— Appendix A — Multivariate Regression Model Outputs Predicting Religious Identity

The table on page 245 contains the output tables for five different multivariate regression models conducted using PRRI’s “2018 American Values Survey.” Each column in the table represents a different model: model one predicts white Christian identity, model two predicts white evangelical Protestant identity, model three predicts white mainline Protestant identity, model four predicts white Catholic identity, and model five predicts white religiously unaffiliated identity.

The variables in the rows are the independent variables—the variables whose influence we are testing. The numbers in rows next to the dependent variable names represent their estimated effects in the analysis, showing how much a one-unit shift in the independent variable changes the dependent variable. A positive number means the probability of being in the dependent variable group (for instance, white evangelical Protestant) goes up as the independent variable increases; a negative number indicates the opposite—that the likelihood of being in the dependent variable group goes down. The numbers in parentheses below each estimate are “robust standard errors” and refer to the uncertainty in the estimate. A quick way to read this table is to look for an asterisk (*** p < 0.01, ** p < 0.05, * p < 0.1), which tells us that the effect is statistically significant; that is, that the effect is distinguishable from 0 (no effect). As a general rule, and unless otherwise specified, I have used the 95 percent confidence level (p < 0.05) as the threshold for significance in the discussion of the findings.

• The Racism Index refers to the composite variable described in chapter 5, with a higher score on the index representing more racist views. A score of 0 (least racist views) is the reference category. The values in the table represent the effect of a score of 1 (most racist views).
• “Party affiliation” was measured in a 5-point scale, with identifying as a Democrat as the reference category.
• “Education” was measured in a 4-point scale, with having no high school diploma as the reference category.
• “Region” was measured across four categories, with living in the Northeast as the reference category.
• “Women” was measured in two categories, with the reference category as men.
• “Age” was measured as a continuous variable running from 18 to 91.
• “Household income” was measured with eighteen categories, from those earning less than $5,000 to those making more than $200,000.
• “Own home” was measured in two categories: owning versus being rented or occupied without payment of cash rent. The latter is the reference category.
• “Church attendance” was measured in six categories, running from those who never attend on the low end to those who attend more than once a week on the high end.
• “Metropolitan area” was measured in two categories: living in a metropolitan area versus living in a nonmetropolitan area. The latter is the reference category.
• The Immigration Index refers to the composite variable described in appendix B, with a higher score on the index representing more anti-immigrant views. A score of 0 (least anti-immigrant views) is the reference category. The values in the table represent the effect of a score of 1 (most anti-immigrant views).

Multivariate Regression Model Outputs Predicting White Christian Identities

DEPENDENT VARIABLE
	Model 1 White Christian	Model 2 White Evangelical	Model 3 White Mainline	Model 4 White Catholic	Model 5 White Unaffiliated
INDEPENDENT VARIABLES
Racism Index (low to high)	3.25***	2.40***	1.51***	2.13***	-0.72
	(0.38)	(0.55)	(0.42)	(0.49)	(0.53)
Party Affiliation
Reference: Democrat
Leaning Democrat	0.14	0.90***	-0.04	-0.19	0.43**
	(0.19)	(0.30)	(0.22)	(0.28)	(0.22)
Independent	-0.59**	0.16	-0.44	-0.45	0.84**
	(0.29)	(0.47)	(0.42)	(0.42)	(0.38)
Leaning Republican	0.18	1.11***	-0.16	-0.25	0.67**
	(0.22)	(0.34)	(0.25)	(0.33)	(0.34)
Republican	0.79***	1.49***	0.00	0.03	-0.90**
	(0.22)	(0.32)	(0.25)	(0.28)	(0.36)
Education Level
Reference: No high school diploma
High school graduate	0.11	0.21	0.86*	-0.66	0.60
	(0.30)	(0.42)	(0.46)	(0.42)	(0.49)
Some college	0.25	0.56	0.78*	-0.68*	0.84*
	(0.28)	(0.38)	(0.46)	(0.40)	(0.47)
BA degree or higher	0.26	0.22	0.93*	-0.55	1.07**
	(0.30)	(0.40)	(0.49)	(0.42)	(0.48)
Region
Reference: Northeast
Midwest	0.04	0.20	0.83***	-0.93***	0.38
	(0.20)	(0.29)	(0.24)	(0.23)	(0.28)
South	-0.77***	0.62**	0.26	-1.63***	-0.23
	(0.20)	(0.28)	(0.25)	(0.24)	(0.27)
West	-1.02***	0.40	-0.02	-2.12***	0.15
	(0.21)	(0.31)	(0.27)	(0.27)	(0.28)
Reference: Men
Women	0.43***	0.19	0.39***	0.07	-0.52***
	(0.13)	(0.18)	(0.15)	(0.18)	(0.18)
Age (in years)	0.02***	0.00	0.02***	0.02***	-0.01*
	(0.00)	(0.01)	(0.00)	(0.01)	(0.01)
Household income (low to high)	0.04**	-0.03	0.02	0.07***	0.03
	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)
Reference: Not homeowner
Own home	0.47***	0.06	0.43**	0.40*	0.48**
	(0.16)	(0.20)	(0.18)	(0.23)	(0.22)
Church attendance (low to high)	1.10***	2.79***	-1.11***	0.17	-6.93***
	(0.21)	(0.31)	(0.24)	(0.23)	(0.54)
Reference: Nonmetropolitan area
Metropolitan area	-0.55***	-0.74***	-0.28	0.16	-0.39
	(0.20)	(0.24)	(0.21)	(0.27)	(0.27)
Immigration Index (low to high)	-0.44	-0.66	0.20	-0.50	-0.05
	(0.56)	(0.83)	(0.63)	(0.77)	(0.78)
Constant	-4.08***	-5.78***	-4.91***	-3.48***	0.05
	(0.54)	(0.76)	(0.66)	(0.66)	(0.72)
Observations	2,337	2,337	2,337	2,337	2,337
SOURCE: PRRI, AMERICAN VALUES SURVEY, 2018.