Introduction

Over the past few decades, scholars and journalists have noted what they consider to be a resurgence of a particularly conservative strain of American civil religion, Christian nationalism, understood generally as a belief that America has been and should always be distinctively Christian in its identity, sacred symbols, values, and policies (Aho 2013; Froese and Mencken 2009; Goldberg 2006; Gorski 2009, 2010; McDaniel et al. 2011). Though concerning for a variety of reasons pertaining to social justice, particularly worrisome are the ideological foundations of Christian nationalism, historically rooted in notions of ethno-racial separatism and imperialism (Aho 2013; Gorski 2009, 2010). Although some researchers believe that the racialist elements of Christian nationalism were greatly weakened over the last half-century (e.g., Gorski 2010), others contend that Christian nationalism in the present day still retains elements of ethno-racial purity, ethnic exclusivism, and supremacy (Aho 2013; Barkun 1994; Goldberg 2006; McDaniel et al. 2011). These authors claim that a resurgence of Christian nationalism could provide ideological support for arguments and policies that covertly reinforce racial boundaries in social life or justify the patterns of systemic racial exclusion.

We argue that, within the “racialized social system” (Bonilla-Silva 1999, 2001) of the United States, the idealized religio-national purity foundational to Christian nationalist thought is inextricably linked with notions of ethno-racial purity and separation. We test this theory by focusing on the link between Christian nationalist ideology and Americans’ views toward family forms where ethno-racial purity is formally violated. Interracial families, in theory or in actual practice, serve as a particularly salient indicator of racial boundaries, and thus, openness to such families has long served as a useful barometer of race relations in the United States (Allport 1954; Bogardus 1933; Gordon 1964; Lee and Bean 2010; Yancey and Lewis 2009). Drawing on survey data from a national probability sample of American adults and employing a multidimensional measure of Christian nationalism, we focus on the link between Christian nationalism and Americans’ support for transracial adoption (TRA)—understood as the legal, permanent placement of minors with parent(s) of another race (Barn 2013; Briggs 2012; Smith et al. 2011; Perry 2013).

Attitudes toward TRA represent a unique and interesting test case in that, unlike racial exogamy, TRA does not signal the biological blending of racial genotypes in mixed-race progeny. Nor does TRA necessarily imply the confrontation of potentially disparate ethno-racial cultures as in interracial dating or marriage.Footnote 1 TRA does not even entail a relationship between peers. Rather, TRA denotes the social and legal uniting of racial groups in a situation where one has guardianship over the other (e.g., white parents rearing black, Latino, or Asian children). Christian nationalist opposition to racial exogamy may be attributed to old-fashioned eugenicist notions of racial supremacy and purity or views about the cultural incompatibility of racial groups in romantic relationships. The idea of TRA as an interracial family relationship, however, is more likely to tap into beliefs that racial groups should be socially and legally separate even when no biological or cultural race-mixing between peers is in view. This research ultimately advances our understanding of the ethno-racial content implicit within Christian nationalist ideology and, thus, validates concern about the resurgence of Christian nationalism in the public sphere as potentially exacerbating racial tensions and ultimately inequalities in the United States.

We begin by briefly surveying the development and ideological content of Christian nationalism in the United States, giving particular attention to its ethno-racial underpinnings and implications for Americans’ racial attitudes. We then survey research on correlates of support for TRA and develop hypotheses about the potential influence of Christian nationalism.

Christian Nationalism and Ethno-Racial Purity in the United States

In surveying the history of Christian nationalism in the United States, Gorski (2009) cautions that, despite its apparent resurgence since 2000, scholars should not mistake Christian nationalism for a recent phenomenon. Rather, it is “older than the Republic” (Gorski 2009:91) and has long been one of the central continuities of American political discourse. He defines Christian nationalism as a “blending of Christian and patriotic narratives and iconography that blurs or erases the line between religious and political community and identity” (Gorski 2009:91). Following Bellah (1967), Gorski (2010:7) distinguishes Christian nationalism from “American civil religion” in that, civil religion views the religious and political spheres as “independent but interconnected,” while Christian nationalists “advocate a total fusion” between the two spheres. Simply put, Christian nationalists “wish the boundaries of the religious and political communities to be as coterminous as possible” (Gorski 2010:7).

Gorski (2010) traces Christian nationalism in the United States to two principal sources: ethno-nationalism and the Hebrew Bible. Envisioning America as the “New Israel,” Christian nationalists read Old Testament passages about Israel’s chosen status and God’s demand for their blood purity through marital-endogamy and blood sacrifice not as a metaphor, but more literally as requiring ethnic separation and the racial purity of his “new” chosen people. Moreover, Christian nationalism has historically become more salient in times of heightened cultural and political conflict. In these times, boundary lines of identity and belonging require reinforcement and out groups become demonized (Gorski 2010).

Although Gorski (2010) theorizes that following the Holocaust and the Civil Rights movement the explicitly racial elements of Christian nationalism were greatly attenuated, others maintain that Christian nationalism, particularly as it pertains to contemporary white America, implies the same desire for racial purity and exclusivity in a more covert form. Aho (2013) and Goldberg (2006), for example, argue that Christian nationalist ideals (e.g., individualism, support for expanding religion’s role in government, reigning in government debt and spending, and enforcing a biblical sexual morality) provide ideological support for contemporary neoliberal policies that staunchly oppose welfare aid to “unworthy” poor minority families and race-based affirmative action policies. Aho (2013) also shows how contemporary Christian nationalists embrace and promote revisionist histories of America’s racial oppression, expressing ambivalence about the slavery of blacks, and even emphasizing the “benefits” that accrued to blacks because of their enslavement (e.g., Barton 2011).

Scholars have also linked contemporary Christian nationalist ideology with antagonism toward foreigners and immigration. Because Christian nationalists believe America’s “Christian heritage” should be defended, they tend to oppose the immigration of non-Christian groups, who also tend to be nonwhite (e.g., Muslims) (Aho 2013). Froese and Mencken (2009) demonstrate that Christian nationalism (what they term “sacralization ideology”) is a strong predictor of whether Americans supported the Iraq War, a war which, following 9/11, revealed a merging of ethno-racial and ethno-religious animus toward Muslims from the Middle East. McDaniel et al. (2011) demonstrate that Christian nationalist ideology is a strong predictor of animus toward immigrants, and that conservative Protestantism only predicts anti-immigrant views when Christian nationalism is not included in models. Though certainly important for linking Christian nationalist ideology and prejudice toward ethnic out groups, immigration as an issue is freighted with other implications that are not explicitly racial that complicate this relationship, including class antagonisms, symbolic-cultural differences (e.g., language), and religious prejudice. The current study turns the focus to TRA, which we argue represents a more direct test case from which to assess the relationship between Christian nationalism and Americans’ racial attitudes.

Attitudes Toward Transracial Adoption in the United States

TRA has been a politically charged issue since the 1970s. While TRA may technically include international adoptions or adoptions involving any combination of races, the vast majority of the research and conflict regarding TRA has centered on the adoption of minority children (and especially black or African American children) by white parents (Bausch and Serpe 1997; Briggs 2012; Fisher 2003; Smith et al. 2011). Although Americans’ attitudes toward TRA in general are overwhelmingly positive today (Evan B. Donaldson Institute 2002; Perry 2014), the issue of minority children being adopted by white families has historically been met with skepticism or outright opposition both from white traditionalists on the right (Herman 2008), and those on the left who question the long-term efficacy of TRA to address deeper issues of vulnerable minority families and their children, as well as the short-term ability of white parents to raise minority children in a racist society (Briggs 2012; National Association of Black Social Workers 1972; Patton 2000; Quiroz 2007).

Most research on TRA focuses primarily on the life outcomes of TRA for children by various metrics (Feigelman and Silverman 1984; Fisher 2003; McRoy and Zurcher 1983; Samuels 2009a, b; Silverman 1993; Silverman and Feigelman 1981; Simon and Altstein 2002), the characteristics of those who transracially adopt (Davis 2011; Fisher 2003; Grow and Shapiro 1976; Jacobson et al. 2012; Hollingsworth 2000a), or the social and political underpinnings of TRA policy and practice (Bartholet 1991, 1995; Briggs 2012; Kennedy 2003; Patton 2000; Quiroz 2007, 2012). Comparatively little work exists comparing what social factors predict general attitudes toward TRA. Most of these studies have examined the TRA attitudes of various subpopulations such as African Americans (Simon 1978), Mexican Americans (Bausch and Serpe 1997), social workers (Fenster 2003; Grow and Shapiro 1976), adoptive parents (Grow and Shapiro 1976), gays and lesbians (Goldberg and Smith 2009), and college students (Chima 1996). Comparatively few studies examine the correlates of attitudes toward TRA among the general public (Evan B. Donaldson Institute 2002; Hollingsworth 2000b; Perry 2010, 2014). Among these studies, researchers find that those who approve of TRA tend to be younger, female, politically liberal, more educated, more racially tolerant, live out of the South, and come from racially integrated backgrounds (Goldberg and Smith 2009; Hollingsworth 2000b; Perry 2010, 2011). Race is an inconsistent predictor of support for TRA. While some find whites to be more supportive of TRA than African Americans (Fenster 2003), others find a negligible to null effect of race on support for TRA (Evan B. Donaldson Institute 2002; Perry 2010).

Only recently have researchers sought to examine the relationship between religious factors and attitudes toward TRA. Within these studies, Protestants, and particularly conservative Protestants, are less favorable toward TRA than Catholics, those of other religious faiths, and the religiously unaffiliated (Fenster 2003; Perry 2010, 2011, 2013). Perry (2010, 2011) also finds, however, that Americans who are more religious in terms of religious practice (worship attendance, prayer, sacred text reading) are actually more likely to support TRA than those who practice their religious faith less frequently. More recently, Perry (2013, 2014) reports that while conservative Protestants or “evangelicals” tend to be less supportive of TRA than other religious traditions on average, they appear to be influenced toward greater support for TRA as certain aspects of their political conservativeness (namely, their intolerance of abortion and government spending on behalf of vulnerable populations) increases (see also Briggs 2012). The relationship between religious and political ideology and TRA support is thus complicated and warrants greater systematic examination. Below, we theorize a connection between Christian nationalism and Americans’ views toward TRA to generate several testable hypotheses.

Christian Nationalism and Views Toward Transracial Adoption

The United States represents a “racialized social system,” where race is structured into the ideological, political, and social life of all society (Bonilla-Silva 1999, 2001). Within this system, racial ideologies provide tacit justification for social relations and patterns of interaction within and across racial groups. At the ideological level, the ethnic element implicit within Christian nationalism has historically provided justification for ethno-racial separation and exclusion, founded on conceptions envisioning America as “God’s new Israel”—originally commanded by God to maintain ethnic distinctiveness (Gorski 2009, 2010). It is unsurprising, then, that researchers on contemporary Christian nationalism propose that ethno-racial purity and separation is still implied in calls for America to “preserve” or “restore” its “Christian heritage” (Aho 2013; Goldberg 2006; McDaniel et al. 2011). Empirical tests of this theory, however, have been limited. While McDaniel et al. (2011) link Christian nationalism to anti-immigrant sentiment, the issue of race was only implied in the issue of immigration, not explicitly stated, and thus, the relationship they find is potentially confounded by issues of class, symbolic-cultural, and religion-based prejudices. Looking at TRA attitudes, our study can more directly assess the ethno-racial content of Christian nationalist thought. We also argue that, as an interracial family relationship, TRA provides us with an interesting test case in that, the social and legal uniting of racial groups is in view, not necessarily biological or even cultural race-mixing. Opposition to TRA as a relationship thus does not stem from eugenicist notions of racial supremacy or even beliefs about cultural incompatibility between ethno-racial groups, but more so respondents’ beliefs about the “ideal” ethno-racial composition of families. Building on previous work on Christian nationalism and empirical research on TRA attitudes, we expect that

Hypothesis 1

Adherence to Christian nationalism will be negatively associated with support for transracial adoption, net of other factors.

Beyond simply establishing an empirical association between Christian nationalism and TRA support, our study seeks to distinguish Christian nationalism in the United States as an ideology that implies and justifies ethno-racial separation and purity rather than one that encourages general racial prejudice or distrust. Previous research on TRA attitudes demonstrates that respondents’ reported level of trust in people of other racial groups is a strong predictor of support for TRA (Perry 2010, 2014), and thus, we control for this in multivariate models to demonstrate the robust and unique effect of Christian nationalism. We hypothesize that

Hypothesis 2

Christian nationalism and respondents’ trust in other races will both have strong and independent effects on TRA attitudes, net of other factors.

Lastly, we argue that Christian nationalism is an implicitly ethno-racial ideology in and of itself, and not to be conflated with or viewed as epiphenomenal of religious conservatism or devotion. Previous work by Perry (2010, 2014) on TRA attitudes finds that Americans’ level of religious commitment is actually positively associated with support for TRA. Thus, while Christian nationalism is expected to be negatively associated with TRA approval, we argue that this effect is not a reflection of religious commitment per se, but rather of the implicit ethno-racial content of Christian nationalist ideology. Therefore, we predict

Hypothesis 3

In contrast with adherence to Christian nationalism, religious commitment (reflected in frequency of religious practice) will be positively associated with support for TRA, net of other factors.

Methods

Data

We test these hypotheses using data from Wave 1 of the Baylor Religion Survey (BRS), which was fielded in 2005. The BRS design was based on the General Social Survey (GSS), but with a goal to comprehensively assess the religious attitudes, beliefs, and behaviors of American adults. Survey data were collected by the Gallup Organization, using a mixed-mode method (telephone and mailed questionnaires). Gallup contacted 7,041 potential respondents by telephone and mailed out 2,603 questionnaires. Of the 2,603 questionnaires distributed, 1,721 completed surveys were returned for a response rate of 66.1 %. The response rate for the entire sampling frame was 24.4 % (1,721/7,041). Though not ideal, this response rate is well above the average response rate for random digit dial surveys (<10 %) (Braunsberger et al. 2007), and Bader et al. (2007) have demonstrated that the BRS data compare favorably with other nationally representative surveys like the GSS on demographic and religious measures. For a more detailed account of the sampling procedure and data collection, see Bader et al. (2007).

Dependent Variable

The dependent variable for this analysis is respondents’ attitudes toward TRA. This is measured with the question, “How do you feel about the following marriage and family related issues…adopting a child of a different race?” Respondents could indicate that they felt it was: (1) always wrong, (2) almost always wrong, (3) only wrong sometimes, or (4) not wrong at all. Because this study is most interested in those who express unequivocal support for TRA, responses were collapsed and dichotomized into a binary variable (“not wrong at all” = 1, “wrong or sometimes wrong” = 0).Footnote 2 Throughout the analyses, respondents with a one for this question are understood to approve of TRA without reservation.

Christian Nationalism Measures

The independent variable of interest for this study is Americans’ relative support of Christian nationalism. To measure this concept, we construct a multi-item index utilizing five level-of-agreement statements, which included “The federal government should advocate Christian values,” “The federal government should defend Christian values,” “The federal government should allow the display of religious symbols in public spaces,” “The federal government should allow prayer in public schools,” and “God favors the United States in worldly affairs.” Responses ranged from (1) strongly disagree to (5) strongly agree. The Cronbach’s alpha coefficient for these measures was 0.88 indicating high reliability. The index ranges from 5 to 25 with a mean of 15.63 (see Table 1). Froese and his colleagues (Froese and Bader 2007; Froese and Mencken 2009) utilize a similar index and name the construct “sacralization ideology.” We would argue that the term “sacralization ideology” is perhaps too broad. Rather, the BRS measures, with varying degrees of specificity, ultimately inquire about respondents’ beliefs that America is a distinctively Christian nation. The statements about the federal government advocating and defending “Christian values” are clear. But it would also be a stretch to argue that the vast majority of American adults are thinking about any other religion but Christianity when asked whether the federal government should allow the display of religious symbols in public spaces, or prayer in public schools, or whether God favors the United States in worldly affairs. Affirmative answers to these questions almost certainly indicate that the respondent envisions the United States as a Christian nation, where Christian identity, values, rituals, and symbols are privileged by the federal government. The high alpha coefficient for these measures (0.88) would support this argument.

Table 1 Descriptive and bivariate statistics for all variables in full model (MI data)
Full size table

Controls

Multivariate models include a number of religious, socio-demographic, and ideological controls following previous research on public opinion toward TRA (Evan B. Donaldson Institute 2002; Fenster 2003; Hollingsworth 2000b; Perry 2010, 2014). Religious controls for this study include religious tradition, theological beliefs, and religious practice. Religious tradition is measured with seven broad categories following Steensland et al. (2000): evangelical, mainline Protestant, Catholic, black Protestant, Jewish, Other, and Unaffiliated. The “Other” category includes Mormons, Muslims, and other religious traditions with too few cases to analyze individually. Evangelicals serve as the reference category. In order to control for respondents’ theological conservatism, which is a consistent predictor of prejudice toward out groups (Kirkpatrick 1993), we also include a measure tapping whether respondents are strict biblical literalists (biblical literalist = 1; not literalist = 0). Lastly, in order to gauge various dimensions of religious behavior, we construct a religious practice scale using three BRS questions concerning respondents’ frequency of religious service attendance, sacred text reading, and prayer. For frequency of religious service attendance and sacred text reading, respondents could indicate (1) never to (9) several times a week. For prayer frequency, respondents chose options from (1) never to (6) several times a day. These measures were standardized into Z scores and summed in order to create the religious practice scale (α = 0.83). Socio-demographic controls included gender (female = 1), child(ren) in the home (children = 1), race (nonwhite = 1),Footnote 3 region (South = 0, West = 1, East = 1, Midwest = 1), and political affiliation (republican = 0, independent/other = 1, democrat = 1). Age is measured in years from 18 to 93. Education is measured in attainment categories ranging from (1) 8th grade or less to (7) postgraduate work/graduate degree. Lastly, racial trust/distrust is measured with respondents’ answer to the question “How much do you trust [people of other races?]” Responses are coded 1 = “a lot,” and 0 = other.

Analysis

The analysis proceeds as follows. Table 1 presents descriptive statistics, bivariate percentages, and correlation coefficients for Americans’ support for TRA and all predictor variables. Second, Table 2 presents binary logistic regression models predicting support for TRA. Models are organized as follows. Model 1 includes only the Christian nationalism index. Model 2 adds the religion controls. Model 3 adds the socio-demographic controls. Model 4, the full model, adds respondents’ reported trust in other races. Formally, the full model takes the following form

$$\log {\text{it}}\left( Y \right) = \log \left( {\frac{\pi }{{\left( {1 - \pi } \right)}}} \right) = \alpha + b_{1} X_{1} + b_{2} X_{2 } + b_{3} X_{3 } + b_{4} X_{4 }$$

where log represents the natural logarithm; π is the probability that the dichotomous outcome variable Y = 1 (respondent supports TRA); α is the Y intercept; b 1–b 4 are unstandardized regression coefficients; X 1 is our Christian nationalism scale; X 2 is a vector of religion controls; X 3 is a vector of socio-demographic controls; and X 4 is the respondents’ reported trust in other races.

Table 2 Logistic regression predicting support for transracial adoption
Full size table

We present both odds ratios and standardized regression coefficients in Table 2. We use standardized coefficients in order to account for substantive significance instead of only statistical significance. Standardized coefficients are estimated as B * yx  = b yx (s x /s y ), and we follow Pampel’s (2000) assumption that the standard deviation of logit(Y) = 1.8138.

To correct for missing data, we utilize multiple imputation procedures, which allow us to take full advantage of the available data and avoid potential bias in standard errors and test statistics associated with listwise deletion (Allison 2002; Rubin 1996). The MI procedure generates five imputations using multiple Markov Chains based on all variables included in each model, resulting in an overall N of 8,240 (1,648 × 5). All results use the MI dataset. The correlations reported in Table 1 and all of the results reported in Table 2 are from the MIANALYZE procedure in SAS. This procedure combines all of the results from each of the five imputations resulting in overall estimates, standard errors, and significance levels. The standardized coefficients and odds ratios for each model were calculated using these overall estimates. The proportional reduction in error (PRE) reported in Table 2 for each model is the average of the PRE for each individual iteration.

Results

In Table 1, we find that just over eight in 10 Americans unequivocally support TRA. However, people who are Christian nationalists tend to be less likely to support TRA (r = −0.13; p < 0.001). Christian nationalism, in fact, has the second strongest correlation with TRA attitudes behind only age (r −0.27; p < 0.001). Interestingly, there is no significant zero-order correlation between more frequent religious practice and support for TRA.

Turning to our logistic regression models in Table 2, we find that in Model 1, Christian nationalism is strongly and significantly associated with attitudes toward TRA. As Americans increase in adherence to Christian nationalism, the likelihood of supporting transracial adoption decreases (β = −0.19; p < 0.001). Put another way, for each one-unit increase on the Christian nationalism scale, the odds a person will support TRA decrease 6 %.Footnote 4

We find that when including a battery of religion control measures, the strength of the relationship between Christian nationalism and support for TRA actually increases. In Model 2, increasing levels of Christian nationalism are significantly and negatively related to support for TRA (β = −0.27; p < 0.001). Increasing one unit on the Christian nationalism scale leads to an almost 9 % decrease in the odds of supporting TRA. In this model, we also find that increasing levels of religious practice lead to greater likelihood of support for TRA (β = 0.17; p < 0.001). Finally, Catholics (β = 0.11; p < 0.05) and the unaffiliated (β = 0.12, p < 0.05) both appear to be more likely to support TRA when compared to evangelical Protestants.

Christian nationalism maintains a significant and negative association with support for TRA even when adding all of the socio-demographic control variables in Model 3 (β = −0.18, p < 0.01). It is worth noting, however, that the relative size of the effect is attenuated somewhat when including these additional controls. Now, for each one-unit increase in Christian nationalism, the odds a person will support TRA decreases by 5 %. The religious practice index maintains a strong and positive association with TRA support, as do Catholics and the unaffiliated when compared to evangelicals. The strongest effect among the socio-demographic measures is age; as age increases, the likelihood of support for TRA decreases (β = −0.44; p < 0.001). Compared to people from the South, those from the West and East are both more likely to support TRA.

In our final model, the association between Christian nationalism and support for TRA continues to be significant and negative (β = −0.17; p < 0.01). Net of all other effects, the odds a person will support TRA decreases 5 % for each one-unit increase on the Christian nationalism scale. Americans who are more religiously active tend to have greater odds of supporting TRA (β = 0.19; p < 0.001), along with Catholics (β = 0.13; p < 0.01) and the unaffiliated (β = 0.12; p < 0.05) compared to evangelicals. Age maintains a strong and negative association with support for TRA while people in the West and East still show greater odds of supporting TRA when compared to people in the South. Finally, Americans who report a lot of trust in other races are much more likely to support TRA (β = 0.20; p < 0.001). These findings support our three hypotheses: Christian nationalism is negatively associated with support for TRA, and this effect is independent of measures for racial trust/distrust and religious commitment.

Figure 1 displays the predicted probabilities of supporting TRA at varying levels of Christian nationalism and religious practice. The predicted probabilities hold all significant effects from each dependent variable’s full model at their means. Graphing these relationships clearly demonstrates the powerful effect of Christian nationalism on the probability of support of TRA. At the lowest levels of Christian nationalism, with all other variables in the model held to their means (including religious practice), the probability of support for TRA is around 80 %, with 10 % lower odds at average levels of Christian nationalism, and over 20 % lower odds at high levels of Christian nationalism (<59 % probability). An opposite relationship exists between religious practice and odds of supporting TRA. With all other variables in the model held to their means (including Christian nationalism), at the lowest levels of religious practice, the probability of support is around 58 %. At average levels of religious practice, the probability of support jumps over 10 %. At the highest levels of religious practice, people have an 80 % probability of supporting TRA.

Fig. 1
figure 1

Predicted probability of supporting transracial adoption for religious practice and Christian nationalism. Note: All significant effects in the final model are set to their means when estimating predicted probabilities. For the religious practice line, Christian nationalism and all other significant effects are set at their means. For the Christian nationalism line, religious practice and all other significant effects are set at their means

Full size image

But certainly there are Christian nationalists who are also faithful practitioners of their faith. While Fig. 1 allows us to focus on the independent effects of Christian nationalism and religious practice, it raises an important question: are people who are Christian nationalists but also very active religiously (high levels of both measures) more or less likely to support TRA compared to people who are not active religiously and not Christian nationalists (low levels on both measures)? Figure 1 would suggest that the effects of the two variables counterbalance. By graphing the predicted probabilities when allowing both religious practice and Christian nationalism to vary, we can investigate whether there is a counterbalancing effect, net of all other effects. It will provide us an answer to whether or not the positive effects of religious practice overwhelm the negative effects of Christian nationalism on the likelihood of supporting TRA.

Figure 2 provides this answer. In it, we see that people who have low levels of Christian nationalism and low levels of religious practice have about a 70 % probability of supporting TRA. Those who are highly active religiously, but who also are Christian nationalists, have about a 65 % probability of supporting TRA. While the difference is small, the negative slope is undeniable. This suggests that the positive effects on the probability of supporting TRA due to religious activity are somewhat negated when the religiously active are also Christian nationalists.

Fig. 2
figure 2

Predicted probability of supporting transracial adoption at varying levels of religious practice and Christian nationalism. Note: All significant effects in the final model except religious practice and Christian nationalism are set to their means when estimating the predicted probabilities shown here

Full size image

Discussion and Conclusions

We theorized that within the racialized social system of the United States, Christian nationalist ideology is inextricably linked with notions of ethno-racial purity and separation. Moreover, we argued that Christian nationalism represents an implicitly ethno-racial ideology that is not merely a general proxy for racial prejudice or distrust but actually implies ethno-racial separation, nor is it a proxy for religious conservatism or commitment. Focusing on Americans’ attitudes toward transracial adoption and drawing on national survey data, our analyses affirm our theory. Net of other factors, Christian nationalism is strongly and negatively associated with support for TRA (Hypothesis 1), and this finding is robust even after the inclusion of socio-demographic and ideological controls, as well as controls for respondents’ reported trust in people of other races (Hypothesis 2), as well as their level of religious commitment (Hypothesis 3). The findings of this study support theories proposed by researchers on Christian nationalism that implicit within calls for America to “preserve” or “restore” its supposed Christian heritage are covert ideologies buttressing racial boundaries and notions of ethno-racial purity (Aho 2013; Goldberg 2006).

Before further discussing the implications of this study, some limitations should be acknowledged. First and primarily, the data are cross-sectional, and thus, causal arguments about the link between Christian nationalism and TRA attitudes must be made with caution. While this study is able to control for a host of relevant predictors of support for TRA consistent with previous research, future research would ideally draw upon longitudinal data, interviews, or participant observation to flesh out the social mechanisms at work in the relationship between Christian nationalism and Americans’ attitudes about race relations. Secondly, the dependent variable consists of a single-item measure of support for TRA with limited values, and thus, the extent to which the analyses can tap respondents’ feelings toward TRA as a relationship is limited (this is also unfortunately the case with our measure of trust in other races). Nevertheless, the findings of the analyses are strong and robust to relevant correlates, and we anticipate that more comprehensive measures of respondents’ opinion toward interracial family relationships like TRA will bear out similar results. Additionally, future research on this topic should examine attitudes toward other interracial relationships like interracial dating and marriage to test whether Christian nationalism evinces similar effects. Lastly, we acknowledge that our measure of Christian nationalism lacks precedent in psychometric testing. The Cronbach’s alpha coefficient for our measure (α = 0.88) is strong, however, and similar measures have been used profitably in studies examining prejudice and other social attitudes (Froese and Bader 2007; Froese and Mencken 2009; McDaniel et al. 2011). These factors, taken together with our broad findings focusing on TRA attitudes, help affirm the usefulness of our Christian nationalism measure.

These few limitations notwithstanding, our study extends the literature on religious nationalism and race relations in the United States. Fundamentally, our findings demonstrate a strong and robust association between adherence to Christian nationalism and Americans’ support for transracial adoption, even after controlling for trust in other racial groups; measures of religious identification, belief, and commitment; and other socio-demographic factors. Taken together with McDaniel et al. (2011), our central finding strongly suggests that movements on the religious right to protect or restore America’s so-called Christian heritage are likely not devoid of racial content, but rather are pregnant with notions justifying ethno-racial separation and social exclusion. Moreover, our results clarify that Christian nationalists are not merely distrustful of other races in a general sense or religiously conservative. If they were, controls for respondents’ trust in other races and the other religion measures in the multivariate models would render the Christian nationalism measure nonsignificant. Rather, Christian nationalism represents an implicitly ethno-racial ideology in and of itself and is neither epiphenomenal of religious commitment or a proxy for racial distrust, but influences Americans’ views toward interracial relationships like TRA above and beyond the effects of trust in other races and religious commitment separately.

Although Christian nationalism obviously contains elements of Christian belief, and certainly staunch Christian nationalists are likely to consider themselves religiously devout, our analyses affirm that the link between Christian nationalism and declining support for TRA is not due to religious devoutness. On the contrary, respondents who more faithfully practice their religious faith were actually more supportive of TRA. Religious orientations theory argues that the direction of the relationship between religious commitment and racial intolerance hinges on how individuals are oriented to their faith (Allport and Ross 1967; Gorsuch and Aleshire 1974; Kirkpatrick 1993; Lenski 1963). Allport and Ross (1967) theorized that the truly devout who have internalized their faith (an intrinsic orientation) would be less prejudiced than those who practice their religion for utilitarian purposes or as a matter of ritualism or institutional devotion (extrinsic orientation). Similarly, Lenski (1963) argued for an “orthodox” religious orientation, which emphasizes doctrinal traditionalism and tends to promote racial intolerance, and a “devotional” orientation, which emphasizes personal communion with God and internalizing one’s faith. This orientation tends to promote greater racial tolerance. Scholars have measured this devotional orientation using private religious practices such as prayer and sacred text reading, finding that greater religious devotion predicts favorable attitudes among whites toward racial integration and government intervention on behalf of the poor (Lenski 1963) as well as friendliness and lower suspicion of out-group members (Ellison 1993). Perry (2010, 2013) draws on religious orientations theory to explain the link between religious devotion and TRA attitudes. Although Perry reported that conservative Protestants tended to be less supportive of TRA compared to others, persons who more frequently engaged in religious practices like church attendance, prayer, and sacred text reading were more supportive of TRA. Following Lenski and others (Lenski 1963; Ellison 1993; Kirkpatrick 1993), frequent religious practice represents a “devotional” orientation that emphasizes the internalization of divine teachings on love and tolerance and thus promotes greater acceptance of intimate interracial relations, like TRA.

While this study has affirmed a strong connection between Christian nationalism and Americans’ racial views focusing on TRA, future research on Christian nationalism and race relations in the United States could ultimately look not only to Americans’ racial attitudes, but to the extent to which Christian nationalism shapes Americans’ actual engagement in interracial relationships (friendship, dating, marriage) and participation in racially diverse volunteer organizations (congregations, civic groups), providing more depth of understanding of the ethno-racial content of Christian nationalism and its implications for race relations in the United States.