The Baylor Religion Survey

The Baylor Religion Survey is an in-depth statistical analysis of Americans' beliefs and practices concerning God and religion, as determined by extensive 2005 polling data. The initial findings were published in September 2006 by the Baylor Institute for Studies of Religion at Baylor University. A Baylor University news article about the initial publication appears here. An independent news article about it may be read here, at the website of The (Baltimore) Sun.

The initial published report, American Piety in the 21st Century: New Insights to the Depths and Complexity of Religion in the U.S., can be downloaded in PDF form by visiting this page or more directly by clicking here. Many statistical details (but not all) are available here, at the Association of Religious Data Archives.

A blog post which questions some of the statistical methodology that was used can be read here, at the "Amor et Labor" blog of Kletos Sumbulos.


The initial finding that is supposedly most surprising to many observers is that Americans are more religious and less secular than other recent surveys have found:

Are Americans losing their religion? Prior national studies with questions on religion, such as the General Social Survey and National Election Study, show an increase in the percent of the population with no religion over the past quarter century ... This growth in “religious nones” is often used by academics and the press to indicate growing secularization in the United States. But are Americans really that detached from organized religion?

Most surveys determine the religious affiliation of respondents by asking them to select their religious family or denomination from a list. This has become increasingly problematic over the years as more and more Americans are losing a strong denominational identity. The rising number of non-denominational congregations as well as congregations that minimize their denominational ties compound the problem. The declining importance of denomination, however, does not mean that religion itself is on the wane. Rather, Americans may simply be more likely to connect with religion at the local level ... To detect religious affiliation today, it is time to look beyond denomination. In addition to presenting respondents with a standard list of denominations, the Baylor Religion Survey asks respondents to give the name and address of their place of worship. Combining these three measures of religious belonging enables us to more thoroughly and accurately sort persons into broader religious traditions.

This method finds that barely 1 in 10 Americans (10.8%) says he or she is not affiliated with a broad religious tradition, much less a specific denomination. Other recent surveys such as the General Social Survey and National Election Study have put the (apparently rising) unaffiliated figure about 3 percentage points higher.

Evangelical Protestant, Mainline Protestant, Catholic, Black Protestant, Jewish, or Other (mainly Buddhist, Christian Science, Church of Jesus Christ of Latter-day Saints [Mormon], Hindu, Jehovah’s Witnesses, Muslim, Orthodox [Eastern, Russian, Greek], and Unitarian Universalist were the broad traditions identified in the survey.

Only 5.2% of Americans emerge from the poll's questions as atheists who are "certain that God does not exist." The other unaffiliated respondents — a roughly equal number of them, in fact — apparently have some belief in a higher power, even if they do not identify with any broad religious tradition.


The survey result which I personally find most provacative is the "Four Gods" hypothesis. The researchers claim (pp. 26ff.) they were able to derive from the poll's responses four meaningful groupings of non-atheists:

1. Those who believe in an "Authoritarian God" (31.4%)
   
2. Those who believe in a "Benevolent God" (23%)
3. Those who believe in a "Critical God" (16%)
4. Those who believe in a "Distant God" (31.4%)

They labeled these believer groups, for obvious alphabetic reasons, Types A, B, C, and D. The four groups were in turn derived from two "clear and distinct dimensions of belief in God":

1. God’s level of engagement – the extent to which individuals believe that God is directly involved in worldly and personal affairs.
   
   
2. God’s level of anger – the extent to which individuals believe that God is angered by human sins and tends towards punishing, severe, and wrathful characteristics.

Here are more details on the four types:

• Type A: Authoritarian God: Individuals who believe in the Authoritarian God tend to think that God is highly involved in their daily lives and world affairs. They tend to believe that God helps them in their decision-making and is also responsible for global events such as economic upturns or tsunamis. They also tend to feel that God is quite angry and is capable of meting out punishment to those who are unfaithful or ungodly.
  
  
• Type B: Benevolent God: Like believers in the Authoritarian God, believers in a Benevolent God tend to think that God is very active in our daily lives. But these individuals are less likely to believe that God is angry and acts in wrathful ways. Instead, the Benevolent God is mainly a force of positive influence in the world and is less willing to condemn or punish individuals.
  
  
• Type C: Critical God: Believers in a Critical God feel that God really does not interact with the world. Nevertheless, God still observes the world and views the current state of the world unfavorably. These individuals feel that God’s displeasure will be felt in another life and that divine justice may not be of this world.
  
  
• Type D: Distant God: Believers in a Distant God think that God is not active in the world and not especially angry either. These individuals tend towards thinking about God as a cosmic force which set the laws of nature in motion. As such, God does not “do” things in the world and does not hold clear opinions about our activities or world events.

Type A ("Authoritarian God") is relatively big on both "engagement" and "anger." Type B ("Benevolent God") emphasizes "engagement" but minimizes "anger." Type C ("Critical God") de-emphasizes "engagement" with this world's affairs but attributes to a remote deity much "anger" at the way things are going. Type D ("Distant God") de-emphasizes both "engagement" and "anger."

This looks like a valuable contribution to the understanding of American religiosity, but a blog post which questions some of the methodology that was used to derive it can be read here, at the "Amor et Labor" blog of Kletos Sumbulos. The main complaint is that the questionnaire (which is shown at the end of the American Piety PDF) may have been constructed in such a way that the two reported dimensions of "engagement" and "anger" were bound to show up in the results with so-called statistical significance.


There are also worries about whether a "cluster analysis" would confirm the four belief types as being meaningfully separate from one another ... again, statistically speaking.

And in a comment I made to the post I said that, per a Wikipedia article on the statistical technique the survey used, called factor analysis: "Interpreting factor analysis is based on using a 'heuristic,' which is a solution that is 'convenient even if not absolutely true' ... More than one interpretation can be made of the same data factored the same way."

In other words, the "Four Gods" hypothesis may be more statistically "convenient" than "absolutely true."


Note also that there is no way to tell from the statistics given in the initial American Piety report how much, in believers' reported notions of God, God's "engagement" and "anger" pertain to the world's affairs in general, rather than the believers' own personal lives. The questionnaire does make such distinctions, but the "Four Gods" groupings do not.

Moreover, I see no way to tell whether respondents who feel God is, for example, "directly involved in my affairs" (see Question 22i) mean that God's personal involvement is at a subjective, spiritual level (as in "God comforts me" or "God seems to be silent when I pray to him") or at the level of objective events (as in "God didn't arrange for me to get the new job I wanted" or "Thanks to God, I finally found the right mate").

To me, that's a crucial distinction in believers' modes of piety. I, for instance, feel God is engaged with me and other believers at a personal, spiritual level but does not actually "make things happen" in the physical world ... or keep them from happening. I'm aware that other believers feel very strongly that God does exactly that. There's no apparent way to respond to the questions in the survey that would make such a distinction clear.