I would suspect that most of you have heard by now that the unofficial vote tracking on Amendment 08-B places the count as 69 yes and 89 no as of last Saturday, a sufficient number to defeat the amendment.  It appears that the “fidelity and chastity” section in the Book of Order for the Presbyterian Church (U.S.A.) will remain for another two years.  That still needs to be verified by the Office of the General Assembly, but based on the official vote tally this appears to be the only item sent to the presbyteries that will not pass. (At this time amendment 08-I is listed as very close to passage but not yet.)  But voting is not over yet — there are still 15 presbyteries that need to vote and the General Assembly recognized that the process, and not just the vote, was important.

However, the results appear certain enough that the Presbyterian News Service has issued an article and the reports are spreading around the news services (exempli gratia Associated Press, The Christian Post, Advocate.com, Dallas Morning News), the advocacy groups (exempli gratia More Light Presbyterians, Presbyterian Coalition, Witherspoon Society), and the blogs (exempli gratia Presbylaw, Psalms Modern, A Classical Presbyterian, Ray’s Net, Mark Time).

Having now had 158 presbyteries vote, and 143 of those presbyteries with vote counts on both 08-B and 01-A recorded at PresbyWeb or the Presbyterian Coalition counting sites, there is a significant amount of data to crunch to compare the two votes and see if it says anything about the PC(USA).

Now, while I have some questions that the two amendments are really comparable since the text of the two is significantly different in content and action, it is still my conclusion that in many quarters they are viewed as similar actions.  For most of this analysis I will take it as a precondition that the two amendments are similar enough in their perceived intent, if not their text, that it is valid to compare the voting numbers.

I will break this analysis into several different posts primarily so as not to overwhelm the casual reader with extensive statistics.  As a research scientist I am used to providing and drinking numerical data through a fire hose.  I am going to try to spare you the experience.  Also, some of the individual case studies will wait until all the presbyteries have voted.  But with over 90% of the data in I will go ahead today with the summary statistics of the population.

Finally, as a research scientist I accept peer review and as a Presbyterian I welcome accountability.  If anyone does want to see my raw data I will gladly send you a copy of my source spreadsheet once I have most of my analysis presented.

And a word on philosophy:  I sometimes wonder if some of my readers view this as “dwelling on the past,” “rearranging deck chairs on the Titanic,” “majoring in the minors,” or “analyzing the obvious.”  I however consider this interesting (yes, I am weird), I am concerned about some of the other statistics and their interpretation I see out in the news, and I do feel that taking a serious look at these things is part of the third note of the True Church – “ecclesiastical discipline uprightly ministered.”

Summary Statistics
From the data sources listed above there are 143 presbyteries that have voted on Amendments 08-B and 01-A for which vote counts are listed in the sources.  In the discussion that follows I will only be addressing presbyteries that have reported numbers for both votes.  There are 15 presbyteries that do not have reported numbers for one or both votes.

On 01-A 42 (29.4%) of these presbyteries voted “yes” and 101 (70.6%) voted “no.”  On 08-B 64 (44.8%) of these presbyteries voted “yes” and 79 (55.2%) voted “no.”  Of the presbyteries that voted “yes” on 01-A two (1.4%) have voted “no” on 08-B.  Of the presbyteries that voted “no” on 01-A 24 (16.8%) voted “yes” on 08-B.

Looking at the number of reported votes, on 01-A there were 21,732 total commissioners voting in these presbyteries.  Of these 9,375 (43.1%) commissioners voted “yes” and 12,357 (56.9%) voted “no.” For 08-B there were 18,562 total commissioners voting in these presbyteries.  Of these 9,189 (49.5%) voted “yes” and 9,373 (50.5%) voted “no.”  (Note: the data sources do not include blank or “abstain” ballots.  From experience these are <5 per presbytery and using an average of 3 per presbytery it could be another 429 ballots or roughly 2% of each vote that I would estimate as an upper limit.)

Between 01-A and 08-B the number of total voting commissioners in these presbyteries declined by 3,170 which represents a loss of 14.6% of the 01-A total votes.  The decline in commissioner “yes” votes is 186, a 1.9% drop relative to the 01-A “yes” total and a 0.8% decline relative to the total number of votes cast.  The decline in commissioner “no” votes is 2,984, a 24.1% drop relative to the 01-A “no” total and a 13.7% decline relative to the total number of votes cast.

Preliminary Analysis Comments
I don’t want to make any substantial comments on the analysis and conclusions until I have spread out some more detailed statistics in front of you.  However, let me set the framework in which I have been studying these numbers.

In modeling the data I have selected five different factors that I think are influential.  These five factors pretty much cover any of the reasons for changes in the vote numbers and so as a whole probably introduce too many degrees of freedom.  However, in working with the numbers it seemed that relying on only the three “general” factors still left out some identifiable variation.  This is part of what prompted my “every presbytery is different” post a bit over a month ago.

The nice thing about working with the overall statistics is that the larger population size should minimize the influence of the special cases and that individual special cases might average, or cancel, out.  I will investigate each of these in detail later, but briefly the three general
factors that I am working with are:

1) Overall, uniform membership changes.  This is the documented membership change (generally decline) in the membership of the PC(USA) and how it would translate into changes in the number of commissioners voting.

2) Vote changes.  This is the switching of commissioner votes from “yes” to “no” or “no” to “yes” between the two votes.

3) Selective decline due to realignment of churches.  This is not the uniform membership decline but the selective departure of churches and individuals of one particular theological perspective that has been happening over the past few years.  The theory is that it is primarily conservative churches that are leaving the denomination so this should manifest itself as a preferential decline in “no” votes.

There are also two special cases that I am considering.

A – Fundamental change in the presbytery.  In some (probably limited) cases there are changes the presbytery has made, apart from typical membership changes, that would influence the number of commissioners voting.  The changes to counting active membership in San Gabriel Presbytery would fall into this category (I discussed that back in March.)

B – Special circumstances of that meeting.  Situations where some external cause influences the number of commissioners at that particular meeting.  There was discussion that the number of commissioners at the John Know Presbytery meeting was significantly reduced (60%) by a winter storm and that there were conflicting conferences that influenced the attendance at the San Francisco Presbytery meeting.

It appears that both of these special cases are very limited.  While it is tempting to consider the factor as uniform across “yes” and “no” votes, if a special circumstance was involved in the San Francisco vote change is was clearly not uniform.  I will drop the special cases for now and return to that topic a few posts from now.

So, looking at the changes in the summary statistics what can we say as a first pass?  The number of both the “yes” and the “no” votes declined but the “yes” only slightly and the “no” substantially.  You can not explain the difference with only changes in the vote.  You can not explain the difference in the votes with any one of these three factors alone.  A combination of two or more is required.

(Factors 1&2) If you want to say that the difference in the total is uniform decline then you could expect about 8000 “yes” votes on 08-B based on the 01-A percentages.  That would mean that there was a net change of 1200 commissioners (6.5% of the 08-B total) changing votes from “no” to “yes.”

(Factors 1&3) You could also interpret the numbers to say that there was no changing of votes, but rather the differences in votes reflects a 0.8% uniform decline (the 186 vote decrease in the “yes” votes) and then an additional 12.9% decline in the “no” votes due to conservative departure.  (That would be the 2984 total “no” vote decline split between 188 uniform decline and 2796 selective decline.)  With a total uniform decline of a bit less than 400 votes in this scenario the conservative departure is clearly dominant and this comes closest to explaining the voting differences with a single factor.

(Factors 2&3)  The other possibility is that there is no uniform decline but the 3170 vote drop in numbers reflects the loss of only conservative “no” votes combined with 186 “yes” votes switching to “no” votes to account for the drop in the number of “yes” votes.

From the summary statistics we can probably say all three of these factors are present but it is difficult to distinguish the level of influence of any of these three factors individually.  As this series of posts progresses I will work my way up to my model where individual presbyteries can be classified as having one or two of these factors dominate the vote changes.  The factors will get limited on a presbytery level so we have an over-determined rather than an under-determined matrix for the model.  (That is mathematical jargon, not a psychological analysis of the matrix.)  And I have found that there are a couple of presbyteries where there is statistically no change in the vote pattern.

But all that is in the future.  For today it is enough to say that from my analysis of these summary numbers the statistic that really jumps out is the 13.7% drop in the number of “no” votes between 01-A and 08-B.  Based on other membership numbers it appears unlikely that this drop could be accounted for in uniform decline alone and it can not be purely vote changes since the total numbers show a similar 14.6% decline.  The question then is how much of the vote shift seen between 01-A and 08-B is truly a shift at the individual level, and how much is a mathematical result of the departure of conservative churches.

Stick with me and I’ll give you an answer to that question.  Next time we move from the denominational level to the presbytery statistics and start including pretty pictures with charts and graphs.