Saturday, September 20, 2008

Why the Vote Was Wrong

Louie Crew does not waste any time in updating his database on Episcopal Church bishops. Accessed just one day after the vote in the House of Bishops to consent to the deposition of the Rt. Rev. Robert Duncan, who is still (as of today, September 20) the diocesan of Pittsburgh, it already shows the see of Pittsburgh as "vacant." (The certificate of deposition will probably not be signed until Monday or Tuesday, in New York.)

It is, of course, a minor quibble, because the signing of the certificate of deposition (which will add one more offense to the catalog of canonical violations committed by the current presiding bishop) is a foregone conclusion. But it is yet another confirmation of the rush to judgment that was the most significant aspect of the vote to "depose" Bishop Duncan. In this post, I want to share the results of some mathematical analysis of the vote, made possible for the first time by the parliamentary request to have a roll-call vote.

(Praise and thanks are due to the stalwarts whose names appear below for the following action taken in the House of Bishops, as recorded in the official minutes, which preserved the vote for the analysis given here:
Bishop William Love then called for division with the following statement: “We the undersigned being voting members of the House of Bishops, hereby call for division of each and every vote of this House on any matter put to vote with respect to the status of the Bishop of Pittsburgh as a bishop and member of this House of Bishops and that each member’s vote, whether it be aye or nay, be entered in the journal.” This above call was signed by Bishops William Love, James Adams, David Reed, Michael Smith, Edward Little, Geralyn Wolf, Mark Lawrence, John Howe and Bruce MacPherson.)
There were 88 recorded votes to consent to Bishop Duncan's deposition, 35 votes against (counting a switch at the conclusion of the vote by the Rt. Rev. Dorsey Henderson of the Diocese of Upper South Carolina, who changed his vote from "aye" to "nay"), and four abstentions. For the purpose of the analysis which follows, I have used the helpful spreadsheet prepared by one of the industrious Elves at Titus OneNine from the official minutes of the meeting, and have made certain corrections with respect to some of the dioceses, and then built upon it, in the ways indicated below. For most of the analysis, I have counted the abstentions as "No" votes, because their effect is to deny a "Yes" vote just as much as a "No" vote does.

All right, let us begin with the first observations. Based just on the raw numbers, it would appear as though the motion to depose carried by a vote of 88 to 39, or 69% affirmative to 31% negative (remember, counting abstentions as a "No" vote). That would appear to be a vote that carried by more than 2 to 1, and so was definitive, would it not? (Incidentally, there is no mathematical way I can see to come out with the 64%-36% tally as reported here. 88 affirmative votes would be 64% of 137 total votes, not 127; while 39 votes would be 36% of 108 votes, so I cannot reconcile Steve Waring's numbers.)

Well, let us ask what the vote was by diocese. This will have the advantage of sifting out the "piling on" effect resulting from when, for example, four bishops from the Diocese of Los Angeles (the diocesan, the suffragan, and two assistant bishops) vote as a bloc. The results are as follows: 60 of 110 dioceses in The Episcopal Church voted for deposition, 33 voted against, plus the Convocation of American Churches in Europe ("CACE"), which is technically not a diocese, abstained, and thus made a total of 34 not voting to depose; and 17 dioceses were absent and not represented in the voting.

(Note to explain the slight discrepancies between my analysis and the one done in the spreadsheet at Titus OneNine: I have only 110 total dioceses, as shown and listed on page 6 of the 2008 Red Book. This counts the ten foreign dioceses in Provinces II, VIII and IX, each of which has a diocesan bishop entitled to a seat and vote in the House, but it does not count (as the Elfgirl did) the Episcopal Church in Micronesia, which is under the jurisdiction of the presiding bishop, and which does not have a separate bishop of its own. CACE is a special "case" [my apologies if you want them, but dry statistics can always use a pun or two]. While also not a Diocese as such, its Bishop is the presiding bishop of TEC, who belongs to no proper diocese, and it also has a "bishop in charge", the Rt. Rev. Pierre Whalon, who is under the supervision of the presiding bishop. Since both of them have a seat and a vote in the House of Bishops, and since both were present at the meeting, the Convocation needs to be counted just like a diocese in tallying the votes. Finally, in counting the number of dioceses who "voted" through a representative bishop, I included the Diocese of Pennsylvania, because even though its current bishop is inhibited and was unable to participate, its retired ["resigned"] Bishop, the Rt. Rev. Allen Bartlett, Jr. [who is currently serving as an Assistant Bishop in the Diocese of Washington] was present and did vote. Likewise, I have included dioceses where the Suffragan or Assistant was present and voted, but not the Diocesan [e.g., the Diocese of Alabama, where the Suffragan Bishop John McKee Sloan voted, while Bishop Henry Parsley was not at the meeting; this was also the situation for the Diocese of Long Island and the Diocese of Massachusetts]. So that explains, I believe, why T19 is reporting that only 56 dioceses voted to consent to deposition, while I have the number at 59. In either event, as both of us show from the numbers, it was not a vote of which TEC can be proud.)

So with those numbers, the percentages now work out like this:

Dioceses FOR: 59 (53%)

Dioceses AGAINST: 34 (31%)
(including CACE)

Dioceses ABSENT: 18 (16%)

Fifty-three percent is quite a different number than sixty-nine percent (the count by number of bishops), is it not? But wait---the numbers get worse.

The 88 to 39 tally counts ten resigned (technical TEC term for "retired") bishops who showed up in Salt Lake City to vote. (That is a story in itself: all ten who showed up voted in favor of deposition---including avowed Duncan opponents such as Bishop Otis Charles, who is [according to Louie Crew, who makes it his business to know] the Church's only other openly homosexual bishop besides the Rt. Rev. V. Gene Robinson of New Hampshire [who also voted to depose].) Also there was retired Bishop of Washington Jane Dixon, whose lawsuit to depose the Rev. Samuel Edwards Bishop Duncan actively opposed. It would thus appear that some of the retired bishops made it a point to come to the meeting so they could vote to depose Bishop Duncan.

But why should just the retired bishops who show up to vote be counted? If we want to compare bishops the way we did dioceses above, we need to count those bishops (both active and retired) who were absent from the meeting in order to get the true picture.

There are 300 current bishops in the House (as of August 2008---see the previous link), of which 152 are retired, and 148 are active. This means that, with only 117 (78+39) active bishops present at the meeting, there were 31 who were absent, as well as 142 retired bishops absent. So if we now recalculate the vote with those numbers, we have:

Bishops FOR: 88 (29%)

Bishops AGAINST: 39 (13%)

Bishops ABSENT: 173 (58%)

This looks very different, does it not? Only 29% of "the whole number of bishops entitled to vote" in the House of Bishops voted for deposition, while 71% did not concur, either by voting actively against, or by staying out of the vote. And yet the leadership of TEC proclaims that Bishop Duncan was lawfully deposed by a vote of his peers!

Now let us tease out the numbers in a different way. Suppose we go back to the vote tallied by diocese, and append the following data to the numbers: the number of parishes in each diocese, the number of baptized active members in each diocese, and the Average Sunday Attendance for each diocese (all data are from the 2008 Episcopal Church Annual, otherwise called the "Red Book"). We then run the totals by "yes" vote and "no" vote of each diocese, and we find:

Parishes represented in "YES" votes: 4,344 (58%)

Parishes represented in "NO" votes: 2,282 (30%)

Parishes not represented at meeting: 884 (12%)

Again, this is hardly an overwhelming majority of TEC parishes---and it ignores the fact that there are so many retired bishops entitled to vote who do not represent any jurisdiction.

(Note to techies: the total of 47 parishes for the "[Episcopal] Diocese of San Joaquin" as shown in the 2008 Red Book had to be adjusted downward to 18, the number currently claimed by that Diocese. And in the numbers that follow, estimates for baptized parishioners and ASA were used---but the overall totals are so large that the margin of error in the estimates could be 100% without affecting the percentages.)

It also turns out that the numbers for baptized members are within a percentage point of the parishes. And the ASA numbers show only that the dioceses that were unrepresented at the meeting had a much lower ASA than those that were: 4,964 versus 7,967 for the dioceses voting "yes" and 7,103 for the dioceses voting "no".

The numbers really start to get interesting, however, when one looks at the geographical spread of the data, and considers the level of each diocese's 2007 contribution to the TEC budget. Here is where the data starts to be telling: it shows that Bishop Duncan was deposed by a combination of the dioceses that are the biggest contributors overall to TEC, as well as by those that are in what has been called, in the political arena, the "blue-county corridors." (Click here for an animated map of how these areas have changed in the presidential elections from 1960 to 2004. Are we surprised?)

Total contributions to TEC by "Yes" dioceses: $20,593,549 (72%)

Total contributions to TEC by "No" dioceses: $ 6,237,162 (22%)

Total contributions by unrepresented dioceses: $ 1,621,881 (6%)

Do you begin to see how TEC is run by the wealthiest players? Only fifty-four percent of the dioceses voted to depose Bishop Duncan, but they contribute 72% of the funds coming to TEC from all the dioceses. (It's also more than three times as much as the "No" dioceses contribute.) The ones that voted against, of course, include dioceses like Dallas, which contributed zero dollars to TEC in 2007. They also include dioceses that are having trouble keeping members, like Central Florida (contributed just 2% of its income to TEC, instead of the recommended 21%), Rio Grande (4%) and Tennessee (4%).

[UPDATE 09/22/2008: A commenter at Titus OneNine suggested it might be more worthwhile to compare dioceses not by their contributions to TEC's budget, but by the sizes of their own budgets. That is easily done, because the TEC data gives the diocese's individual contribution as a percentage of its total budget---and one has also to correct for the exemption in the first $100,000 of income which TEC grants each diocese. The results come out as follows:

Total budgets of "Yes" dioceses: $129,911,718 (67%)

Total budgets of "No"/"Abstain" dioceses: $ 47,155,287 (24%)

Total budgets of absent dioceses: $ 15, 537, 963 (8%)

So the skew still holds: the "money" dioceses, while contributing 53% of the votes to depose, spent 67% of the total amount spent in 2007 by all dioceses in ECUSA; those dioceses spending just 24% of the total voted "No", with 31% of the total votes. From these figures one can also derive the following average comparison (in which all the "zero" contributions are smoothed out):

"Yes" dioceses contributed, on average, 16% of their non-exempt income to TEC, while

"No" dioceses contributed, on average, 13% of their total non-exempt income to TEC, and

Unrepresented dioceses contributed just 10% of their non-exempt income to TEC.]

The situation is even worse if we look just at Province I. Of the seven dioceses in that northeasternmost Province, only one---Rhode Island---voted against the deposition, and it contributed its 21% of income, or $436,335 to TEC in 2007. This amounted to 11% of the funds coming from that Province. The other six dioceses (Maine, Vermont, New Hampshire, Connecticut, Massachusetts and Western Massachusetts), however, all voted to depose, and together (89% of provincial contributions) accounted for a total of $3,395,472 to TEC, or a hefty 12% of contributions from all dioceses. (All dioceses were 21% tithers to TEC in 2007, with the exception of Vermont, which gave TEC 15% of its income in 2007, and Massachusetts, which gave 22%.)

Likewise, in the huge western Province VIII, only the absolute smallest of TEC's 110 dioceses---the Navajoland Area Mission, with just 9 parishes---did not vote for the deposition (it abstained), while four were absent and thirteen dioceses voted to depose. Just as in Province I, these latter dioceses were responsible for 89% of the total contributions to TEC from Province VIII. Together, the nineteen dioceses voting for deposition in Provinces I and VIII alone (17% of the total dioceses in TEC) were responsible for 28% of the overall contributions to TEC's budget.

The only province in which more dioceses voted against deposition (7) than voted for it (3) was the midwestern Province 7, which stretches from Kansas and Western Missouri south to the Texas and the Rio Grande.

Here are the diocesan voting and absentee totals organized by Province:

Province I: Yes - 6; No - 1; None absent.
Province II: Yes - 6; No - 4; 1 absent.
Province III: Yes - 8; No - 3; 2 absent.
Province IV: Yes - 10; No - 9; 1 absent.
Province V: Yes - 7; No - 4; 4 absent.
Province VI: Yes - 4; No - 3; 1 absent.
Province VII: Yes - 3; No - 7; 2 absent.
Province VIII: Yes - 13; No - 1; 4 absent.
Province IX: Yes - 2; No - 2; 3 absent.

A final useful comparison we can draw from the spreadsheet is when the vote is arranged by the length of tenure in office of the bishops. (As noted earlier, all ten retired bishops present voted for the deposition.) Here is a table showing the votes tallied by years of consecration, starting with the most recent first (and taking account of absences):

2005-2008: Yes - 18; No - 10.
2001-2004: Yes - 18; No - 11 (absent: 3).
1997-2000: Yes - 18; No - 9 (absent: 4, including +Duncan).
1993-1996: Yes - 16; No - 4 (absent: 6).
1967-1992: Yes - 18; No - 5 (absent: 1 [+Wimberley]).

These subtotals tell us that the level of opposition to +Duncan remained fairly constant despite the length of time a bishop had been in the Church, but that the number of his friends decreased with age. That trend might explain something about the current difficulties in TEC, and show how difficult it will be for it to reverse its course any time soon. [UPDATE 09/22/2008: Another commenter has pointed to the fact that the "No" votes increase in number as the bishops get younger, and cites this as a positive indication that the younger bishops care more about following the canons than do the older ones. Two different ways of seeing the figures---but saying what amounts to the same thing: it was chiefly the older bishops (those consecrated before 2001, who cast 59% of the votes to depose) who decided that Bishop Duncan had to go!]

What can we conclude from all these numbers? Given the ruling that only a majority of bishops present and voting are needed to depose, we have the following possible anomaly. Only active bishops are counted towards a quorum of the House, so of the 148 total active bishops in the House, 75 are needed for a quorum. A bare majority of such a quorum would then be 38 bishops. It would theoretically be possible, if the very largest dioceses and the very smallest dioceses made up the 75 present for a quorum, to depose a bishop on the 38 votes of those bishops who collectively represented just the least populated dioceses. (Due to the difficulty of presenting tables that work in all types of browsers, you can view the tabulation at this link.)

The tabulation shows that under the interpretation given by TEC's current leadership to Canon IV.9's language that it requires a "majority of the whole number of Bishops entitled to vote" in the House of Bishops, it would theoretically be possible for just 38 diocesans, representing just 1,263 of TEC's 7,510 parishes (about one-sixth), and 202,349 of its 2,310,094 members (less than one-tenth!) to unseat any bishop in the House without a trial, on charges that he/she had "abandoned . . . the doctrine, discipline or worship of this Church," or, in the language of the proposed revision to Canon IV.9 (to be renumbered IV.16), "in any other way."

By doing so, those 38 diocesans could flout the will of the 37 largest dioceses in the Church, having 4,184 parishes and 1,487,109 baptized members, or more than 64% of the membership of the Church.

Such will be the result if the changes to the Canons now being talked about are made at General Convention 2009 in Anaheim. And such already is the de facto situation in TEC, under the arbitrary interpretation of Canon IV.9 imposed on the Rt. Rev. Robert William Duncan.


  1. Question: How many of the Retired Bishops who voted for Deposition usually attend the HOB Meetings?

    Lois Smith-Ellingson
    Rocky River OH

  2. Lois, there are no published records I am aware of that would enable one to answer your question. I imagine it depends a lot upon where the meeting is held, and who lives close enough to come, etc. I can tell you this: as the number of retired bishops has gotten larger, and as not very many of them show up for meetings (10 out of 152 at this one), the pressure is building for a change to the Constitution that would take away their right to a vote altogether. (It passed the House of Bishops once, but did not pass the House of Deputies.)