the previous point, but let me move on first to the next set of figures, having to do with programs. Here the issue again is that we’re looking at churches that have grown, so it’s not surprising that churches that are larger have more resources to devote to all the various things that are held to correlate to growth. Even the most promising finding, the correlation between service diversity and growth, is not immune: given the typical 8 & 10 two service pattern, one suspects that 10 is likely also to be traditional; but with three services it is common to see a “9 contemporary/11 trad” pattern, so it may be that being larger allows more service diversity.

I don’t mean to be too hard on Hadaway on this, as some analysis is better than none. And there are some nuggets there, such as Figure 25, which shows pretty clearly that there’s an optimal window for clerical tenure (and it’s not exactly five years—that would mean moving everyone at their peak effectiveness). The temptation in reading this study is to say, “do what big parishes do”. But a parish that’s not big now and is not in the “right” place (geographically or historically) is limited in its ability to conform to that model, and never mind whether applying that model to a small parish would result in growth. Part of our heritage is serving large areas besides the newer suburbs that are easy pickings for evangelical megachurches. I’ve attached a graph which shows that only about 5% of parishes are in that desirable “new suburb” category. OK, so what do we do for the 35% that are in small towns or out in the country, or the 24% in middle-sized towns? These don’t fit well into the model because growth for them doesn’t mean getting big: it means getting less small.

And hanging over it all is 2003. Almost every graph of ECUSA numbers starts a strong downward trend in that year.

I place greater value in data like average Sunday attendance than in the data about membership. I think clergy have wildly different approaches to membership numbers, but not about attendance figures. All such data tends to be inaccurate, but I think attendance is likely to be fairly consistently accurate/inaccurate. There have been periods in TEC’s history when reported membership was in decline and attendance was increasing. I pay attention to membership numbers only once a year - when preparing the parochial report - but I pay attention to attendance numbers all year.

Daniel, part of it depends on where you think the inaccuracy of membership lies. It’s striking how, over the church as a whole, the ratio between membership and ASA is essentially constant over the years, at least within the last 10-15 years. This suggests that whatever inaccuracy is going on is at least consistent. It would be of interest to investigate what produces the ratio: some is produced by people missing Sundays (for whatever reason), and some is produced by the C & E crowd, and some is produced by the failure to keep the rolls clean, and some is produced by leniency as to what constitutes activity. For my purposes the membership numbers are more directly useful because baptisms and burials figure directly in that number.

You are keeping your rolls up-to-date, right?

I’ve attached the single most depressing chart, from this 2004 report, where it appears as figure 7. Note that this report is exactly old enough to where the post-2003 decline had yet to appear, which you should bear in mind in reading it. Anyway, what this chart shows is a depressingly strong correlation between mainline growth/decline rate and the white birth rate (.94 for all mainlines, .89 for ECUSA). Either we’ve got to get Episcopalians (and for that matter, Baptists and Methodists, since there are so many more of them) to breed more, or we need to investigate, far more aggressively, why people are coming and going.