INTRODUCTION
Are guns polarizing? A conventional answer to that timely question would look for evidence of bimodal chasms in policy preferences or ideological commitments related to guns. Consider that, on the one hand, a growing percentage of Americans—58% today, versus 41% in 1999—agree that gun ownership increases safety, a number that exceeds the roughly 30% who actually own a gun. On the other hand, nearly four-in-ten Americans disagree that guns enhance safety (Murray Reference Zhang and Reid2018) and six-in-ten also say that gun laws should be stricter (Pew Reference Zhang and Reid2018). Beyond policy preferences and ideology, however, there is another way of looking at guns and polarization that focuses on identity. Shanto Iyengar and his colleagues, among others, have identified affective polarization as a key manifestation of the deep divides in American politics today (Iyengar and Westwood Reference Iyengar and Westwood2014; Mason Reference Mason2018; Iyengar et al. Reference Iyengar2019). They argue that political divisions are not merely the result of policy or ideological disagreements, but rather reflect the divergence of deeply ingrained social identities that are hardened by the realities of intergroup conflict.
Our goal in this note is to take some initial steps at examining how social identity shapes the politics of guns in America. Our approach starts with two types of group attachment—religious and gun-owning—that is established in the literature as predictors of how Americans relate to guns and gun policy. We reconceptualize and measure those group attachments as forms of identity, and then test a distinctive gun-owner identity as a factor that interacts with religion. The analysis suggests a convergence of identities—religious and gun-owning—that appear to be mutually reinforcing. The role of religion in shaping gun identity is not as simple as a correlative of a general religious identity but rather is impacted by religiosity and the specifics on an individual's religious tradition. We use the comparison of the two major Protestant groups—evangelicals and mainliners—to analyze the effects of tradition and religiosity.
SOCIAL IDENTITY AND GUNS: BUILDING THEORY
Scholars have leveraged theories of social identity to explain a host of social phenomena, from consumer choices to political polarization. While these theories vary widely in their assumptions and applications, they share a similar focus on an individual's self-understanding and psychological investment as a member of a social group (Tajfel and Turner Reference Tajfel, Turner, Austin and Worchel1979). In the past few decades, political scientists have tapped these theories to examine how social identity shapes political attitudes and mobilization (Monroe, Hankin, and Van Vechten Reference Monroe, Hankin and Van Vechten2000; Kalin and Sambanis Reference Kalin and Sambanis2018). The resulting literature has become rich, increasingly sophisticated, and vast, including applications to policy-related beliefs (Conover Reference Conover1984; Citrin, Reingold, and Green Reference Citrin, Reingold and Green1990; Kinder and Winter Reference Kinder and Winter2001; Schmermund et al. Reference Schmermund, Sellers, Mueller and Crosby2001; Hurwitz, Peffley, and Mondak Reference Hurwitz, Peffley and Mondak2015; Cramer et al. Reference Cramer2017; Cramer et al. Reference Cramer2017), political participation (Leighley Reference Leighley1996; Greene Reference Greene1999; Schildkraut Reference Schildkraut2005; Miller and Conover Reference Miller and Conover2015; Achen and Bartels Reference Achen and Bartels2016), and even distinctive forms of political identity (Huddy Reference Huddy2001).
Given the prevalence and power of social identity as an analytical framework for politics, it is surprising how little scholars know about how social identity shapes attitudes about gun ownership or policy. Our goal is to examine the interrelation of two such identities: one related to religion, the other to gun-ownership itself. We draw from the extensive literature on religion-as-identity and a burgeoning literature focused on gun owners as a distinctive group (Lacombe Reference Lacombe2019). We argue that for many Americans, belonging to the category of “gun owner” has taken on the characteristic self-referencing qualities of social identity.
We know a great deal about the social characteristics of gun owners from the literature on gun culture in the United States (Hofstadter Reference Hofstadter1970; Wright Reference Wright1995; Haider-Markel and Joslyn Reference Haider-Markel and Joslyn2001; Kahan and Braman Reference Zhang and Reid2003; Abramowitz and Saunders Reference Abramowitz and Saunders2008; Spitzer Reference Spitzer2012). But it is one thing to say that gun owners share social characteristics and political interests; it is another to claim that they share an identity. Group attachment is an indispensable seedbed for social identity, but only under the right conditions. Merely owning a gun—fitting into the objective category of “gun owner”—is insufficient as an identity marker because it has no necessary impact on one's basic self-concept. For gun-ownership to operate as a social identity, owners must understand themselves as attached to and emotionally invested in a group. Moreover, because social identity is rooted in the dynamic of in- and out-groups (Tajfel and Turner Reference Tajfel, Turner, Austin and Worchel1979), owners would likely understand their attachment as a matter of intergroup contrast. They are psychologically committed to a group of gun-owners who define themselves partly in contradistinction to—and perhaps derogation of—opposed groups. That intergroup dynamic is generally felt acutely in climates of threat or perceived threat to one's imagined community, which heightens the salience and emotional resonance of group membership and has the power to propel group members into the collective political response (Zhang and Reid Reference Zhang and Reid2013).
Some social science hints at a connection between a “gun-identity” and attitudes about guns and gun-related policy. Joslyn et al. (Reference Joslyn, Haider-Markel, Baggs and Bilbo2017), for example, use gun ownership as a proxy for membership in two different gun “cultures” with opposed “clusters of values,” which they find is associated with different partisan vote outcomes (see also Celinska Reference Celinska2007). Some work has focused on social psychological constructs that are associated with identity. Joslyn's and Haider-Markel's (Reference Joslyn and Haider-Markel2017) study of responses to mass shootings uses attribution theory to argue that gun ownership increases self-serving attributions about the causes of gun violence and leads to resistance of gun control. Stroebe, Pontus Leander, and Kruglanski (Reference Stroebe, Pontus Leander and Kruglanski2017) draw from social-cognitive theories to show how threat construal shapes the likelihood of gun ownership. They argue perceptions of both diffuse (“it's a dangerous world out there”) and specific (“I'm likely to be assaulted at some point in my life”) threat increase the likelihood of buying a gun.
Yet these studies are only suggestive about an independent role for gun identity in shaping attitudes and behavior related to guns. Joslyn et al. (Reference Joslyn, Haider-Markel, Baggs and Bilbo2017) raise the question of whether gun ownership amounts to an indirect measure of an “emerging” political identity and Lacombe (Reference Lacombe2019) suggests that the National Rifle Association uses identity-language to inspire its members to action. But, as we have already noted, gun ownership alone does not confer an identity.
Like gun ownership, religion is an oft-cited, albeit less obvious, the source for attitudes about guns. Many gun studies focus on imprecise associations and collapse categories into a simplistic “Protestant” versus “non-Protestant” dichotomy. Wright and Marston (Reference Wright and Marston1975), Young (Reference Young1989), Little and Vogel (Reference Little and Vogel1992), and Cox, Navarro-Rivera, and Jones (Reference Cox, Navarro-Rivera and Jones2013), among others, have noted that over time, Protestants are more likely than other religious traditions to own guns and less likely to support gun control. These findings remain consistent, but these scholars provide either no theoretical explanation for the finding or treat it as an epiphenomenon of other factors. Moreover, few studies move beyond the broad religious category of “Protestant” into other measures of religion, including specific religious traditions or intensity of religious commitment—a puzzling omission, given the enormous literature on the political relevance of those specifics factors (see, e.g., Green et al. Reference Green, Guth, Smidt and Kellstedt1996; Steensland et al. Reference Steensland, Park, Regnerus, Robinson, Bradford Wilcox and Woodberry2000; Smidt Reference Smidt2013; Wielhouwer Reference Wielhouwer, Kellstedt and Smidt2009; Hertzke et al. Reference Hertzke, Olson, Den Dulk and Fowler2018). As Yamane (Reference Yamane2016) argues, religion has too often been little more than a narrowly defined control variable in studies of gun ownership and attitudes about policy rather than fully explored as its own theoretical puzzle.
In his more sophisticated analysis, Yamane (Reference Yamane2016) concludes that “Protestant” as a coherent category of affiliation is irrelevant to gun ownership, controlling for other factors. He does find, however, that evangelical Protestants and theological conservatives exhibit relatively high levels of personal handgun ownership, though the white evangelical category differs from other religious traditions in interesting ways. For example, there is no significant difference between white evangelicals and black Protestants in ownership levels, while mainline Protestants do indeed differ in their lower levels compared to evangelicals. Moreover, religious involvement is negatively associated with ownership, controlling for religious tradition and other factors.
We choose to focus on the specific subtraditions within Protestantism: evangelicals and mainliners. Unlike a simple Protestant-versus-non-Protestant dichotomy, the distinction between these subtraditions within Protestantism has proven a robust explanatory variable for myriad political behaviors and attitudes, including attitudes about gun policy. Our goal is to move beyond attitudes about the gun policy to how religion shapes a deeper sense of group identity around guns. More specifically, does a Protestant subtradition matter to the depth of gun-owner identity?
DATA AND METHODS
One of the serious challenges to the analysis of the role of social identity––both religious and gun-owning—is the lack of extant surveys with sufficient range of relevant questions. The Pew Research Center, however, provides us with questions that capture both gun owner identity and certain components of religious identity. As part of its American Trends series in 2017, Pew surveyed a total of nearly 4,000 respondents and a subsample of 1,269 gun owners.Footnote 1 The survey included questions about the nature and extent of gun ownership, views of gun-related policy, and demographic characteristics, religious tradition and religiosity among them.
Our first dependent variable of interest is gun owner identity, which is measured using the question, “How important is being a gun owner to your overall identity?” Respondents answered on a four-point scale from “Not at all important” to “very important.” Of the total, 41.79% of gun owners responded that gun ownership was either “somewhat important” or “very important” to their overall identity.
We utilize an additional dependent variable to capture other components of identity. First, we measure the commitment that gun owners may have to their gun ownership. The question asks gun owners to agree or disagree with the following statement: “I can never see myself NOT owning a gun.” Second, we use an original scale measuring the level of participation in gun owner activities. The scale is derived from the following five statements, in which respondents indicate the frequency of their participation on a four-point scale: visit websites about guns, hunting or other shooting sports, watch TV programs about guns or watch gun-oriented videos, listen to gun-oriented podcasts or radio shows, participate in online discussion forums about guns, and attend gun shows. The resulting additive index ranges from 1 to 15, with the questions loading onto a single factor with an Eigenvalue of 2.15 and a Cronbach's Alpha of 0.79.
Our third area of interest is attitudes toward gun policy and policy's effectiveness in preventing gun violence. We measure this opinion using the question in which respondents are asked which statement comes closer to their view: “People who want to kill or harm others will find a way to do it whether they have a gun or not” or “People who want to kill or harm others are less likely to do it if they don't have a gun.” In total, 68.13% of the sample gravitated toward the former. This question is asked of both gun owners and non-gun owners and we include all respondents in our analysis.
Religion in various forms serves as a key set of predictors. We utilize religious identity within the Protestant tradition (mainline Protestantism and evangelicalism) in our analysis. The Pew dataset does not include a full battery of denominational indicators, which precludes a full “reltrad” analysis (Steensland et al. Reference Steensland, Park, Regnerus, Robinson, Bradford Wilcox and Woodberry2000), but it does include the standard question, “Would you describe as a ‘born-again’ or evangelical Christian, or not?” While not an ideal measure (but see Burge and Lewis Reference Burge and Lewis2018), we coded a “yes” to this question as “evangelical,” and all other Protestants as “mainline.” We utilize dummy variables for mainline and evangelical identification, with all other respondents as the reference category, and a dummy variable that compares evangelicals to mainliners alone (evangelical = 1). We control for race through a dummy variable in each model (white = 1). We additionally consider the strength to which respondents identify as a religious person (as indicated by their self-placement on a five-point scale). We also controlled for several demographic and geographic variables suggested by the literature on guns: gender, party identification, rural upbringing, age, and education. We use both logistic (when dependent variables are dichotomous, e.g., perceptions of discrimination, commitment to owning guns, and attitudes toward gun policy), and ordered logistic (when dependent variables are ordered response categories, e.g., gun owner identity and the gun activity scale) regression to investigate the relationship between gun ownership, gun owner identity, religious tradition, and attitudes toward gun policy.
RESULTS
At a first descriptive glance, neither mainline nor evangelical Protestants appear more likely than the other to claim a gun identity. Figure 1 shows evangelicals slightly edging out mainliners on the gun-identity scale, but not at the level of statistical clarity that we might expect.
Figure 1. Data: Pew 2017.
Following Yamane (Reference Yamane2016), however, we suggest that this relationship is mediated by the strength of religious identity. Figure 2 shows that evangelicals report similar levels of gun owner identity regardless of religious identity strength; mainline Protestants, however, are impacted by how strongly they identify as religious. Mainline Protestants who identify strongly as a religious person report higher levels of gun owner identity than mainliners who do not strongly identify as religious; religiosity acts as intensifying factor regarding gun owner identity within Protestant subtraditions.
Figure 2. Data: Pew 2017.
Tables 1–3 test this descriptive suggestion in several ways. Table 1 analyzes gun identity as a function of religious tradition and religiosity, first by examining evangelicals and mainliners separately against all other religious groups and various controls (Models 1 and 2), and then by isolating Protestants and comparing evangelicals and mainliners to each other (Models 2 and 3). The results both confirm and challenge the descriptive analysis. Evangelical Protestants score higher on the gun-identity scale than all other religions, while mainliners score lower than the others. Given Models 1 and 2, it is not surprising, when we constrain the sample to Protestants only, that evangelicals are more likely to score high on the scale than mainliners (Model 3, where evangelical = 1 and mainline = 0). That evangelical-versus-mainline pattern holds even when we introduce a religiosity variable, though the shifting levels of significance suggest that religiosity does mediate the gun-identity scores of one or both subtraditions.
Table 1. Religion and Gun Owner Identity
t statistics in parentheses.
DV: Gun Identity, Data: Pew 2017.
Model: Ordered Logistic Regression.
* p < 0.05, ** p < 0.01, *** p < 0.001.
Table 2 explores two additional components of gun owner identity: commitment to lifelong ownership and participation in gun-related activities. Here, we compare evangelicals to mainline Protestants only. Commitment to gun ownership is statistically significant and in the expected direction—i.e., evangelical gun owners are more likely to claim that they will maintain their gun ownership for a lifetime, even controlling for a quite powerful predictor in partisan identification. Evangelical and mainline Protestants are not distinguishable when it concerns their engagement in gun-related activities. Religious identity strength does not matter for either component.
Table 2. Religion and Additional Identity Measures
t statistics in parentheses.
DV: Additional Gun Identity Components, Data: Pew 2017.
Model: Logistic Regression (Model 1) and Ordered Logistic Regression (Model 2).
* p < 0.05, ** p < 0.01, *** p < 0.001.
Table 3 examines the impact of religious identity on a specific assumption about human motivation with implications for how respondents understand the effectiveness of the gun policy, i.e., the view that people will find a way to kill regardless of gun access policy. Even controlling for key variables, evangelicals are more likely than mainline Protestants to suggest that gun policy will fail to prevent violent death because people will find ways to kill whether or not they have access to a gun. It is remarkable that even gun identity itself, which we include in this model as an independent variable, does not impact the significance of evangelical tradition.
Table 3. Religion, Gun Owner Identity, and Policy Attitudes
t statistics in parentheses.
DV: Gun Violence will Occur Regardless of Legislation, Data; Pew 2017.
Model: Logistic Regression.
* p < 0.05, ** p < 0.01, *** p < 0.001.
In Models 2 and 3, we assess the impact of the interaction between religious identity, the strength of that identity, and gun owner identity on policy attitudes. For evangelicals, it appears not to matter the strength of religious identity nor the strength of gun owner identity; mainline Protestants, however, are affected. Figure 3 allows us to better explain these results. Higher religious identification among evangelicals only leads to a stronger belief that gun laws are ineffective among those with a high gun owner identity (4 = strongest gun identity). This visualization applies to guns a standard distinction between Protestant subtraditions, namely, that mainline Protestants tend to focus on addressing public problems as matters of structure (and therefore are relatively open to legal interventions), while evangelicals tend to focus on addressing problems through campaigns to change individual attitudes and behaviors (hence the campaigns to change the “hearts and minds” of the people).
Figure 3. Data: Pew 2017.
DISCUSSION
In a general way, the specific Protestant identity of the gun owner does indeed matter to whether they claim a gun identity. Evangelicals are more likely than non-evangelicals to say gun ownership is important to their overall identity; mainline Protestants are less likely than their counterparts to report the same ownership-identity pattern. Moreover, when the sample is constrained only to Protestants, evangelicals are more likely than mainliners to connect ownership to identity. The patterns hold despite a battery of political and demographic controls, as well as self-described intensity of religious commitment. The role that the Protestant subtradition plays in gun-ownership-as-identity is not a mere epiphenomenon of party identification, rural upbringing, or race.
The measures that focused more on the self-assessment of commitment and behavior reveal mixed results. On the one hand, evangelical Protestants are more likely than mainliners to report that they will be life-long gun owners, and in fact, the only other statistically significant factor in that self-report is Republican party identification. On the other hand, neither measure of religion—Protestant subtradition nor levels of religiosity—appears to have an impact on the level of gun-related activity.
While we see some unexpected directions in the results, in general, we can say that evangelical and mainline Protestants have different responses to guns and gun identity, and that the strength of religious identity has a different impact on gun policy attitudes; but, the patterns in Table 3 warrant further elaboration. Whereas strong identification as a religious person led to increased gun owner identity, it counterintuitively leads to less-extreme attitudes toward gun policy. In other words, regardless of the intensity of gun owner identity, mainline Protestants who identify strongly with their religion are less likely to believe that gun laws are ineffective in reducing gun violence, which favors greater gun control. The same is not true for evangelicals, who show strong opposition to the idea that gun laws can reduce gun violence.
CONCLUSION
In 2008, then-candidate Barack Obama worried aloud that some embittered citizens “cling to guns or religion” as a salve for their frustrations. The comment flared into a controversy, with his rival Hillary Clinton stoking the flames. But our analysis suggests that Obama might have been on to something, at least in connecting two sources for identity that do matter to how citizens understand politics and policy.
One of the key questions from these analyses is the precise mechanism that would explain differences between gun identity among different religious identities, especially among Protestants. Gun identity does appear to have an amplifying effect in attitudes about policy, which social identity theory would suggest is rooted in gun owners' self-concept and intergroup contrast. We do not have specific survey questions to dive more deeply into social-psychological dispositions about guns or perceptions of intergroup relations. Following Iyengar et al. (Reference Iyengar2019), we imagine future research employing feeling-thermometer scales or quasi-experimental designs to explore the affective dimensions of membership in “gun-ownership” as a distinctive social group.
The mechanisms that could explain the different attitudes of mainline and evangelical Protestants are myriad. It is clear, however, that the differences are not simply a reflection of different beliefs about God, salvation, sacred texts, or outreach; if these beliefs matter, they are mediated through the traditions' different tendencies in political attitudes and values, including their relative openness to change at the structural versus individual level. We also suspect that the differences have much to do with the nature of their differing social networks and behavioral patterns within the traditions. The analysis here hints in that direction.
