Warning: This is another one of my posts where the analysis is going to get really geeky really fast.  So be it — just jump to the end for the bottom line if your eyes start to glaze over.

In working on a couple of other current issues I decided that for my own edification I needed to find a metric for the theological viewpoint of the membership, not the leadership, of the Presbyterian Church (U.S.A.) and how that might be changing with time.

One motivation for this is the contention that the PC(USA) is preferentially losing conservative members.  I have previously commented that 1) the total membership loss is much higher than what can be attributed to congregation level realignment out of the PC(USA) and that 2) change in presbytery level membership can not be correlated to leadership theological views.  I had been holding the position that membership loss in the PC(USA) is broadly across the theological perspectives.  I may be wrong about that.  Here is an analysis of a different data set…

I looked at the last five Presbyterian Panel surveys: 1996, 1999, 2002, 2005, 2008.  These are the initial surveys of each new panel which serves as the “sample population” for the PC(USA) for the next three years.  That is, the 1996 survey was for the 1997-1999 panel.

In those surveys I found five questions that were asked the same way in all five surveys that pertain directly to doctrinal issues giving a direct measure of an individual’s theological viewpoint.  The five questions are:

1. Which one of the following terms best describes your current stand on theological issues?
2. All the world’s different religions are equally good ways of helping a person find ultimate truth
3. The only absolute truth for humankind is in Jesus Christ
4. Only followers of Jesus Christ can be saved
5. There is a life beyond death

I really wish the question about the respondent’s view of the Bible had been asked the same way every time because that would also have given a good perspective on the individual’s viewpoint.  And there are a couple other questions that appear in every survey that could be considered theological indicators as well, such as “Have you ever tried to encourage someone to believe in Jesus Christ or to accept Him as a personal savior?” but these are more about spiritual practices and I thought the questions could be answered either way across the theological spectrum so were not as good of indicators..  (For the record, on this question of accepting Jesus Christ as personal savior it is very close to 60% “yes” and 40% “no” in all five surveys with no trend or statistical variation.)

Other technical details I need to mention:  The margin of error is reported as +4%.  I will only be looking at the “members” category but as I opined before 57% of “members” are ordained officers of the church and for elders they are those not currently serving on session.

Now, the first shall be last and the last shall be first so let me deal with the fifth one at the beginning.  This is easy – over the five surveys there is virtually no change with always 84-86% who agree or strongly agree, 12-14% who are not sure, and 1-3% who disagree or strongly disagree.  I would also note that there was a statement on four of the five surveys (missing in 1999) that “Jesus will return to earth some day.” The last three surveys are indistinguishable at 66-69% agree or strongly agree, 24-27% not sure, and 6-7% disagree or strongly disagree.  The first survey was a bit higher for the two agree categories (75%) with equal drops (3-4% each) in the not sure and combined disagree.  For these statements there is no indicator of change with time.

For the statement “all the world’s different religions are equally good ways of helping a person find ultimate truth” there is an interesting statistically significant variation, but not a trend.  (Note that on all these tables I have added the “combined agrees” and “combined disagrees” categories to simplify graphing and they show up as “all agrees” and “all disagrees” on the chart.)
 

All the world’s different religions are equally good ways of helping a person find ultimate truth.	1996	1999	2002	2005	2008
Strongly Agree	9	7	8	9	11
Agree	31	28	27	23	26
Combined Agrees	40	35	35	32	37
Not Sure	18	18	19	25	19
Disagree	25	29	28	24	24
Strongly Disagree	18	18	18	20	19
Combined Disagree	43	47	46	44	43

It is not clear what happened here in the 2005 survey where the “agree” dropped and the “not sure” jumped up. Except for that point the responses to this question in the other surverys are all statistically indistinguishable with no clear suggestion of a trend.

When it comes to the statements about the significance of Jesus Christ, and that is not the significance in the statistical sense, there are clear trends of the sample populations moving away from the orthodox or conservative position.  The two statements are 1) “The only absolute truth for humankind is in Jesus Christ” and 2) “Only followers of Jesus Christ can be saved.”  And yes, I am taking the two agree categories as reflecting the conservative position.  Here are the numbers…

The only absolute truth for humankind is in Jesus Christ.	1996	1999	2002	2005	2008
Strongly Agree	43	46	41	39	38
Agree	29	27	28	24	21
Combined Agrees	72	73	69	63	59
Not Sure	1 7	15	17	20	20
Disagree	8	9	10	12	13
Strongly Disagree	3	2	3	5	7
Combined Disagree	11	11	13	17	20

Only followers of Jesus Christ can be saved.	1996	1999	2002	2005	2008
Strongly Agree	27	26	23	26	25
Agree	19	20	20	15	14
Combined Agrees	46	46	43	41	39
Not Sure	25	25	23	25	25
Disagree	20	20	23	21	19
Strongly Disagree	8	10	11	14	17
Combined Disagree	28	30	34	35	36

In graphical form (and yes, the first graph is the “absolute truth” question not the “ultimate truth” question above)

In each of these there is an apparent trend with the number of those in some agreement with the statement decreasing with time, the number disagreeing increasing, and those not sure mostly to very constant.

Finally, we have the survey question asking each respondent to self-identify their theological viewpoint.  I am not a big fan of the “conservative” and “liberal” labels but I have used it throughout this post because those were the options given in the survey for this question:

Which term best describes your current stand on theological issues?	1996	1999	2002	2005	2008
Very Conservative	8	5	5	6	6
Conservative	31	33	33	35	28
Combined Conservatives	39	38	38	41	34
Moderate	48	47	43	40	41
Liberal	11	12	14	14	18
Very Liberal	3	3	4	5	7
Combined Liberals	14	15	18	19	25

And graphically

It is interesting that in the first four surveys the shift seems to be from the moderates to the liberals with the conservatives fairly constant and then in the last survey group the liberals increase and the conservatives drop.  While interesting, I am hesitant to put too much weight on that last point because we saw the 2005 “bump” on the different religions question was a one-survey event.  In three years we will see if it is a new trend.

Now having laid the data out there, what does all this mean?  First, and to my surprise, there was more of a shift than I expected in these indicators from conservative to liberal.  The view of the denomination that it is growing more liberal may hold up. But what is actually changing?

One interpretation is to say that the changes in the panels represents the changes in the members of the denomination as a whole and the changes in attitudes in the survey group is explained by those joining and leaving the PC(USA).  This is still a wildly under-determined problem (that is mathematical jargon) so many different distributions of those joining and those leaving would produce this result.  For instance, you could say that those leaving broadly represent the membership but those joining are more liberal.  Or you could explain it the other way, that those joining are broadly representative and those leaving are more conservative.  And of course many different combinations in between.

The other explanation of course is that people’s minds are changing about these statements.  Rather than members with fixed opinions moving in and out of the denomination we could say that there are people remaining in the denomination that are changing their viewpoint over time.

And these are two possible end-members and the best interpretation is probably some combination of the two and the precise balance between them would require tracking over time or questions specifically designed to test for time-variability of viewpoint.

We can narrow the possible range a little bit by looking at how this breaks down for the self-identified categories for each panel year.  I do realize that the total membership number includes Ministers of Word and Sacrament as well but they represent about 1% of the total membership and so I am going to consider the effect too minor to worry about correcting for this back of the envelope calculation.  Here is how the membership numbers would be split out based on the declared theological viewpoint of the sample population:

Year	Total Membership	Conservative Members	Moderate Members	Liberal Members
1996	2,631,466	1,026,272	1,263,104	368,405
1999	2,560,201	972,876	1,203,294	384,030
2002	2,451,969	931,748	1,054,347	441,354
2005	2,313,662	948,601	925,465	439,596
2008	2,140,165	727,656	877,468	535,041

Looking at the numbers we can see that the conservative and moderate declines can, with one exception (con
servative 2005), be explained within the denominational membership loss.  The reverse is true for the liberal component — with the exception of 2005 all the other changes show an increase in the absolute, not just the relative, numbers.   But none of these changes can be attributed to just those leaving or joining the church.  The volume of the turnover is significantly larger than the actual net loss so each group must have members added and members lost and what is listed here is the net.  (For specifics consider the 2008 membership numbers – the church had 103,488 members join, and 138,768 leave (not counting deaths).  That represents a 5% annual turnover, or to put it another way, every 20 years the PC(USA) is a whole new church.  More on that another time.)

Finally, you could speculate that the results reflect the way the respondents thought they should answer, either because of what they think the research group wants or because of how they see themselves even if their basic theological perspective has not changed.

So whether by membership turnover or change in opinion there is evidence that over the last 14 years the PC(USA) is indeed becoming a more liberal denomination at the level of the total membership. 

Finally, a note about a paradox in this data:  “Conventional wisdom” says that younger generations are more liberal, more questioning, more tolerant of other viewpoints like those the “truth” and “only way to salvation” questions ask.  Does that mean that the changing viewpoints seen in the survey questions is due to an influx of younger members?  Unfortunately not — In the 12 years between the 1996 panel to the 2008 panel the median age of the panel members has crept up from 55 to 60 years old.  The interpretation is left as an exercise for the reader.