— 5 — Mapping The White Supremacy Gene in American Christianity

In previous chapters, I’ve discussed the prominent, historical role white Christians, churches, and other institutions such as seminaries have played in creating and sustaining white supremacy in America. But in order to assess the impact of this history on the present, we’ll need some new tools that can give us a reliable understanding of the current relationship between white supremacy and Christian identity among whites. We’ll also need a systematic approach that looks for any imprint that centuries of accommodation to white supremacy may have left on contemporary white Christians.1 Specifically, I aim to answer these central questions: How prevalent are racist and white supremacist attitudes among white Christians today? To the extent that they exist, are these attitudes merely incidental to, or have they come to be, over time, actually constitutive of white Christian identity? And is this relationship limited to white evangelicals or white Christians in the South, or do these attitudes also persist among white mainline and white Catholic Christians outside that region? These are sensitive, difficult questions to sort out, and this chapter proceeds one step at a time.

The Living Legacy of Slaveholding Among White Americans

When people think about the contemporary effects of slavery, they typically consider external effects such as the continued economic and social inequalities between black and white Americans, which are indeed stark. But in the book Deep Roots: How Slavery Still Shapes Southern Politics, political scientists Avidit Acharya, Matthew Blackwell, and Maya Sen demonstrate something less acknowledged: the remarkably enduring impact of slavery on how contemporary white people think, feel, and act today.

In a rigorous statistical analysis linking county-level slave ownership from the 1860 US Census and public opinion data collected between 2006 and 2011 by the Cooperative Congressional Election Study (CCES), a large-scale national survey of the American electorate conducted by nearly forty universities, they find that whites residing in areas that had the highest levels of slavery in 1860 demonstrate significantly different attitudes today from whites who reside in areas that had lower historical levels of slavery: (1) they are more politically conservative and Republican leaning; (2) they are more opposed to affirmative action; and (3) they score higher on questions measuring racial resentment.2 After accounting for a range of other explanations and possible intervening variables, Acharya and his colleagues conclude that “present-day regional differences, then, are the direct, downstream consequences of the slaveholding history of these areas.”3

One remarkable feature of this research is that the results are sensitive at the county level, not just within the South but also within individual states. While on average 36.7 percent of the population in southern counties was enslaved in 1860, there was a wide variation even within the same state, as two examples from Arkansas demonstrate. In the northwestern county of Benton, which was less suited to cotton farming, only 4.1 percent of the population was enslaved. But in the southeastern county of Chicot—where my parents lived while I was in graduate school, with a backyard emptying into Lake Chicot and a front yard emptying into cotton fields as far as you could see—81.4 percent of the population was enslaved.4

Even smaller variations in slave-owning percentages produce measurable attitudinal differences today. For example, moving from Alabama’s Clay County to Barbour County (associated with about a 25 percent increase in slave owning in 1860) is associated today with a 5.6 percent decrease in the share of whites who identify as Democrats, a 7.8 percent decrease among whites in support for affirmative action, and a 4.8 percent increase in attitudes reflecting racial resentment.5

This research demonstrates that the deep racial prejudice that was created by a slaveholding society is still measurably present in the contemporary South, and that this relationship is not just correlational but causal. Even when comparing neighboring counties that differ in slaveholding percentages and adjusting for state-by-state variation and cotton-farming suitability, the relationship between the level of slaveholding in 1860 in a county and its current political and racial conservatism remains robust. As Acharya and his coauthors summarize the findings, “It’s not simply that more conservative people live in these areas—these are more conservative areas because of their past.”6

These findings are provocative. But for our purposes, they have two limitations. First, the authors give scant attention to religion in their analysis. Second, by design, the study is limited to measuring the effects of southern slaveholding. To answer the broader questions I am raising about white Christian identity, I’ll draw on recent national public opinion surveys by PRRI in order to bring religion into the analysis more deliberately and widen the field of vision beyond the South.

The Distinctive Racial Attitudes of White Christians Today

One of the most remarkable, consistent findings in contemporary public opinion data is the chasm between two groups who otherwise share both geographic proximity and a common evangelical religious orientation: black Protestants and white evangelical Protestants. On a range of religious measures—belief in God, belief in a literal heaven and hell, belief in a literal interpretation of the Bible, frequency of church attendance and prayer—black and white evangelical Protestants are largely aligned. But as these religious beliefs and behaviors are refracted through the lens of race, they produce starkly divergent opinions and behaviors in political space.7

More recently, as the ranks of religiously unaffiliated Americans have grown—from single digits in the 1990s to a quarter of Americans today—the gap between white Christians of all kinds and white religiously unaffiliated Americans has become another defining feature of the religious landscape. As we proceed through the analysis below, we’ll keep all of these groups in view: white evangelical Protestants, white mainline Protestants, white Catholics, and religiously unaffiliated whites. For comparison, we’ll also examine the views of these white groups alongside the views of African American Protestants.

Attitudes about African Americans and racism can be challenging to assess through public opinion surveys. The biggest hurdle is that a researcher obviously cannot get accurate results from asking respondents outright whether they are white supremacists or racists. Even with online surveys, where participants complete surveys privately on their own devices and with assurances of anonymity, many may be reluctant to reveal their true views. Or they may privately hold white supremacist or racist views that they themselves would not identify with those labels.

The slipperiness of individual questions on the sensitive topic of race can be seen in the following example from PRRI’s “2018 American Values Survey.”8 White Christians were asked to say how warmly they feel toward African Americans as a group on what social scientists call “a feeling thermometer”: a scale ranging from 1 to 100, where 1 is cold, and 100 is warm. White mainline Protestants (mean = 65) and white Catholics (mean = 66) on average report views close to the general population (mean = 67), while white evangelical Protestants report even warmer feelings (mean = 71).9 But when white Christian attitudes are illuminated by more specific questions about the symbols of white supremacy, about economic and social inequality between African Americans and whites, or about unequal treatment of African Americans in the criminal justice system, white Christian attitudes appear considerably less warm, and the differences between white Christians and other Americans are revealed in stark relief.

The Confederate Flag and Confederate Monuments

Not surprisingly, attitudes about what the Confederacy symbolizes today are one of the most powerful differentiators among these groups. PRRI’s 2019 “American Values Survey” found that 86 percent of white evangelical Protestants, along with 70 percent of white mainline Protestants and 70 percent of white Catholics, believe that the Confederate flag is more a symbol of southern pride than of racism. By contrast, only 41 percent of white religiously unaffiliated Americans and 16 percent of African American Protestants agree; approximately six in ten religiously unaffiliated whites and three-quarters of African American Protestants report that they see the Confederate flag mostly as a symbol of racism.10

On a similar question, a 2018 PRRI survey found that more than eight in ten white Christians overall—including 85 percent of white evangelical Protestants, 85 percent of white mainline Protestants, and 80 percent of white Catholics—say that Confederate monuments are more a symbol of southern pride than of racism. These views are shared by only 54 percent of religiously unaffiliated whites and a mere 24 percent of their fellow African American Protestants.11

Image

FIGURE 5.1 Perception of Confederate Symbols, by Religious Affiliation

Source: PRRI, American Values Surveys, 2019 (Flags), 2018 (Monuments).

The Killing of African American Men by Police

Current events related to racial injustice also produce differing opinions among white Christians than among religiously unaffiliated whites. Nearly two-thirds (64 percent) of white Christians overall believe the killings of African American men by police are isolated incidents rather than part of a broader pattern of how police treat African Americans. There is some daylight here between white evangelicals (71 percent), white Catholics (63 percent), and white mainline Protestants (59 percent), but each group has solid majorities agreeing with this statement. And there is a 26-percentage-point gap between white Christians overall and religiously unaffiliated whites (38 percent agree they are isolated incidents) and a nearly 50-percentage-point gap between white Christians and African American Protestants (15 percent agree).

There is an even larger attitude gulf related to protests about this issue, such as NFL players kneeling during the national anthem. More than seven in ten (72 percent) white Christians overall believe that professional athletes should be required to stand during the national anthem at sporting events, a view held by only 34 percent of religiously unaffiliated whites and only 21 percent of African American Protestants. Again, all white Christian subgroups have majorities agreeing with this statement: 81 percent of white evangelicals, 71 percent of white Catholics, and 65 percent of white mainline Protestants.12

Notably, white Christian objections are not due to a misunderstanding of the nature of these protests by African American players. PRRI’s January 2018 sports poll found that more than seven in ten white Americans correctly identified these athletes’ actions as protesting police violence against African Americans.13

Perceptions of Structural Injustice

White Christians also stand out as a group on questions related to structural injustice and perceived barriers to black social mobility. More than three-quarters of white Christians overall—including 83 percent of white evangelicals, 75 percent of white Catholics, and 71 percent of white mainline Protestants—believe that racial minorities use racism as an excuse for economic inequalities more than they should. While these views are also shared by 52 percent of religiously unaffiliated whites, the gap between them and the closest white Christian subgroup is nearly 20 percentage points. Only 30 percent of black Protestants agree. Similarly, two-thirds of white Christians overall, including similar numbers of all white Christian subgroups, agree that black Americans should be able to overcome prejudice and “work their way up without any special favors.” Only about half of religiously unaffiliated whites and roughly three in ten African American Protestants agree.14

Perhaps most fundamentally (see figure 5.2), more than six in ten white Christians overall disagree with this basic statement: “Generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class.” Sixty-seven percent of white evangelical Protestants, 62 percent of white mainline Protestants, and 57 percent of white Catholics disagree with this sentiment, compared with only 40 percent of religiously unaffiliated whites. Only 31 percent of black Protestants disagree with this statement, while more than two-thirds agree.15

Across a range of questions, the overall pattern that emerges is abundantly clear. On the one hand, white Christians explicitly profess warm attitudes toward African Americans. At the same time, however, they strongly support the continued existence of Confederate monuments to white supremacy and consistently deny the existence not only of historical structural barriers to black achievement but also of existing structural injustices in the way African Americans are treated by police, the courts, workplaces, and other institutions in the country. And, notably, Christian affiliation remains a powerful differentiator among whites, with differences between white Christians and religiously unaffiliated whites running from 20 to nearly 40 percentage points across these questions. In every case, it is religiously unaffiliated whites who stand closer than white Christians do to their African American Christian brothers and sisters.

Image

FIGURE 5.2 Impact of Structural Injustice on Black Economic Mobility

Source: PRRI, American Values Survey, 2018.

Foreign Protectionism: White Christians and Negative Attitudes About Immigrants

We find similar attitudinal patterns across questions related to resisting demographic changes in the country and to protecting a perceived “American way of life” from changes by foreign influences. The largest gulf between white Christians and religiously unaffiliated whites appears on a general question about cultural protectionism. Approximately two-thirds of white Christians overall—including 71 percent of white evangelical Protestants, and 62 percent of both white Catholics and white mainline Protestants—believe that the American way of life needs protecting from foreign influence. This attitude puts them strongly at odds with religiously unaffiliated whites and with African American Protestants, among whom less than half perceive such a cultural threat (38 percent and 48 percent, respectively).

There is also notable support among white Christians for President Donald Trump’s signature immigration policies. While most of the policy debates have focused on illegal immigration, one of the strongest areas of agreement among white Christians is actually in the area of limiting legal immigration. Nearly two-thirds (65 percent) of white Christians overall, including nearly identical numbers of white Christian subgroups, favor stricter limits on the number of legal immigrants coming to the United States. That level of support is 20 to 30 percentage points higher than support among religiously unaffiliated whites (37 percent) or African American Protestants (45 percent).

Similarly, white Christians are also strongly behind Trump’s Muslim travel ban policy. Nearly two-thirds of white Christians favor temporarily preventing people from some majority Muslim countries from entering the United States. White evangelical Protestants are particularly supportive of this policy (72 percent), but 63 percent of mainline Protestants and 60 percent of white Catholics also back it. By contrast, only 35 percent of religiously unaffiliated whites and 34 percent of African American Protestants support this policy.

Finally, majorities of all white Christian subgroups support Trump’s signature project, building a wall on the southern border. White evangelicals stand out for their particular enthusiasm: two-thirds of white evangelical Protestants favor building a wall along the US border with Mexico, including nearly four in ten who strongly favor this policy. Fifty-six percent of white Catholics and 52 percent of white mainline Protestants also support this policy. Only 28 percent of religiously unaffiliated whites and African American Protestants agree, reflecting a 30-percentage-point difference compared with white Christians overall (58 percent).

Compared with the questions about African Americans, there is more variation here among white Christian subgroups, with white evangelicals generally staking out more anti-immigrant positions than either white mainline Protestants or white Catholics. But even with this variation, the differences between white Christians overall and religiously unaffiliated whites are on average a robust 30 percentage points, and it remains true that religiously unaffiliated whites are closer than white Christians to the attitudes of Christians of color. Religiously unaffiliated whites are far less likely than their Christian counterparts to perceive demographic and cultural changes as negative or to support policies designed to protect the country from such perceived external threats.

What Role Do Racial Attitudes Play in Structuring White Christian Identity?

The correlations in the descriptive statistics above are consistent and clear. But in order to assess what role racist attitudes independently play in structuring white Christian identity, we have to employ some additional statistical tools. This more rigorous analysis proceeds in two steps. First, we want to ensure that the particular questions we analyzed above are not peculiar in their results, and that they have not been chosen in a way that produces a skewed set of findings. To address this concern, I have integrated these questions into a broader Racism Index composed of a total of fifteen separate questions related to this history of white supremacy and perceptions of African Americans. By combining these questions into a single scale, we can ensure that we are measuring a more general underlying sentiment rather than what might be an outlier response to the specifics of a single question.

Second, I test this Racism Index for the possibility that the correlations between racial attitudes and white Christian identity are explained by some other intervening variable. For example, perhaps white evangelicals hold negative attitudes about African Americans not because they are evangelical Christians but because they overwhelmingly identify with and support the Republican Party, which has consistently opposed civil rights legislation since the 1960s, or because they reside overwhelmingly in the former states of the Confederacy, or because of some other demographic attribute. If that is the case, the correlation between holding racist attitudes and white Christian identity may be coincidental or even spurious. But if this relationship holds up in statistical models that account for these other possible explanatory variables, we gain confidence that this relationship is real—in other words, that holding racist attitudes is directly and independently linked to white Christian identity.

I realize that, for nontechnical readers, references to “statistical models” may prompt a search for the beginning of the next chapter. But we need these tools if we want to know whether racist attitudes are important drivers of white Christian identity, or the reverse—whether white Christian identity is an important driver of racist attitudes—not just in the past but also in the present. Generally speaking, these models are the way social scientists determine the strength of relationships between attitudes, behaviors, and concepts. Humans are complex animals, and it is impossible to determine every element that explains our political and social attitudes. However, statistical analyses allow us to explain, in this case, white Christian identity, and to isolate the impact of holding racist attitudes relative to other factors such as partisanship, socioeconomic status, or where one lives. While we’ll need a working knowledge of these concepts for the discussion that follows, I’ve sequestered most of the technical discussion and the math to the endnotes and appendices. For nontechnical readers who may still struggle with the statistical analysis, I have included a summary of the findings, sans data, at the conclusion of the chapter.16

A more general way of thinking about this exercise is to imagine it as an attempt to sociologically map the genome of white Christianity to see whether white supremacist attitudes have become integrated into its DNA as part of what it means to be a white Christian in America. If the initial correlations we see between white supremacist attitudes and white Christianity cannot be explained away by other factors, white Christians have some serious soul-searching to do.

The Racism Index

The table below lists the fifteen questions included in the Racism Index.17 These questions cover a lot of ground: attitudes about Confederate symbols; racial inequality and African American economic mobility; racial inequality and the treatment of African Americans in the criminal justice system; and general perceptions of race, racism, and racial discrimination. For the analysis below, I have combined these fifteen questions into a single composite index that is calibrated from 0 to 1, with 0 representing the least amount of racist attitudes across all questions and 1 representing the highest amount of racist attitudes across all questions. The breadth of this scale, along with the fact that these questions produce an internally consistent scale, give us a high degree of confidence that we are measuring a broader underlying sentiment.18

The Racism Index • Individual Question Wording

Confederate Symbols

Do you see monuments to Confederate soldiers more as symbols of southern pride or more as symbols of racism?

Just your opinion: What should be done with Confederate monuments that are currently standing on public property such as statehouses, county courthouses, public universities or city parks? Should they be:

Racial Inequality and African American Economic Mobility

Generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class.*

It’s really a matter of some people not trying hard enough; if blacks would only try harder, they could be just as well off as whites.

Irish, Italians, Jews, and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.

Over the past few years, blacks have gotten less than they deserve.*

Racial minorities use racism as an excuse more than they should.

White people in the U.S. have certain advantages because of the color of their skin.*

Racial Inequality and the Treatment of African Americans in the Criminal Justice System

Do you think recent killings of African American men by police are isolated incidents, or are they part of a broader pattern of how police treat African Americans?

Professional athletes should be required to stand during the national anthem at sporting events.

A black person is more likely than a white person to receive the death penalty for the same crime.*

Perceptions of Race, Racism, and Racial Discrimination

I am fearful of people of other races.

Racial problems in the U.S. are rare, isolated situations.

I am angry that racism exists.*

Today discrimination against whites has become as big a problem as discrimination against blacks and other minorities.

* Note: Response options for these questions were reverse-coded so that they run the same ideological direction as other questions in the scale.

SOURCE: PRRI, AMERICAN VALUES SURVEY, 2018.

Image

FIGURE 5.3 Distribution of Racism Index Scores among White Religious Subgroups

Note: Median lines shown.

Source: PRRI, American Values Survey, 2018.

Analysis of the composite Racism Index confirms the general pattern from the individual question analysis above: white Christians overall are more likely than white religiously unaffiliated Americans to register higher scores on the Racism Index, and the differences between white Christian subgroups (white evangelical Protestants, white mainline Protestants, and white Catholics) are largely differences of degree rather than kind. Figure 5.3 shows the median scores with vertical lines—the score at which there are equal numbers of group members falling above and below—and the distribution of attitudes on the Racism Index among each white Christian subgroup, along with the comparison group of whites who are religiously unaffiliated.

Even a brief glance at the median scores for each group shows the similarity of attitudes among each white Christian group. Not surprisingly, white evangelical Protestants have the highest median score (0.78) on the Racism Index. But the median scores of white Catholics (0.72) and white mainline Protestants (0.69) are not far behind. To put these scores into perspective, the median score for each white Christian subgroup is significantly above the median scores of the general population (0.57), white religiously unaffiliated Americans (0.42), and black Protestants (0.24).

The shapes of the distributions are also informative. Among white evangelicals, attitudes are distributed nearly perfectly on either side of a high peak near the median score, indicating a great degree of homogeneity in opinion around this mark. The distribution of white Catholics looks similar to that of white evangelical Protestants, but with the median score just to the left of the curve peak—reflecting slightly more respondents with lower scores on the Racism Index. White mainline Protestants have the most variation in attitudes. Like white evangelicals, this group also shows a (lower) peak around 0.8, but the median score is well to the left of that peak, reflecting the presence of significantly more white mainline Protestant respondents toward the lower end of the Racism Index scale compared with the other two white Christian subgroups. By contrast, the distribution of religiously unaffiliated whites is significantly heavier toward the low end of the index, with a peak around 0.1.

We can now compare the scores from the Racism Index to the previously mentioned findings from the “feeling thermometer” question. Although all white Christian groups record similarly warm feeling thermometer scores to Americans overall in the analysis above, each group’s score on the Racism Index is significantly higher than the general population’s. White evangelical Protestants, for example, report the warmest attitudes toward African Americans (an average score of 71 on a scale of 1 to 100), while simultaneously registering the highest score on the Racism Index (0.78 on a scale of 0 to 1). While this stunning contradiction is most pronounced for white evangelical Protestants, the pattern also exists among white mainline Protestants and white Catholics. In other words, by asking multiple questions about concrete policies and specific attitudes rather than relying on more general sentiments, the Racism Index helps us see a paradox that is critical for understanding racial attitudes among white Christians. White Christians think of themselves as people who hold warm feelings toward African Americans while simultaneously embracing a host of racist and racially resentful attitudes that are inconsistent with that assertion. If we want to understand the legacy of white supremacy in American Christianity, we’ll need to move beyond self-reported general sentiment toward African Americans, and the Racism Index gives us a lens to help us see its concrete effects.

This first step in the analysis confirms that the Racism Index is a stable, broad measure of underlying racial attitudes about African Americans, and that attitudes on this index are strongly related to the three largest white Christian subgroups: white evangelical Protestants, white mainline Protestants, and white Catholics. This tells us that there is a robust relationship between holding racist attitudes and identifying as a white Christian. But it’s still possible that this correlation is the result of some other intervening variable.

In order to eliminate that possibility, we need to tap some more sophisticated statistical tools that allow us to control for the influence of a range of other variables. That analysis will allow us to measure how much holding racist attitudes predicts independently the probability of identifying as a white Christian. As another way to understand the relationship between these two concepts, we’ll also flip the direction of the analysis, measuring how much identifying as a white Christian predicts independently the likelihood of holding racist attitudes. In short, we’d like to know whether holding racist attitudes makes one more likely to identify as a white Christian and, conversely, whether identifying as a white Christian in itself makes one more likely to hold racist attitudes.

Does Holding More Racist Attitudes Increase the Likelihood of Identifying as a White Christian?

To sort this out, we turn to what social scientists call “multivariate regression models,” a fancy term for a statistical technique that can unravel the simultaneous influence of multiple concepts in predicting how another attitude or behavior changes. In our case, we need to measure how strongly holding racist attitudes predicts four different white Christian identities: (1) white Christian overall, (2) white evangelical Protestant, (3) white mainline Protestant, and (4) white Catholic. For comparison, we also need a fifth model to predict being white and religiously unaffiliated.19

To ensure that the relationship between the Racism Index and white Christian identity is a direct and independent one, we also need a robust set of independent control variables representing other factors that could theoretically influence the relationship between racial attitudes and white Christian identity. Unless otherwise noted, the control variables in the analysis below are political party affiliation, education level, region, gender, age, household income, home ownership, frequency of church attendance, and living in a metropolitan area. Additionally, I included an Immigration Index variable, a composite variable based on nineteen separate questions measuring attitudes about immigrants and immigration policy.20 The Immigration Index was included to ensure that the attitudes in the Racism Index were, in fact, measuring negative attitudes about African Americans specifically and not negative attitudes about nonwhite racial groups or general threats to cultural change.

One way of understanding the function of a regression model is to think about it as a science project constructed to understand electrical circuitry. We can imagine the science project as a box with multiple electrical switches attached to its left end and a single lightbulb inserted into a socket on the right end. The box itself contains complex circuits connecting all the switches to the light in a way that accounts for how the currents running between these switches and the light might interact with one another to produce a final current to the light. When the bulb is first inserted into the socket, the default state of the experiment—with all switches set in neutral positions—produces a light with a glow set by the average level of electrical current specific to that bulb. The goal of the science project would be to understand the effect that manipulating a single switch might have on the light.21

Turning back to our regression models, we can envision the independent variables (the items whose influence we are interested in testing, such as the Racism Index) as the set of switches attached to one side of our box. We can imagine the dependent variable (the item we are curious to see impacted, such as a specific white Christian identity) as the lightbulb inserted into the socket on the opposite side of the box. And we can think of the regression model as the box itself.

Using this thought experiment, a number of things could theoretically happen when a switch is flipped. If there is no significant, independent relationship between a switch and a particular bulb, we could flip the switch on and off in vain, seeing no effect at all on the light. For example, with other controls in place, education level had no independent impact on white Christian identity, so flipping that switch would have no effect on the light. If there were a weak relationship between an independent variable and white Christian identity, flipping the switch would produce only a modest change in the light: slightly brighter for a positive relationship and slightly dimmer for a negative relationship. For example, with other controls in place, higher household income was positively predictive of white Christian identity (in other words, more income = more likely to identify as white Christian), but the magnitude of this effect is very modest, so flipping that switch would only slightly brighten the light in our thought experiment. Only if there were a strong relationship between an independent variable and a particular white Christian identity would flipping the switch produce a very bright light. Because of the work of the box circuitry, we would know this effect was not caused by any other variable.

The results below spell out what happens when we flip the Racism Index switch on the box and look to the light on the other side representing some variety of white Christian identity. Figure 5.4 summarizes the independent impact of the Racism Index on the likelihood of white Christian identity overall, among the three largest white Christian subgroups, and among whites who are religiously unaffiliated. In other words, if we insert the lightbulb representing each of these target groups one at a time and flip the Racism Index switch, here’s what happens.

These results are striking. Even with all the statistical controls in place, the Racism Index remains an independent predictor of white Christian identity overall for each of the three white Christian subgroups individually, and—in the opposite direction—for religiously unaffiliated whites. In the model, when the RI shifts from least racist to most racist (a move from 0 to 1 on the RI), that shift independently makes an average respondent 18 percentage points more likely to identify as white mainline Protestant, 19 percentage points more likely to identify as white evangelical Protestant, and 20 percentage points more likely to identify as white Catholic. For white Christian identity overall, which accounts for the combined effect of all three identities, maximizing the Racism Index score makes a respondent 57 percentage points more likely to identify as a white Christian.22 By contrast, the corresponding shift in the RI has only a very weak effect on white religiously unaffiliated identity.23

Translating these results to our thought experiment, we can imagine having a lightbulb—one for each of these white Christian identities (white evangelical Protestant, white mainline Protestant, and white Catholic) plus white religiously unaffiliated identity—whose brightness ranges from 0 to 100. We attach these bulbs to our science fair box one at a time and flip on the switch representing the RI to its maximum value. In each case, flipping the RI switch produces a light that is approximately 20 percentage points brighter than that bulb’s default state. And when the bulb representing all white Christians is attached, it produces a truly incandescent light, nearly 60 percentage points brighter. By contrast, when the RI switch is flipped with the white religiously unaffiliated bulb plugged in, the light dims slightly.

Image

FIGURE 5.4 Change in Predicted Probability of White Religious Identities Accompanying a Shift from Least to Most Racist on the Racism Index

Source: PRRI, American Values Survey, 2018.

The models reveal that, in the United States today, the more racist attitudes a person holds, the more likely he or she is to identify as a white Christian. And when we control for a range of other attributes, this relationship exists not just among white evangelical Protestants but also equally strongly among white mainline Protestants and white Catholics.24 And there is also a telling corollary: this relationship with racist attitudes has little hold among white religiously unaffiliated Americans; if anything, the relationship is negative.

A look at the role of church attendance levels casts further light on the relationship between holding racist attitudes and white Christianity. Some have argued, in defense of white Christian churches and institutions, that this link is driven primarily by those who claim a Christian identity but who have little connection to Christian churches. There’s even an acronym for this condition: CINO (Christians in name only). Those loosely connected white Christians, the theory goes, are more likely to hold racist views, while those who attend religious services more often—with more exposure to sermons, Sunday school, Bible study, and other forms of Christian discipleship offered within congregations—are more likely to be in solidarity with their African American brothers and sisters.25

But there is no evidence that higher church exposure has any mitigating effect on racist attitudes; if anything, the opposite is true. For all white Christian subgroups, there is a positive relationship between holding racist attitudes and white Christian identity among both frequent (weekly or more) and infrequent (seldom or never) church attenders. For white Catholics, there are no attendance differences: a move from least racist to most racist on the Racism Index makes both frequent and infrequent church attenders more likely to identify as white Catholic (21 percentage points and 19 percentage points, respectively). For white mainline Protestants, infrequent church attenders see a bigger boost in probability of identification related to holding more racist views (22 percentage points), but the identification boost due to racist views among frequent church attenders is also positive and significant (12 percentage points).

For white evangelical Protestants, there is, strikingly, a stronger boost in likelihood of affiliation due to racist attitudes among frequent church attenders than among infrequent church attenders. A move from least racist to most racist on the Racism Index makes frequent church attenders 34 percentage points more likely to identify as white evangelical Protestants, compared with an increase of only 9 percentage points among infrequent church attenders. In other words, holding racist views is nearly four times as predictive of white evangelical Protestant identity among frequent church attenders as among infrequent church attenders.

Moreover, if we ask the question of where church attendance has the largest influence on white evangelical Protestant identity, we get a startling answer, which can be seen in the gap between the two lines in figure 5.5. The largest gap between the frequent and infrequent church attenders is among those with the highest (most racist) scores on the Racism Index. It’s the opposite pattern that anyone thinking of the church as a moderating force on race relations would expect. Among Americans holding the most racist views (Racism Index = 1), frequent church attenders are 31 percentage points more likely than infrequent church attenders to identify as white evangelical Protestant. By contrast, among Americans with the least racist views (Racism Index = 0), the likelihood of white evangelical Protestant identification between frequent and infrequent church attenders is nearly indistinguishable.

Image

FIGURE 5.5 Predicted Probability of White Evangelical Protestant Identity, by Racism Index Score and Church Attendance Frequency

Source: PRRI, American Values Survey, 2018.

Racist Attitudes and White Christian Identity at the Regional Level

As explosive as these findings are, the national numbers tell only part of the story. An examination of the regional variations within each of these models provides important, unanticipated insights into how the relationship between racial attitudes and white Christian identity functions closer to the ground. Historically, and still today, each of these white Christian traditions has been culturally dominant in different regions of the country. For example, white evangelical Protestants are dominant in the South, and white Catholics are dominant in the Northeast. White mainline Protestants have been more evenly distributed, but they have historically been dominant in the Midwest, notable as the prevailing expression of white Protestantism in the Northeast, and, to a lesser extent, influential in the peripheral South, where white Methodists are abundant. The West stands out as the census region that has not had a clear, dominant white Christian tradition. In fact, today the most dominant “religious” group in the West is the religiously unaffiliated, which constitutes about three in ten westerners, compared with only about one in ten westerners who identify with one of the three white Christian subgroups.

If we understand negative racial attitudes to be about white supremacy—that is, expressions of support for white social dominance and control—we might expect the relationship between racist attitudes and white Christian identity to be stronger for each white Christian subgroup within the region in which they are culturally dominant. This is exactly what we see at the regional level, as the chart on the next page indicates. In figure 5.6, the black dots represent the increase in probability of each type of white Christian identity, within each region, that accompanies a shift from least to most racist (from 0 to 1) on the Racism Index. The lines on each side of the dots are “error bars,” indicating how precise an estimate is, based on the sample sizes and distribution of each subgroup. The shorter the error bar, the more precise the estimate. 26

For white evangelical Protestants, the largest boost in the predicted probability of affiliation due to increased racist attitudes is in the South. Whereas the predicted probability of identifying as a white evangelical Protestant due to racist attitudes is 19 percent nationally, it jumps to 29 percent in the South. Similarly, whereas the predicted probability of identifying as a white Catholic due to racist attitudes is 20 percent nationally, it soars to 54 percent in the Northeast. For white mainline Protestants, who are geographically more distributed, the regional differences are weaker; but the predicted probability of identifying as a white mainline Protestant due to racist attitudes moves from 18 percent nationally to 23 percent in the Northeast, the region with the highest effect. Notably, for white mainline Protestants, the predicted probabilities in the Midwest (19 percent) and the South (21 percent) are also substantial.

These regional contingencies suggest that within specific geographic regions, especially where a particular white Christian group holds a dominant cultural position, the connections between racist attitudes and white Christian identity are notably stronger. In the South, white evangelicalism receives a significant affiliation boost due to racist attitudes. While white evangelicalism benefits little from increasing racist attitudes in the Northeast, white Catholicism and, to a lesser extent, white mainline Protestantism become the beneficiaries in that region. White mainline Protestants receive a more evenly distributed boost in the probability of affiliation due to racist attitudes across the Northeast, Midwest, and South. Each form of white Christian identity receives a boost in the probability of affiliation as a result of racist attitudes, and for white evangelicals and white Catholics in particular, this boost is turbocharged in the regions where they are culturally dominant.

Image

FIGURE 5.6 Change in Predicted Probability of White Religious Identities Accompanying a Shift from Least to Most Racist on the Racism Index, by Geographic Region

Source: PRRI, American Values Survey, 2018.

The combined white Christian model also reveals some surprising insights. Most notably, the region in which white Christianity receives the highest boost in affiliation due to racist attitudes is not the South but the Northeast. This unexpected finding is due to the fact that unlike the South, where white evangelicals are the single dominant group, the Northeast has two culturally dominant white Christian groups: one Protestant (mainline) and one Catholic. Like product diversification in the marketplace, the presence of two groups that are compatible with white supremacist affinities effectively gives whites with strong racist attitudes two possible Christian identities with which to affiliate. The South clocks in as the region with the second-highest boost in affiliation due to racist attitudes, based principally on the strong effects among white evangelical Protestants, with a smaller boost from white mainline Protestants. The Midwest ranks third, primarily because of the presence of white mainline Protestants and, to a lesser extent, white evangelical Protestants. With no culturally dominant white Christian group, the West ranks last, with the smallest boost in white Christian identity due to racist attitudes.

Compared with white Christian identities, the differences in the patterns predicting white religiously unaffiliated identity are telling. Increased racist attitudes have virtually no impact on the probability of white religiously unaffiliated identification in the West or the South. In the Midwest, while these effects are modest, they flow in a negative direction: higher scores on the Racism Index make one slightly less likely to be white and religiously unaffiliated. And in the Northeast—the region where both white Catholicism and white mainline Protestantism are strong beneficiaries of racist attitudes—holding more racist attitudes results in a 34 percent reduction in the likelihood of identifying as white and religiously unaffiliated. In other words, particularly in the Northeast, holding racist attitudes has become a strong differentiator between white Christians and white religiously unaffiliated residents; those holding the strongest racist attitudes are more likely to feel at home with either white mainline or Catholic Christians, as compared with religiously unaffiliated whites.

The data suggest that white Christian churches, both Protestant and Catholic, have served as institutional spaces for the preservation and transmission of white supremacist attitudes. Rather than deconstructing this racist ideology, most white Christian churches have protected white supremacy by dressing it in theological garb, giving it a home in a respected institution, and calibrating it to local cultural sensibilities. And this analysis shows just how much this legacy has become embedded in the DNA of white Christianity today.

Flipping the Analysis: How Much Does White Christian Identity Predict Higher Racist Attitudes?

Reversing the direction of the analysis above also helps us confirm the strong relationship between racist attitudes and white Christian identity. If we shift the Racism Index to be the dependent variable (make it the bulb instead of a switch in our imaginary science experiment), and move white Christian identities to be independent variables (make them switches), we can see how much each white Christian identity predicts independently an increase in racist attitudes as measured by the RI. This analysis largely confirms the previous findings. (See appendix C for full model output.27)

Using the same control variables in the models above, being affiliated with each white Christian identity is independently associated with a nearly 10 percent increase in racist attitudes, compared with those who do not identify as a white Christian: 9 percent for white evangelical Protestant identity, 8 percent for white mainline Protestant identity, and 9 percent for white Catholic identity.28 By contrast, there is no significant relationship between white unaffiliated identity and holding racist attitudes.

Notably, looking at the analysis in this direction, church attendance has no significant impact on the relationship between white Christian identities and holding racist views, confirming the findings of the analysis above. In other words, there is no evidence that going to church every week, at least at the churches white Christians are currently attending, makes a white Christian any less likely to be racist. Whatever Christian formation and discipleship is happening is not impacting the white supremacist attitudes that are deeply embedded in white Christian institutions of all types. White evangelical Protestants, white mainline Protestants, and white Catholics who go to church frequently are as likely as their less-frequently-attending counterparts to hold racist attitudes.

The relationship, then, between the Racism Index and white Christian identity is a broad two-way street: an increase in racist attitudes independently predicts an increase in the likelihood of identifying as a white Christian, and identifying as a white Christian is independently associated with an increased probability of holding racist attitudes.

A Summary of the Statistical Findings

We’ve covered a lot of statistical ground in this chapter. Below is a summary of the main findings:

This analysis leaves us with some remarkable conclusions. If you want to predict whether an average person is likely to identify as a white Christian, and you could know only one attribute about that person, you would be better off knowing how racist he or she is than how often he or she attends church. Or, to put it even more bluntly, if you were recruiting for a white supremacist cause on a Sunday morning, you’d likely have more success hanging out in the parking lot of an average white Christian church—evangelical Protestant, mainline Protestant, or Catholic—than approaching whites sitting out services at the local coffee shop.

Conclusion

While the persistence of racist attitudes among white Christians today may seem astonishing at first blush, the three authors of Deep Roots, Acharya, Blackwell, and Sen, note just how close the Civil War and the institution of slavery are to us. The last known person to have been born into slavery, Sylvester Magee, died in Hattiesburg, Mississippi, in 1971,29 and Ruth Odom Bonner, a child of former slaves, lived long enough to stand alongside President Barack Obama to ring the bell opening the National Museum of African American History and Culture in 2016, at the age of ninety-nine.30

Avidit Acharya and his coauthors also demonstrate how cultural beliefs and behaviors can persist over time. They argue that beliefs and behaviors can become socially self-reinforcing; once a fork in the cultural road has been chosen, especially among a social group, it becomes harder to deviate from that path because of the social pressure that accumulates over time from parents and grandparents, settled communal norms, and institutions that codify and preserve these beliefs and bridge generations. They sum up this dynamic as follows: “Similar to religion and language, attitudes—including political and racial attitudes—are passed down from generation to generation, fostered and encouraged by families and social structures, such as schools and churches.”31

While Acharya and his colleagues explicitly mention religion and churches early in their book as general examples of the kind of institutional, connective tissue they theorize is at work in passing down attitudes about race across generations, this mechanism largely drops out of their statistical analysis and findings. At the conclusion of the book, they are perplexed that the successes of the civil rights movement and the fall of Jim Crow laws have not done more to mitigate the continued effects of slaveholding on the contemporary attitudes of whites in the South. They note that these sweeping social and legal changes have been less effective than one might expect at attenuating “differences in behavioral outcomes—including political attitudes, opinions on race, and support or opposition for race-related policies.”32 They conclude simply, “The fact that we can still detect this kind of divergence between whites living in these areas and whites living in other parts of the South is a testament to slavery’s lasting political and cultural legacy.”33

This much is true. But the authors miss the opportunity to identify a key conduit of the ongoing cultural transmission of white supremacy: white Christian churches.34 Even as Jim Crow laws have been struck from the books in the political realm, most white Christian churches have reformed very little of their nineteenth-century theology and practice, which was designed, by necessity, to coexist comfortably with slavery and segregation. As a result, most white Christian churches continue to serve, consciously or not, as the mechanisms for transmitting and reinforcing white supremacist attitudes among new generations.

This chapter demonstrates—with rigorous quantitative evidence—a disturbing fact: that Frederick Douglass’s nearly two-hundred-year-old observations about the positive correlation between white supremacy and Christianity continue to be supported by the contemporary evidence. Not only in the South but nationwide, higher levels of racism are associated with higher probabilities of identifying as a white Christian; and, conversely, adding Christianity to the average white person’s identity moves him or her toward more, not less, affinity for white supremacy. White supremacy lives on today not just in explicitly and consciously held attitudes among white Christians; it has become deeply integrated into the DNA of white Christianity itself.

That last statement, standing alone, sounds shocking. But an honest look at the historical arc of white Christianity in America suggests that we should instead be astonished if it were otherwise. For centuries, through colonial America and into the latter part of the twentieth century, white Christians literally built—architecturally, culturally, and theologically—white supremacy into an American Christianity that held an a priori commitment to slavery and segregation. At key potential turning point moments such as the Civil War and the civil rights movement, white Christians, for the most part, did not just fail to evict this sinister presence; history confirms that they continued to aid and abet it. The weight of this legacy is indeed overwhelming. But as the next chapter shows, there are signs that at least some white Christians are facing the reality of this history and have taken a few steps along a new path toward repentance and repair.