Are religious people more ethnocentric?

Does loyalty to one group you belong to predict loyalty to other groups?  I suspect that religious people are also more likely to value their ethnicity.  The General Social Survey (GSS) asked people how important their ethnicity is to them with answers ranging from unimportant (1) to very important (4). 

I categorized people as not religious if they attend religious services no more than once per year. All others I lumped into the religious group. The means for thinking ethnicity is important to you are listed below--the means for the religious in parentheses, the means for the irreligious without parentheses (sample size = 2,110)

"My ethnicity is important to me"-- means

American Indian  3.50  (3.50)
Black  3.47  (3.53)
Chinese  3.40  (3.75)
Mexican  3.22  (3.47)
Puerto Rican  3.19  (3.60)
Italian  2.85  (2.72)
Jewish  2.76  (3.26)

Total  Sample  2.60  (2.84)

Scottish  2.55  (2.44)
Irish  2.45  (2.60)
German  2.35  (2.50)
Swedish  2.29  (2.67)
Russian  2.25  (2.95)
English/Welsh  2.21 (2.50)
Polish  2.13  (2.67)

For most of the ethnic groups, ethnocentrism is higher for the religious group. The differences are generally not large but look, for example, at how religious Swedes are almost as ethnocentric as secular Jews.

The difference at the extremes is large: the gap between religious Chinese Americans on the high end and the irreligious Poles on the bottom is well over one standard deviation. The typical Chinese person who goes to church says his ethnicity is 'moderately' or 'very important.' Compare that to the average Polish American who is not religious: he says his ethnicity is 'slightly important.'

Group solidarity among whites

MIDUS Study participants were asked how closely do they identify with their 1) ethnicity and 2) race; how much do they prefer to be with their own 1) ethnicity and 2) race; how important is it to marry someone of the same 1) ethnicity and 2) race; and how closely do they identify as an American.

I calculated the Pearson correlations for each possible pair of questions for whites only (sample size = 4,075).  All of the 21 correlations are positive, and the average is .30 which indicates a moderate, positive relationship. The strongest correlations are among the ethnicity items (around .60), and also among the race items. The weakest correlations are between the American question and other measures. The vast majority of study participants identify strongly as Americans so the lack of variation reduces the correlations.

Bottom line: whites either tend identify with groups at all levels or no groups at any level. To put it in exaggerated terms, ethnicists are racialists are patriots. Individualists are individualists are individualists.  

More on racial identification and well-being

As a follow-up to the last post, I wanted to check the race-feeling of other groups. For whites and blacks, I found that measures of psychological well-being were basically unrelated to identifying with your race. Both groups, however, vary little on racial identification: most identify with their race. American Indians, by contrast, identify with their race less. Here is their distribution:

How closely do you identify with your race?--Percent

Very close 37.7
Somewhat close 35.1
Not very close 10.4
Not at all close 16.9

Still skewed, but believe or not, American Indians are less ethnocentric than whites. Here are the correlations between identity and psychological measures:

Correlations (sample size = 77)

Sad .15
Hopeless .06
Worthless .16
In good spirit -.17
Peaceful -.18
Feel belonging -.08
Feel pride .09
Feel confident .04

The correlations are stronger, but they suggest that greater racial solidarity is associated with less well-being.