Some fear, more loathing? Threats and anxieties shaping protective gun ownership and gun carry in the U.S.

Abstract

Given notable recent spikes in gun purchases in the U.S., we revisit the ‘fear and loathing’ hypothesis of firearm demand by (1) establishing how crime/victimization fears are shaped by broader economic, cultural, and racial status anxieties (those emerging from group status threats [loathing]) and (2) illustrating how both fear and loathing matter for protective gun ownership and gun carry (among owners), and openness to future protective ownership among non-owners.

Using data from a nationwide survey of adults in the U.S. (n = 2,262) collected in 2019, we find that fears of crime and victimization are often more strongly associated with status anxieties than with safety threats. Both status anxieties and victimization are associated with protective ownership and carry. Among non-owners, those higher in cultural anxiety are especially likely to be open toward future protective gun ownership. This study illustrates the multidimensional fear-guns link, wherein both status-related threats and victimization-related fears shape why individuals own guns, and how they use guns.

In March 2020, as the COVID-19 global pandemic prompted lockdowns across the U.S., a record-breaking 3.7 million firearm background checks were completed; this record was broken only a few months later, with 3.9 million firearm background checks during the widespread social unrest that unfolded following the murder of George Floyd by a law enforcement officer (Lee and Chabria 2020; Levine and McKnight 2020). While the reasons behind such spikes in demand¹ are complex, both events sparked a climate of uncertainty, and uncertainty (even if broad/abstract) is inherently threatening (Caliso, Francisco, and Garcia 2020; Maftei and Holman 2020). Threats, in turn – whether real or perceived – can spur defensive responses (Bennett, Dickmann, and Larson 2018; Tanovic, Gee, and Joormann 2018). Defensive firearm acquisition in response to threat is consistent with the shifting cultural meaning of firearms in the U.S. over the past few decades. Firearm ownership has transitioned away from hunting, sports, or recreation, towards personal protection and safety (Yamane 2017), with recent survey data showing that 67% of owners list ‘protection’ as a major reason they own a gun (Gramlich and Schaeffer 2019).

Emphasis on personal protection (from ‘ordinary’ crime/victimization to extreme events [e.g., mass shootings, terrorist attacks]) – i.e., the ‘good guy with a gun narrative,’ – has been a key theme in messaging from the National Rifle Association (NRA) and advertisements in major gun magazines (Lacombe 2019; Yamane, Ivory, and Yamane 2019). Although crime rates have generally been declining since the 1990s, many Americans remain fearful, with public perceptions of crime often out of step with actual patterns (Gramlich 2016, 2020). This past year Gallup polls did observe a notable increase in the percentage of Americans saying there is more crime in the U.S. now than there was a year ago (78%); however, perceptions of increased crime in one’s local area continued its trend downward (McCarthy 2020). In fact, the gap between the share of Americans believing crime is up nationally compared to those saying it is up locally (78% vs. 38%) is the widest gap Gallup has ever recorded (Gramlich 2020). Thus, if both actual crime and fear of local crime have been decreasing while protective gun ownership remains high, one must ask from what, exactly, is it that Americans are seeking protection? What is driving Americans’ crime/victimization fears? Why were there record spikes in gun sales alongside a global health crisis and during nationwide protests against racial injustice? To answer such questions, this study pursues two key objectives. First, we examine how (and which) threats spur fear – both broad crime concerns and more proximate victimization fears. Next, we explore how fears and/or threats shape protective gun behaviors and gun interest. By illustrating the links between immediate fears of and experiences with crime and victimization and broader social, cultural, and economic threats, we interrogate the ways in which both may shape why individuals own guns, and the protective ways in which they may use those guns.

Background

Although criminological attention to firearms largely focuses on gun violence and victimization, such investigations should have, at their foundation, an understanding of gun demand. Demand is a precursor to availability, and firearm availability facilitates harm (Braga et al. 2021). Indeed, an average of 100 Americans die from gun violence each day; over one-third of gun deaths are homicides (which disproportionately affect communities of color); the majority of intimate partner homicides are perpetrated with guns; and the second-leading cause of death of children under age 18 is gun-related (Giffords Law Center 2021). According to Mother Jones’ open-source database on mass shootings, over three-quarters of the guns used in U.S. mass shootings since 1982 were obtained legally (Follman, Aronsen, and Pan 2021). In 2019, nearly three quarters of homicides involved firearms; of those, 62% were handguns (FBI 2019).

Advancing policies that reduce gun violence and injury requires focusing on how (and why) people live with guns, not just how they are harmed by them (Metzl 2019b, p. 2). Gun scholarship to date has centered on several different correlates of firearm ownership: socialization; fear, victimization, and collective security concerns; and prejudice toward/proximity to racial/ethnic minorities (Cao, Cullen, and Link 1997; Celinska 2007; Filindra and Kaplan 2016; Jiobu and Curry 2001; Kleck et al. 2011; Schreck et al. 2018). Yet often overlooked is how such factors operate in conjunction with broader group attitudes and identities, which themselves may be vulnerable to threat (see Isom Scott and Andersen 2020; Steidley and Trujillo 2020). Given the shift toward protective ownership amidst generally declining crime, criminologists should consider the complexity of factors shaping Americans’ sense of threat, and how such threats may spur gun demand. One early explanation from Wright and colleagues (1983) may be a good place to start.

Fear and loathing

In their synthesis of early gun scholarship that followed the social unrest and racial tensions of the Civil Rights movement, Wright, Rossi, and Daly (1983) argued that the work to date speculated a ‘fear and loathing’ hypothesis of firearm demand, wherein gun ownership was shaped by fears of crime and a sense of loathing toward groups threatening the social order. However, despite noting the plausibility of the fear and loathing hypothesis, Wright, Rossi, and Daly (1983) concluded that work had not demonstrated its effect. Subsequent studies conceptualized ‘loathing’ in terms of racial prejudice, linking fear of crime and racial prejudice to firearm ownership in general (Delmas and Bankston 1993) and handgun ownership specifically (Kleck and Kovandzic 2009; Young 1985). However, far fewer studies have considered how this loathing component of the original hypothesis may capture broader threats to social order (beyond racial prejudice) that inform fear, shaping not just firearm demand but also behaviors like gun carrying. We see this as an area in need of attention. That is, we may need to recognize that a sense of loathing toward groups threatening the social order involves both the group(s) and the threat.

Group threat and racial resentment

The presence and/or perception(s) of racial minorities is deeply intertwined with presumptions of threat in U.S. culture. Blumer’s (1958) work on racial prejudice and Blalock’s (1967) on discrimination provide an important orienting framework for connecting group threat and protective actions. For instance, in his foundational work, Blumer (1958) outlined four basic feelings emergent from the positional arrangements of majority and minority groups: superiority, alienation, proprietary claim, and fear of encroachment. Threat is felt in response to challenges to group position – which may come in various forms (e.g., an affront to group superiority; a challenge to group power/privilege; economic/political competition), and which often spur reactive social control efforts. Racial prejudice is one such ‘ … defensive reaction … a protective device [that] functions, however shortsightedly, to preserve to integrity and the position of the dominant group’ (Blumer 1958, p. 5). Blalock (1967) extended this work, noting that economic, political, and even status threats posed by minority racial groups may motivate the majority to engage in social control efforts, such as political discrimination (e.g., voter disenfranchisement), segregation, and threat-oriented ideologies (e.g., stereotypical representations/assumptions of Black criminal threat).

As Chiricos and colleagues (2020) noted, the identification of criminal threat as a consequence of minority group presence was critical for the application of group threat theory in criminology. Indeed, changing demographic context (often measured via increasing percent Black) has been linked to numerous punitive criminal justice outcomes, from police force size to sentencing decisions. The theorized process here is that increasing minority presence spurs the perceived threat of crime, which then informs punitive and aggressive social control response (Liska, Lawrence, and Benson 1981) in order to protect and reinforce dominant group power and privilege. There are two important issues here for connecting group threat theory to firearm-related behaviors. First, such reactive social control responses occur not only at the level of policy or institution, but the group threat/social control linkage can be observed at the micro-level as well, in individual attitudes, preferences, and behaviors (Chiricos, Pickett, and Lehmann 2020). Thus, considering social control responses as a means to mitigate/protect against threat (especially to physical safety), and combining this with Black’s (1983) notion of self-help (a form of social control that occurs when individuals take self-protective measures, see also J. D. Carlson 2012), we can view defensive gun ownership as a source of social control achieved via self-help (see also collective security perspectives, e.g., Schreck et al. 2018; Smith and Uchida 1988).

The second issue relevant for connecting group threat and personal firearm behavior is that threats may be real or perceived – that is, threat perceptions can exist fully independent of one’s proximate demographic surroundings and circumstances (Chiricos, Pickett, and Lehmann 2020). Indeed, this is captured in what Blalock (1967) termed ‘threat-oriented ideologies,’ which persist today via modern, laissez-faire racism (Bobo, Kluegel, and Smith 1997). Such ‘color-blind’ attitudes continue to view White majorities as ideal (Bonilla-Silva 2014), emphasizing values associated with individualism, and castigating people of color who are perceived as not living by traditional American values of independence and hard work, instead relying too much on the government for ‘special rights’ (Drakulich et al. 2020; Kinder and Sanders 1996).

Supporting the idea that such modern racism (e.g., racial resentment) continues to inform firearm demand, research shows that White legal concealed gun carriers often hold heightened fear of hypothetical Black offenders, and deep skepticisms of structural racial inequality in the United States (J. Carlson 2015a; Stroud 2012). O’Brien and colleagues (2013) linked racial resentment with gun ownership and support for gun rights (also Filindra and Kaplan 2016). Much of the cultural framing and literature produced for gun owner consumption (primarily intended for White audiences) reify such messages (Melzer 2009; O’Neill 2007), which may account for why White gun owners feel more physical, moral, and emotional empowerment from possessing firearms compared to other races (Mencken and Froese 2019). Indeed, self-reliance is a core component of gun rights messaging (Melzer 2009).

Extending loathing beyond race: group threat and status anxiety

While the ‘fear and loathing’ hypothesis focused initially on racial animosity and tensions as determinants for gun behaviors, ‘loathing’ may extend beyond racial prejudice specifically. Wright and colleagues’ (1983, p. 82) own description of the hypothesis was predicated on the assumption that the population is ‘ … arming itself as a hedge against a fearful and unknown future’ and ‘preparing to shoot one another to death over cultural, racial, ideological, or social disputes’ (emphasis added). Group threats that may spur firearm-related attitudes and actions oriented around self-help need not be limited to race-based threats. Although Blumer (1958) and Blalock’s (1967) frameworks focused on racial prejudice and discrimination, they can be applied to threats and challenges across many group hierarchies (Branscombe et al. 1999). Because individuals are motivated to maintain a positive social identity, we emphasize the positive values of all the groups with which we identify (Tajfel and Turner 1979); group status threats – and resulting anxieties and actions – can occur when we perceive any of our groups’ (presumed) relative higher status to be unstable or under threat from competing groups (Major, Blodorn, and Major Blascovich 2018).

There have been numerous structural and cultural changes in the U.S. in recent decades that represent threats (real or perceived) to several historically privileged groups. Changes such as gender equity, racial/ethnic heterogeneity, changing family structures, and expanding LGBTQ rights have occurred alongside diminishing economic opportunities for upward mobility, representing broader group status threats encompassing anxieties stemming not only from racial dynamics, but from challenges to ‘traditional’ (White) masculine hegemony and values. Such changes may translate into status anxiety for some (especially men), with detrimental consequences for health, safety, and well-being (Braman et al. 2005; McVeigh and Maria-Elena 2009). Recently economists and demographers have observed ‘deaths of despair’ – increased mortality among non-Hispanic White adults that began at the turn of the 21^st century (when mortality among all other groups was decreasing) (Case and Deaton 2015, 2017). Such deaths – driven in part by firearm suicide, and observed among both young and middle-aged men (Monnat 2017) – have been attributed to deteriorating economic conditions (e.g., stagnant wages, loss of manufacturing jobs, Case and Deaton 2017; Cherlin 2014) and fears of cultural displacement (Carrasquillo 2019; Hochschild 2018; Siddiqi, Sod-Erdene, Hamilton, Cottom, & Darity Jr, 2019) – that is, to ways in which many middle- and working-class, primarily White Americans feel they are losing ground (Boehme and Isom Scott 2019; Graham and Pinto 2017; Kimmel 2017; Meara and Skinner 2015) and falling behind (Metzl 2019a; Wuthnow 2019).

A growing body of work links economic insecurity and anxiety to gun ownership, gun use, and attitudes toward guns and gun control. For instance, Mencken and Froese (2019) found that gun owners who had experienced or feared experiencing an economic set-back (e.g., losing a job, facing home foreclosure) viewed guns as morally and emotionally empowering, asserting higher levels of agreement with statements such as ‘owning a gun makes me feel responsible; . . ..in control; … confident; … valuable to my community.’ Gun owners who felt more empowered by their guns were more likely to view guns as instruments to solve problems, make communities safer, and improve one’s quality of life. Indeed, Carlson (2015a) shows how declining economic opportunities constrain access to the traditional ‘breadwinner model,’ motivating some men to adopt a ‘citizen protector’ model wherein ‘ … guns provide an alternative means of claiming masculine duty, authority, and dignity in the household’ (J. Carlson 2015a, p. 97; Stroud 2012).

Economic well-being and status are key components in constructions of masculinity. Townsend (2002) argues that breadwinner status is one of the four central pillars (along with homeownership, marriage, and fatherhood), comprising the ‘package deal’ of masculinity. When one of these pillars is threatened or unavailable, men may ‘repackage the deal’ (J. Carlson 2015b; Stroud 2012; Townsend 2002), renegotiating their masculine identities by doubling down on attainable aspects of masculinity via a number of compensatory behaviors, ranging from increased parental engagement to increased gun acquisition (Cassino and Besen-Cassino 2020). Work examining non-gun owners supports this as well – e.g., adherence to traditional masculine ideals (e.g., a man should be physically tough) was linked with openness to future protective gun ownership among both men and women (Warner 2020). In a recent study testing the provider-to-protector shift, economic insecurity was associated with protective gun ownership only among those most deeply invested in stereotypical masculine ideals, and this pattern held for both men and women (Warner et al. 2021).

Anxieties emerging from such economic threats (real or perceived) are also informed by or operate alongside cultural anxieties – resentment at the ‘line-cutters’ (Hochschild 2018): women, racial/ethnic minorities, immigrants who are seen as getting ahead economically, via government intervention, at the expense of predominantly White, older, working-class men (Boehme and Isom Scott 2019; Drakulich et al. 2020; Metzl 2019a; Mutz 2018). Mencken and Froese (2019, p. 6) suggest that ‘White men are searching for positive identities in a society that they perceive to be turning against them.’ Utilizing general strain theory (GST), Isom Scott and Andersen (2020) recently described perceived anti-white bias (‘whitelash’) as a status-based strain prompting corrective action (anger and offending). Kimmel (2017) offers an even more pointed view, describing aggrieved entitlement among ‘angry white men’ who are no longer able to reap the expected benefits of their diligence and hard work, as the American Dream becomes increasingly unavailable (see also Bridges and Tober 2016). Marginalized groups, presumed to have usurped the success and status to which dominant groups have been accustomed, are blamed wholesale by members of the dominant group, thus further fueling potential alternative strategies to assert one’s dominance. It is in this context that the increasing expansion of firearm-related self-defense has emerged as a response to the ‘collapse of spaces of domination’ (Kautzer 2015, p. 174), and a means to reconstruct moral identity, autonomy, and authority that is particularly observed among White men (Mencken and Froese 2019). Therefore it is not just that economic opportunities are being lost, but it is also who is perceived as responsible for such losses and for threatening the status of dominant groups (Boehme and Isom Scott 2019; Kimmel 2017).

Indeed, gun advocates have historically framed threats to gun rights as challenges to traditional American identity and masculinity from threatening forces such as immigration, feminism, and LGBTQ movements (Bhatia 2019). Lacombe’s (2019) analysis of almost 80 years of NRA literature and letters to editors in major U.S. newspapers shows that gun ownership has become a social identity constructed over time by gun rights advocates (e.g., framing owners as patriotic, law abiding, ordinary citizens), mobilized for political action via portrayals of threat from ‘big city, urban, elitist’ politicians (Lacombe 2019, p. 1348; Lacombe, Howat, and Rothschild 2019; Melzer 2009). Warner and Ratcliff (2021) recently observed that among gun owners, both cultural anxiety and masculine ideals were linked with gun identity and empowerment.

Current study

Protective gun acquisition may provide ‘symbolic empowerment’ against a variety of forces perceived to be threatening (Jiobu and Curry 2001): lack of confidence in the federal government, presumed inadequacy of law enforcement, insecurities about crime rates, and/or a limited access to or faith in social institutions (Gresham and Demuth 2019; Mencken and Froese 2019; Warner 2020; Warner and Thrash 2020). It may be a response to both a specific perceived threat of assault and a diffuse threat of a dangerous world (Stroebe, Leander, and Kruglanski 2017). Yet what makes the world (appear) dangerous, however, may be as much about crime (specifically) as it is about abstract anxieties over social, economic, and demographic change (Britto 2013; Hirtenlehner and Farrall 2013). Access to weaponry is very much seen as an ‘appealing hedge’ (Wright, Rossi, and Daly 1983, p. 102) against future uncertainty and/or unspecified threatening others. Understanding gun ownership and gun behaviors requires identifying and understanding the multiple cultural and social underpinnings of fear and threat in America. Attention to the intersection of racial animus, economic hardship, and various group status threats has been missing from criminological analyses of both fear of crime and gun ownership/gun behaviors. This is a serious limitation, as group threat/status anxieties may shape both firearm ownership/demand (J. Carlson 2015a; Cassino and Besen-Cassino 2020 ; Mencken and Froese 2019; Steidley and Kosla 2018; Stroud 2012; Warner 2020) and firearm violence (Bridges and Tober 2016; Kalish and Kimmel 2010).

Despite Wright, Rossi, and Daly (1983) initial skepticism that ‘fear and loathing’ would influence firearm behaviors, guns are indeed a cultural symbol of reassurance in the face of threats to (and anxieties about) race, ethnicity, nativity, gender, and social class. Racial animus may shape motivations around firearms among some who feel non-White racial groups have unfairly gained success under progressive policies, while other identity threats (e.g., gender, political) may also spur status anxieties shaping protective motivations. Drawing attention to racial resentment and cultural/economic anxiety is not to suggest that any one of these is the most important driver of gun behavior, or to suggest that most or all gun owners harbor racial resentments, or experience other broad anxieties. Rather our objective here is to revisit and extend the ‘loathing’ aspect of the ‘fear and loathing’ hypothesis to better understand social, cultural, and even emotional factors that may be working independently or together to shape crime/victimization fears and, in turn, inform gun ownership motivations.

Methods

Data are from two surveys administered in January 2019 (Study 1) and March 2019 (Study 2) to a crowdsourced nationwide sample of adults (18 and older) in the United States. The study design and protocol received Institutional Review Board approval for human subjects’ research prior to data collection. Participants were recruited from Amazon’s Mechanical Turk (MTurk) platform. MTurk ‘workers’ voluntarily sign up to participate in various ‘human intelligence tasks’ (HITs) for money. The use of MTurk in the social sciences has increased significantly since 2010, and although the data are not nationally representative, comparison studies have found MTurk samples to be more representative than college student samples, college town community samples, and other online sources (Chandler and Shapiro 2016). MTurk samples have been used for observational studies across several topics, including gun attitudes (e.g., Lacombe, Howat, and Rothschild 2019; Logan 2019; Warner 2020). Data collection procedures are detailed in Appendix A.

To improve data quality, and following best practices of other MTurk research (Pickett, Roche, and Pogarsky 2018), participation was restricted to U.S. workers, those with approval ratings on prior HITs of at least 95 percent, and those having completed at least 5,000 prior HITs. The rate of missing data for both surveys – from survey breakoff and item nonresponse – was low. For Study 1, 1,610 workers began the survey; 1,592 (99%) completed it. For Study 2, 898 workers began the survey; 879 (98%) completed it. The sample for both studies combined was n = 2,471. We exclude from this sample 16 respondents with missing data on the outcome and/or key predictors and 193 suspected duplicate respondents (see Appendix A), leaving a final analytic sample size of n = 2,262.

MTurk draws participants from across the U.S., but it is not meant to be nationally representative. The study sample, however, is comparable to Pew Research Center (2017) nationally-representative American Trends Panel (ATP), with respect to gun-ownership, gender, and self-reported region of residence; deviations are in the expected directions for an internet convenience sample (e.g., the MTurk sample contains more middle-aged persons [30–49], more Whites; for a Table and full discussion of these differences, see Appendix B). Our purpose is not to draw generalizable conclusions, but to explore associations between a broader array of correlates than has yet to be examined in gun research. Importantly, the current survey is the only one to include all of the measures described below, and we position this study and findings as an important first step toward unpacking the complexity of fear, status threat, and gun owning behaviors – we return to this in the Limitations.

Measures

The objectives of this study investigate two types of outcomes: (1) crime/victimization fears and perceived threats and (2) gun behaviors/attitudes. Four measures capture concerns about crime and victimization. Fear of Victimization is a mean-rating scale based on responses to the six items listed in Table 1 (e.g., ‘how often do you worry … someone will break into my home;’ ‘ … being attacked by someone with a gun’ [responses ranged 0 = never to 4 = often]) (Gainey, Alper, and Chappell 2011; Henson and Reyns 2015). Fear of Crime is based on responses to ‘I worry that crime in the U.S. is increasing.’ Perceived Risk of Victimization (Rountree and Land 1996) is a three-item index of respondents’ assessment of the safety of their home and neighborhood (see Table 1). Belief in a Dangerous World (Altemeyer 1998) is from the question, ‘Compared to 50 years ago, how safe would you say the world is?’ (responses 0 = very safe to 4 = not safe at all). These sets of items are consistent with prior work on fear, and their specific wording is taken from the Pew Research Center’s American Trends Panel.

Table 1.

Operationalization of indicators and indices, and original sample means per item.

Survey Items	Mean
Fear of Crime
How often do you worry about the following? (responses 0 = never – 3 = often):
1. I worry that crime in the U.S. is increasing.	1.645
Belief in a Dangerous World
1. Compared to 50 years ago, how safe would you say the world is? (responses 0 = not at all safe – 3 = very safe)	1.653
Fear of Victimization (mean-rating scale; alpha = 0.89)
How often do you worry about the following? (responses 0 = never – 3 = often):
1. I worry someone will break into my home.	1.342
2. I worry about being attacked by someone with a gun.	1.782
3. I worry about having my money/possessions take from me.	1.329
4. I worry about being the victim of a violent crime.	1.158
5. I worry about being the victim of a mass shooting.	0.883
6. I worry about being the victim of a terrorist attack.	0.767
Perceived Risk of Crime (mean-rating scale; alpha = 0.79)
Responses range 0 = very safe – 3 = not safe at all)
1. How safe would you say your local community is from crime?	0.762
2. How safe do you feel being outside alone at night in your neighborhood?	0.741
3. How safe do you feel being home alone?	0.349
Personal [Vicarious] Victimization (dichotomous indicator; 1 = any experience, 0 = no experience)
Please indicate whether or not you personally [a family member/someone else you know] experienced any of the following in the last 5 years:
1. Someone stole something from you [them].	0.260
2. Someone broke into your [their] home.	0.064
3. Someone tried to steal something from you [them] by force or by threatening you.	0.041
4. Someone threatened you [them] with a weapon.	0.046
5. Someone physically attacked you [them].	0.088
6. Someone shot you [them].	0.009
Police Distrust
1. How much would you say you trust the police? (responses 0 = a lot – 3 = not at all)	1.106
Economic Precarity (summated scale; alpha = 0.46)
1. In the past five years, have you or someone from your household lost their job because of the bad economy (1 = yes; 0 = no)?	0.443
2. In the last five years, have you or someone you know faced home foreclosure (1 = yes; 0 = no)?	0.212
3. Do you ever feel at risk of falling to a lower social class (1 = yes; 0 = no)?	0.179
4. When it comes to the availability of good jobs in America, would you say that America’s best years are behind us, or that our best years are yet to come (1 = behind us; 0 = yet to come)?	0.453
Economic Alienation (mean-rating scale; alpha = 0.78)
Indicate your level of agreement with the following statements (0 = strongly disagree – 3 = strongly agree)
1. The economic system is working against me.	1.440
2. The rich get richer while the poor get poorer.	2.165
3. Hard work and determination are not a guarantee of success for most people.	1.822
4. The political system is working against me.	1.550
Cultural Anxiety (mean-rating scale; alpha = 0.84)
Indicate your level of agreement with the following statements (0 = strongly disagree – 3 = strongly agree)
1. Poor people have become too dependent on the government.	1.472
2. We should restrict and control people coming into our country more than we do now.	1.484
3. Things have changes so much that I often feel like a stranger in my own country.	1.268
4. The American way of life needs to be protected from foreign influence.	1.413
5. Today, America is in danger of losing its culture and identity.	1.306
6. God has granted America a special role in human history.	1.075
Masculine Role Attitudes (mean-rating scale; alpha = 0.76)
Indicate your level of agreement with the following statements (0 = strongly disagree – 3 = strongly agree)
1. These days society seems to punish men just for acting like men.	1.205
2. It is essential for a man to get respect from others.	1.377
3. A young man should be physically tough, even if he’s not big.	1.226
Racial Resentment
Indicate your level of agreement with the following statements (0 = strongly disagree – 3 = strongly agree)
Today discrimination against whites has become as big a problem as discrimination against blacks and other minorities.	1.154

Note: For dichotomously-coded items, the mean is the proportion coded 1 on that item.

Three measures capture gun behaviors and attitudes: Protective Gun Ownership is measured among the full sample (n = 2,262), and is coded 1 for all respondents who indicated that they themselves personally owned a gun (responding ‘yes’ to the question ‘Do you currently own any guns (NOT including air guns, such as paintball, BB, or pellet guns)?’), and who identified ‘for protection’ as a major reason they did so (all other respondents are coded 0). Protective Gun Carrying is measured by responses to the question (asked only of current handgun owners): ‘How often, if ever, do you carry a handgun or pistol outside your home for your protection?’ Original response options (ranging from never to all of the time) were collapsed into 1 = at least some of the time, 0 = never. Among all non-owners (n = 1,380), respondents were coded as being open toward Prospective Protective Ownership (=1) if they indicated that ‘for protection’ was a ‘major reason’ (vs. ‘minor’ or ‘not a reason’) they would consider owning a gun (all other respondents coded 0). We use the term ‘prospective’ here because the survey did not gauge specific plans regarding future gun acquisition.

Additional correlates of both fear and gun ownership include Personal Victimization (past 5 year experience of any of 6 types of violent or property crime), Vicarious Victimization (family member and/or friends victimized), and Police Distrust (see Table 1 for exact item wording and coding).

‘Loathing’ – broader social, economic, and cultural status anxieties – is captured via four multi-item measures – Economic Precarity, Economic Alienation, Cultural Anxiety, and Masculine Role Attitudes (specific indicators and Cronbach’s alpha for each are listed in Table 1). Economic Precarity (Pew Research Center 2017) measures personal or vicarious experiences of economic loss (e.g., losing a job; facing home foreclosure). Economic Alienation (Pew Research Center 2017) measures cynicism toward the attainability of upward economic mobility (e.g., ‘the rich get richer while the poor get poorer’). Cultural Anxiety captures themes of antagonism toward social and cultural change observed in recent qualitative research (Hochschild 2018; Metzl 2019a; Wuthnow 2019) (e.g., ‘Things have changed so much that I often feel like a stranger in my own county’) – specific item wording is taken from the Public Religion Research Institute (PRRI) White Working Class survey (Public Religion Research Institute 2017). Masculine Role Attitudes is a mean-rating index of respondents’ agreement (0 = strongly disagree – 3 = strongly agree) with the stereotypical ideals about and perceived threats toward masculinity (Pleck, Sonenstein, and Ku 1993). Racial Resentment is captured via an item that gauges perceived ‘reverse discrimination’ (Isom Scott and Andersen 2020): ‘today discrimination against whites has become as big a problem as discrimination against blacks and other minorities.’ The multi-item racial resentment index often used in prior literature (Drakulich et al. 2020; Filindra and Kaplan 2016) was not included in Study 1; therefore we use the single item available across both studies (we return to this in the Limitations).

Finally, the analyses include a number of control variables: Gender (dummy variable for Men); Age (centered at 18); Race (dummy variable for White); Education (0 = less than high school – 5 = graduate degree); Political Affiliation (dummy variable for Republican); Income (0 = < $10,000–8 = ≥ $150,000); Region (dummy variable for South); and Rural (dummy variable for self-reported residing in rural area [distributions across other geographic regions shown in Appendix B]).

Analyses

The analyses proceed as follows: First, we examine the correlates of broad crime-related fears (fear of crime increasing and belief in a dangerous world). Because the proportional odds assumption was violated for these two measures, both were dichotomized at the highest category (often; not at all safe, respectively), and analyzed via binary logistics regression. More proximate fears – fear of victimization and perceived risk of victimization – are continuously-distributed indices, and analyzed via OLS regression. Next, we examine how fear and loathing shape protective gun ownership (among the full sample, n = 2,262), protective gun carry (among current handgun owners, n = 552), and prospective protective gun ownership (among those with no current or prior gun ownership, n = 1,380) via a series of binary logistic regressions.

Results

Sample descriptives

Descriptive statistics for the full sample and by gun ownership status are presented in Table 2. For parsimony, here we present descriptive statistics on current and former owners combined (n = 882), and contrast these with non-owners (n = 1,380). We present descriptive statistics disaggregated between current and former owners in the Appendix (Table C1). Overall, one quarter (26%) of all respondents own/owned a gun(s) for protection (among the current/former gun owners, 67% report owning for protection). Just under half of current handgun owners reported carrying their weapon outside the home for protection at least some of the time. Among the 1,380 non-owners, half (51%) were open to protective gun ownership in the future.

Table 2.

Sample descriptive statistics, overall and by gun ownership: means (standard deviations) and t tests (n = 2,262).

	Full Sample (n = 2,262)		Current/Former Gun Ownersa (n = 882 [39%])		Never Owners (n = 1,380 [61%])
Variables	Meanb	(SD)	Mean	(SD)	Mean	(SD)	Sig.
Outcomes
Protective gun ownership	0.260	–	0.667	–	–	–	–
Protective gun carryingc	–	–	0.467	–	–	–	–
Prospective protective gun ownership	–	–	–	–	0.513	–	–
Fears, Risks, and Threats
Fear of crime	0.160	–	0.211	–	0.127	–	***
Belief in a dangerous world	0.151	–	0.166	–	0.142	–
Fear of victimization	1.110	0.658	1.156	0.662	1.080	0.655	**
Perceived risk of victimization	0.617	0.552	0.584	0.546	0.639	0.555	*
Personal victimization	0.306	–	0.371	–	0.265	–	***
Vicarious victimization	0.630	–	0.701	–	0.585	–	***
Police distrust	1.106	0.954	0.985	0.937	1.183	0.957	***
Loathing: Status Anxieties
Economic precarity	1.286	1.109	1.371	1.172	1.232	1.064	**
Economic alienation	1.744	0.680	1.682	0.694	1.783	0.669	***
Cultural anxiety	1.336	0.759	1.591	0.749	1.174	0.721	***
Masculine role attitudes	1.269	0.753	1.465	0.721	1.144	0.746	***
Racial resentment	1.154	1.057	1.498	1.058	0.934	0.996	***
Control Variables
Gender
Men	0.504	–	0.523	–	0.491	–
Women	0.497	–	0.477	–	0.508	–
Age	41.384	12.156	43.329	12.306	40.141	11.898	***
Race/Ethnicity
White	0.809	–	0.878	–	0.765	–	***
Black	0.066	–	0.040	–	0.083	–	***
Hispanic	0.063	–	0.049	–	0.072	–	*
Other	0.062	–	0.034	–	0.080	–	***
Political party affiliation
Republican	0.373	–	0.509	–	0.286	–	***
Democrat	0.439	–	0.287	–	0.537	–	***
Independent/Other	0.187	–	0.204	–	0.177	–
Socioeconomic status
Education d	3.236	1.237	3.090	1.221	3.330	1.239	***
Income e	4.258	2.056	4.314	2.011	4.223	2.085
Geography
Region/Community Type
South	0.364	–	0.441	–	0.315	–	***
Town	0.273	–	0.295	–	0.259	–
Rural area	0.135	–	0.218	–	0.082	–	***

Notes:

^aContrasts between current and former gun owners are available in Appendix Table C1;

^bMeans for dummy variables can be interpreted as the proportion of the sample coded 1 on that indicator;

^cProtective carry reported among 552 current handgun owners (carrying was not asked of former owners);

^dCategorical variable ranging from 0 = less than high school to 5 = postgraduate;

^eCategorical variable ranging from 0 = less than $10,000 to 8 = $150,000.

^*p < 0.05,

^**p < 0.01,

^***p < 0.001 (two-tailed tests).

In terms of fears and safety threats, 16% of all respondents ‘often’ worried that crime in the U.S. was increasing, and 15% thought the world was ‘not at all safe’ compared to 50 years ago. Fear of crime was significantly higher among owners than non-owners (21% vs. 13%). Both fear of and perceived risk of victimization were fairly low in the full sample, but higher among gun owners. About 30% of the sample had personally been victimized (most having experienced theft); twice that reported knowing family/friends who were victims, with owners more likely than non-owners to report both personal and vicarious victimization. Regarding loathing, gun owners report more economic precarity (perceived/experienced economic instability) than non-owners, but economic alienation (negativity/cynicism toward the economic institution) is actually lower among gun owners. Cultural anxiety (e.g., negative attitudes toward the poor, immigrants; concern over American identity), endorsement of stereotypical masculine ideals, and racial resentment are all considerably higher among owners than non-owners (p < 0.001); supplemental analyses (see Appendix) show these latter three measures are highest among current owners, followed by former owners, and they are lowest among non-owners. These basic patterns have not yet been observed simultaneously in criminological firearm scholarship to date.

The control variables are patterned across ownership status as expected. Gun owners are more likely than non-owners to be White, and to identify as Republican. Owners report a lower education; they are more likely to live in the South and in a rural area.

Correlates of broad and proximal fears

Table 3 explores the correlates shaping fear, offering insight into the gap between perceptions about national compared to local crime, and the factors motivating high rates of protective gun ownership in the U.S. We first examine broad fears: fear that crime in the U.S. is increasing (Model A), and belief that the world has become more dangerous (Model B). Crime/victimization-related threats to safety are examined first (Model 1), followed by loathing indicators (Model 2); the final fully-adjusted model (Model 3) illustrates if threats, loathing, or both are associated with fear. As Model A1 shows, vicarious (but not personal) victimization is positively associated with fear of crime; surprisingly, police distrust is negatively associated with crime fear. None of these safety threats are associated with belief in a more dangerous world (Model B1), but almost all five loathing indicators are associated with crime fears (Model A2) and nearly all are associated with dangerous world beliefs (Model B2). In the final models, it is primarily loathing that remains associated with broad fears. For instance, each additional experience of economic precarity (e.g., job loss, home foreclosure) is associated with a 16% increase in the odds of often worrying that crime is increasing (exp[0.150]). The role of cultural anxiety is particularly strong, where a 1-unit increase is associated with 3.8 times the odds of crime fear. Those distrustful of police are more likely to think the world has become more dangerous; endorsement of stereotypical masculine ideals is unrelated to broad crime fears.

Table 3.

Binary logistic regression of association between victimization, anxiety, and broad crime fears (n = 2,262)^a.

	Model A: Fear of Crime
	Model 1		Model 2		Model 3a
	b	SE	b	SE	b	SE
Intercept	−1.783***	0.119	−4.554***	0.269	−4.274***	0.410
Safety Threats
Personal victimization	0.192	0.129			0.093	0.142
Vicarious victimization	0.368**	0.132			0.267	0.144
Police distrust	−0.174**	0.063			0.010	0.078
Loathing: Status Anxieties
Economic precarity			0.169**	0.058	0.150*	0.061
Economic alienation			0.417***	0.103	0.449***	0.114
Cultural anxiety			1.417***	0.144	1.322***	0.159
Masculine role attitudes			−0.348**	0.124	−0.230	0.131
Racial resentment			0.149	0.084	0.175*	0.087
−2 Log Likelihood	1965.758		1707.983		1660.055
	Model B: Belief in a Dangerous World
	Model 1		Model 2		Model 3a
	b	SE	b	SE	b	SE
Intercept	−1.963***	0.121	−4.590***	0.272	−4.393***	0.429
Safety Threats
Personal victimization	0.097	0.133			−0.052	0.144
Vicarious victimization	0.115	0.131			0.033	0.141
Police distrust	0.116	0.061			0.210**	0.076
Loathing: Status Anxieties
Economic precarity			0.236***	0.057	0.195**	0.060
Economic alienation			0.704***	0.106	0.600***	0.116
Cultural anxiety			1.148***	0.143	1.047***	0.159
Masculine role attitudes			−0.346**	0.124	−0.149	0.131
Racial resentment			0.006	0.086	0.003	0.090
−2 Log Likelihood	1915.522		1728.936		1661.835

Notes:

^aModel 3 also includes controls for gender, age, race/ethnicity, education, income, political affiliation, community type, and region (as listed in Table 2).

^*p < 0.05,

^**p < 0.01,

^***p < 0.001 (two-tailed tests).

Associations between safety threats, loathing indicators, and proximate fears (fear of victimization [Model A], perceived risk of victimization [Model B]) are examined in Table 4. Contrary to the findings for broader crime fears (Table 3), both personal and vicarious victimization experiences are positively associated with fear of victimization (Model A3); the former is associated with perceived risk of victimization (Model B3). But, importantly, victimization fears and perceived risks are also shaped by cultural anxiety. Indeed, comparisons of the partially standardized beta coefficients – which allows for a ranking of effect sizes in a given model (Menard 2011) – shows that cultural anxiety has the strongest association with victimization fear (stb = 0.240 [compared to stb = 0.114 for personal victimization and stb = 0.057 for vicarious victimization]; Model A3). Economic alienation also has a stronger association with victimization fear than does actual victimization experiences (stb = 0.126). For perceived risk of victimization (Model B), personal victimization, police distrust, and cultural anxiety have similarly-strong associations (the effects for economic precarity and alienation are attributable to gender and income). Altogether, the patterns observed here are consistent with arguments by Hirtenlehner and Farrall (2013), who described fear of crime as ‘the smallest common denominator for a series of entirely different – social, cultural, economic, ecological, and political – insecurities: a metaphor for everything that is dissatisfying about the new, altered living conditions in a rapidly changing globalized world’ (Hirtenlehner and Farrall 2013, p. 7; Stroebe, Leander, and Kruglanski 2017; Warner and Thrash 2020). This provides important context for understanding the myriad of factors driving gun demand and protective ownership, which we explore below.

Table 4.

OLS regression of association between victimization, anxiety, and proximate victimization fears/risks (n = 2,262)^a.

	Model A: Fear of Victimization
	Model 1		Model 2		Model 3a
	b	SE	b	SE	b	SE
Intercept	0.960***	0.027	0.526***	0.050	0.665***	0.078
Safety Threats
Personal victimization	0.197***	0.031			0.162***	0.030
Vicarious victimization	0.128***	0.030			0.078**	0.028
Police distrust	0.008	0.014			0.002	0.016
Loathing: Status Anxieties
Economic precarity			0.079***	0.013	0.065***	0.013
Economic alienation			0.124***	0.022	0.122***	0.023
Cultural anxiety			0.199***	0.030	0.208***	0.032
Masculine role attitudes			−0.022	0.026	0.006	0.026
Racial resentment			0.024	0.019	0.022	0.019
R2	0.036		0.087		0.163
	Model B: Perceived Risk of Victimization
	Model 1		Model 2		Model 3a
	b	SE	b	SE	b	SE
Intercept	0.455***	0.022	0.275***	0.043	0.620***	0.066
Safety Threats
Personal victimization	0.178***	0.026			0.140***	0.025
Vicarious victimization	0.033	0.025			0.025	0.024
Police distrust	0.079***	0.012			0.073***	0.013
Loathing: Status Anxieties
Economic precarity			0.043***	0.011	0.017	0.011
Economic alienation			0.101***	0.019	0.028	0.019
Cultural anxiety			0.050	0.026	0.101***	0.027
Masculine role attitudes			0.014	0.023	0.023	0.022
Racial resentment			0.023	0.016	0.024	0.016
R2	0.049		0.036		0.142

Notes:

^aModel 3 also includes controls for gender, age, race/ethnicity, education, income, political affiliation, community type, and region (as listed in Table 2).

^*p < 0.05,

^**p < 0.01,

^***p < 0.001 (two-tailed tests).

Fear, loathing, and protective gun ownership

As we argued above, there is at least some dimension of fear and/or threat implicated in protective motivation(s) for gun ownership, driving the need or desire to protect oneself and/or others. The current analyses illustrate that the source of such fears extends beyond immediate reactions to actual safety threats (victimization) or rational calculations of potential vulnerability (police distrust). Social, cultural, and economic status threats and anxieties fuel fear, which in turn may spur protective gun ownership. We next explore this process directly, first assessing how loathing and fear are associated with owning guns for protection (Table 5).

Table 5.

Loathing, fear, and protective gun ownership, binary logistic regression (n = 2,262).

	Model 1		Model 2		Model 3		Model 4
	b	SE	b	SE	b	SE	b	SE
Intercept	−2.132***	0.193	−2.021***	0.163	−2.574***	0.222	−3.168***	0.343
Loathing: Status Anxieties
Economic precarity	0.066	0.049			0.001	0.051	0.011	0.052
Economic alienation	−0.092	0.083			−0.139	0.090	−0.127	0.095
Cultural anxiety	0.473***	0.112			0.440***	0.123	0.300*	0.132
Masculine role attitudes	0.079	0.100			0.092	0.102	0.155	0.107
Racial resentment	0.282***	0.069			0.252***	0.070	0.211**	0.073
Fears, Risks, and Safety Threats
Fear of crime			0.284***	0.072	0.017	0.078	−0.013	0.080
Belief in a dangerous world			0.044	0.068	0.011	0.071	−0.066	0.074
Fear of victimization			0.296**	0.090	0.337***	0.094	0.407***	0.098
Perceived risk of victimization			−0.181	0.101	−0.206*	0.103	−0.132	0.108
Personal victimization			0.345**	0.111	0.323**	0.115	0.352**	0.117
Vicarious victimization			0.361**	0.112	0.371**	0.116	0.366**	0.118
Police distrust			−0.161**	0.055	0.014	0.062	0.064	0.066
Control Variables
Men							0.051	0.113
Age							0.012*	0.005
White							0.301*	0.148
Republican							0.178	0.130
Education							−0.147***	0.044
Income							0.066*	0.029
South							0.535***	0.105
Rural area							0.422**	0.145
−2 Log Likelihood	2409.020		2484.334		2362.307		2297.507

^*p < 0.05,

^**p < 0.01,

^***p < 0.001 (two-tailed tests).

As Model 1 (Table 5) shows, among the full sample, both cultural anxiety and racial resentment are positively associated with protective gun ownership. For instance, each 1-unit increase in cultural anxiety is associated with 60% higher odds of protective ownership. Not surprisingly, many of the fear indicators are also associated with protective ownership, in expected directions (with the exception of police distrust; Model 2). Such patterns largely persist when loathing and fear are analyzed together (Model 3). Individuals who are fearful of or have had experience with victimization are more likely to own a gun primarily for protection; but persons experiencing greater cultural anxiety or harboring a racially resentful attitude are also more likely to be protective owners. In the fully-adjusted model (Model 4), racial resentment and cultural anxiety remain significant. That is, individuals expressing greater agreement that ‘discrimination against whites’ has become a problem are more likely to report protective gun ownership, and both this effect (stb = 0.123) and the effect for cultural anxiety (stb = 0.126) are greater magnitude (stb = 0.123) than the effect for personal victimization (stb = 0.089). These findings suggest that both fear (of victimization) and loathing shape protective gun ownership. This should also be considered in light of the ways in which loathing and fear are themselves connected (as illustrated above). In fact, the effect of fear of crime on protective ownership observed in Model 2 is explained completely by its antecedent, cultural anxiety.

Fear, loathing, and protective handgun carry

Moving beyond identifying the correlates of protective gun ownership, we explore the factors associated with how individual owners use their guns. Table 6, 7 examines carrying for protection outside the home at least some of the time (among current handgun owners). Note, respondents were specifically asked about carrying ‘ … outside your home for your protection’ – this measure is not capturing carrying solely for transport purposes (e.g., to/from work, a shooting range, etc.). Economic precarity and racial resentment (Model 1) are positively associated with protective gun carry, and the effect of precarity remains when fear and loathing variables are examined together (Model 3). Victims (Model 3), have 1.5 times the odds of protective carry, compared to non-victims. Because one’s motivation for ownership may shape how owners use guns, the final model controls for protective ownership (the outcome examined in Table 5). In the fully-adjusted model (Model 4), only economic precarity remains associated with a higher likelihood of protective carry among current handgun owners, even after accounting for owning one’s gun for protection in the first place. Each additional increase in precarity is associated with a 20% increase (exp(0.186) in the odds of protective carry. The significance of economic precarity observed here aligns with Carlson’s (2015a) observation of the emergence of ‘citizen protectors’ among economically precarious gun owners in Michigan – a pattern she infers from her in-depth ethnographic work, and which we support quantitatively here.

Table 6.

Loathing, fear, and protective handgun carry, binary logistic regression (n = 552 current handgun owners).

	Model 1		Model 2		Model 3		Model 4
	b	SE	b	SE	b	SE	b	SE
Intercept	−1.127**	0.351	−0.588*	0.290	−1.254**	0.417	−2.211**	0.700
Loathing: Status Anxieties
Economic precarity	0.211*	0.085			0.175*	0.089	0.186*	0.093
Economic alienation	−0.231	0.143			−0.327*	0.156	−0.311	0.161
Cultural anxiety	0.187	0.195			0.408	0.218	0.402	0.235
Masculine role attitudes	0.228	0.168			0.131	0.175	0.066	0.185
Racial resentment	0.250*	0.124			0.247	0.128	0.218	0.133
Fears, Risks, and Safety Threats
Fear of crime			0.187	0.127	−0.048	0.140	−0.003	0.146
Belief in a dangerous world			−0.198	0.121	−0.267*	0.130	−0.247	0.136
Fear of victimization			0.264	0.162	0.341*	0.170	0.279	0.178
Perceived risk of victimization			0.033	0.175	0.027	0.183	0.084	0.190
Personal victimization			0.357	0.196	0.414*	0.206	0.409	0.210
Vicarious victimization			−0.053	0.203	−0.128	0.214	−0.128	0.219
Police distrust			0.022	0.099	0.177	0.117	0.127	0.123
Control Variables
Protective ownership							0.554*	0.242
Men							0.319	0.206
Age							−0.007	0.008
White							0.252	0.308
Republican							0.114	0.227
Education							0.083	0.082
Income							−0.017	0.054
South							0.239	0.185
Rural area							0.122	0.231
−2 Log Likelihood	725.598		748.343		708.492		696.651

^*p < 0.05,

^**p < 0.01,

^***p < 0.001 (two-tailed tests).

Table 7.

Loathing, fear, and prospective protective gun ownership, binary logistic regression (n = 1,380 non-owners).

	Model 1		Model 2		Model 3		Model 4
	b	SE	b	SE	b	SE	b	SE
Intercept	−1.028***	0.219	−0.525**	0.168	−1.241***	0.244	−0.465	0.375
Loathing: Status Anxieties
Economic precarity	0.091	0.060			0.051	0.062	0.093	0.064
Economic alienation	−0.149	0.099			−0.175	0.107	−0.118	0.111
Cultural anxiety	0.642***	0.139			0.669***	0.149	0.608***	0.158
Masculine role attitudes	0.305*	0.119			0.302*	0.121	0.280*	0.128
Racial resentment	0.155	0.084			0.135	0.084	0.075	0.088
Fears, Risks, and Safety Threats
Fear of crime			0.221**	0.081	−0.078	0.090	−0.082	0.091
Belief in a dangerous world			−0.024	0.077	−0.093	0.084	−0.056	0.087
Fear of victimization			0.322**	0.105	0.391***	0.112	0.375**	0.116
Perceived risk of victimization			0.065	0.115	−0.036	0.122	−0.089	0.127
Personal victimization			0.228	0.136	0.204	0.144	0.179	0.146
Vicarious victimization			0.208	0.119	0.281*	0.126	0.266*	0.129
Police distrust			−0.251***	0.061	−0.037	0.071	−0.088	0.074
Control Variables
Men							0.031	0.130
Age							−0.017**	0.006
White							−0.129	0.147
Republican							0.561***	0.165
Education							−0.121*	0.052
Income							−0.020	0.032
South							0.285*	0.129
Rural area							−0.070	0.220
−2 Log Likelihood	1725.870		1841.676		1699.654		1663.175

^*p < 0.05,

^**p < 0.01,

^***p < 0.001 (two-tailed tests).

Fear, loathing, and non-owners’ openness to protective gun ownership

A key challenge in the current study (one facing much gun scholarship to date), is that given the cross-sectional nature of the data, it is not possible to discern the temporal ordering of these fear/loathing-protective ownership/carry associations; indeed, both directions are conceptually plausible (e.g., protective gun carry neutralizes perceived victimization risk, see Hauser and Kleck 2013). In an effort to isolate more clearly these possible associations, our final focal models examine loathing and fear on possible future protective gun ownership. Kelley and Ellison (2021) describe this group as ‘maybes,’ and their recent analysis found them to be similar to never-owners in some respects (more likely to be women, lower income, live outside rural areas) and similar to current owners in others (more likely to be Republican or conservative).

Among respondents who do not currently and have never owned a gun (n = 1,380), we tested the odds that non-owners would consider owning a gun in the future primarily for protection. Cultural anxiety and masculine role attitudes (Model 1) and crime and victimization fears (Model 2) are associated with an increased openness to future protective ownership. Consistent with the model of protective ownership among the full sample (Table 5) police distrust is (unexpectedly) negatively associated with prospective protective ownership. We speculate this may be due to differences by race/ethnicity – non-white respondents (who are generally less likely to own guns) report higher police distrust than whites. Alternatively, although the negative sign for police distrust is contrary to expectations from the collective security hypothesis (Kelsay et al. 2018), interpreting this coefficient in terms of police trust (such that more trust in police is associated with higher odds of ownership) is consistent with the pro-gun angle of the pro-police (e.g., Blue Lives Matter) movement, the pro-police angle of the NRA, and the spikes in gun sales amidst calls to ‘Defund the Police’ (J. Carlson 2020; Lang and Jamison 2021). A more thorough investigation of such patterns is beyond the scope of the current paper (we return to this in the Discussion).

In the final adjusted model (Model 4), both loathing and fear remain positively associated with prospective protective ownership. Culturally anxious non-owners, and those endorsing stereotypical masculine ideals (‘men should be tough’) and perceived threats to ‘traditional’ masculinity (‘society punishes men for acting like men’) are more open to future protective ownership. Those who are fearful of victimization or who have experienced it vicariously are more open to future ownership. It is important to note that cultural anxiety is a key correlate of crime and victimization fears (as shown in Tables 3 and 4), and it continues to exert a unique association with prospective protective gun ownership, independent of the effect of victimization fears. Cultural anxiety also explains the effect of fear of crime on prospective ownership. This again echoes arguments discussed above suggesting that fear of crime may really be capturing reactions to social, cultural, and moral change and upheaval (Hirtenlehner and Farrall 2013).

In supplemental analyses (Appendix Table C2) we explored if such patterns might differ by race/ethnicity, re-estimating the full models for protective and prospective ownership among white respondents, and respondents from communities of color (the latter group an admittedly rudimentary grouping of respondents identifying as Black, Hispanic, Asian, and other race/ethnicities). Among whites, racial resentment, fear of victimization, and vicarious victimization are key correlates of protective gun ownership; among respondents of color, it is only personal victimization that matters. Among non-owners, cultural anxiety is associated with increased odds of prospective protective ownership for both white respondents and respondents of color. We see this as another indicator of our original assertion that cultural anxiety – while likely conceptually connected to racial attitudes and threats – is capturing an additional (and unique) dimension of status anxiety, with important implications for protective actions and attitudes.

Conclusion and discussion

The characteristics of and the factors motivating gun owners are multifaceted and complex. Guns in the United States remain most often acquired for protection, a paradox in light of historically low crime rates that have generally been in decline for decades. Yet these protective motivations inherently implicate at least some dimension(s) of perceived threat and/or fear. Gun rights groups, gun advertisements, and gun advocates frame guns as useful tools facilitating the restoration of law and order. Firearms instructors encourage students to view the world as dangerous and unpredictable (Shapira and Simon 2018), and some gun owners wax nostalgic about a much safer, and more ordered, past (J. Carlson 2015b). Prior work on correlates of gun ownership often focuses on fear of crime and/or victimization. Indeed, the current study shows that fear of victimization is a significant correlate of both current and prospective protective gun ownership. Yet – as the current study also shows – what these fears actually encompass is far more complex than one’s response to immediate (perceived) threats of victimization or physical safety (Warner and Thrash 2020).

This study extends criminological and sociological scholarship on gun issues, returning to and expanding the ‘fear and loathing’ thesis (Wright, Rossi, and Daly 1983), incorporating group threat perspectives to first examine the sources of broad crime and proximate victimization fears that may fuel status anxieties among traditionally dominant groups. That is, we argue that ‘fear’ of crime may be a proxy for anxieties about the state of society and the perceived declining moral order, and ‘loathing’ is directed at those challenging not just the racial status quo, but the economic and gendered one as well. Understanding the individual-level associations between fear of crime and behaviors like gun ownership (Hauser and Kleck 2013) or gun carrying (J. Carlson 2015a) requires awareness of both the meaning of guns (Metzl 2019b) and the social contexts from which the various threats influencing gun behaviors may arise.

Continuing to evaluate how gun motivations arise and are shaped remains crucial to inform effective public health and safety policies. Using data from two relatively recent, nationwide crowdsourced surveys, we find first that crime and victimization fears are as much – and in some cases more – about diffuse status anxieties (loathing) rather than direct safety threats and experiences. Loathing (alongside fear) shapes gun behaviors. We summarize the findings below.

Status anxieties (economic precarity and alienation; cultural anxiety; masculine role attitudes; racial resentment) are significantly associated with broad crime fears: that crime is increasing and the world is more dangerous. Safety threats (victimization) are largely unrelated. Economic and cultural threats are associated with fear of victimization, with cultural anxiety having the strongest association with both fear of and perceived risk of victimization – and this is net of victimization experiences, police distrust, and all other controls. Indeed, as Chadee and Ying note (2013), for individuals with little actual experience with crime (residents of low crime areas, non-victims) rational calculations of crime/victimization risk are not tenable; as such, these persons may be influenced more by their global fear response and symbolic notions of crime – that is, their ‘threat-oriented ideologies’ (Blalock 1967) – that are shaped by diffuse anxieties about the moral order.

Such ‘threat-oriented ideologies’ matter for protective gun behaviors, both directly, and through their role in shaping fear. Fear of victimization and victimization experiences are associated with protective gun ownership – but so too are cultural anxiety and racial resentment. Personal victimization is associated with protective handgun carry; so too is economic precarity. Even among non-owners, victimization fears and experiences are associated with openness to future protective ownership – cultural anxiety and masculine role attitudes are as well, and the strength of the anxiety association is almost twice that of victimization fear (stb = 0.242, stb = 0.135, respectively).

While prior criminological work on gun ownership often focuses only on crime/victimization-related fears, threats, and experiences, we show that other economic, social, and cultural threats shape these fears specifically, and influence gun ownership independently. Thus, what may be motivating protective gun ownership and protective gun behaviors are both actual victimization fears as well as broader group status threats (both of which are connected). Although some scholars implicate a diffuse fear that the world is out-of-control, dangerous, and plagued by ‘evildoers’ intent on inflicting harm (Stroebe 2013), this study suggests that these threats may be quite specific. Owning a gun for protection is reactive – to tangible fears (for one’s safety), but also to social, cultural, and economic changes that present threats (real or perceived) to the status quo and that engender feelings of fear. That is, many of the most pressing threats are those perceived threats to the social hierarchy and dominant groups’ place there (Blalock 1967; Blumer 1958). Gun sales in early 2020 reached unprecedented rates as the coronavirus pandemic spurred a surge in gun sales (Lee and Chabria 2020). Our emphasis on group threat, broadly conceptualized, helps contextualize why individuals may turn to guns in unsettled times (Burnett 2020), why gun stores were deemed ‘essential businesses’ during COVID-19 pandemic shutdowns (Allyn 2020), and why gun sales continued to surge alongside the nationwide social unrest that erupted in response to systemic racism and police violence (Alcorn 2020).

Limitations

Our study’s findings should be situated within some key limitations. First, this study uses a convenience sample of U.S. adults, and as such, findings are not generalizable to the U.S. adult population. Although differences between MTurk workers and population-based survey respondents can be reduced when controlling for key demographic features (Levay, Freese, and Druckman 2016), as the current study has attempted to do (and our sample is comparable to nationally-representative data across several indicators [see Appendix B]), we do not intend to draw broader conclusions here, but rather suggest that our findings are a first step to guide future research with more representative samples. Second, the study cannot make causal conclusions. While findings show how fears and anxieties are linked with protective gun ownership, reverse-causality is possible (e.g., gun ownership makes status anxieties more acute). Unfortunately, studies addressing this temporal-ordering issue are essentially nonexistent (Hauser and Kleck 2013). That said, we believe it unlikely that protective gun ownership or carry would necessarily spur cultural anxiety; furthermore, we attempt to address this with our analysis of prospective protective ownership, which revealed very similar patterns with respect to the roles of fear and loathing. A third limitation concerns some of the unexpected associations observed (regarding economic alienation and police distrust). Although we provided some brief insights into these findings, we recognize them as speculative, and note this as an area in need of future research. Fourth is our limited ability to investigate thoroughly these patterns by race and ethnicity – an important avenue for future research, given the findings illustrated in the preliminary supplemental analyses we examined here. A deeper exploration of fears, anxieties, and protective gun ownership among persons from communities of color is indeed warranted. Fifth is that our measure of personal victimization do not capture incidents experienced more than five years prior (which may have influenced ownership among current owners), nor does it capture explicitly all types of victimization (e.g., intimate partner violence, sexual assault). A final limitation is the absence of respondents’ current geographic location (MTurk by design is anonymous). Though we control for self-reported rural context and region, gun behaviors and attitudes are highly place-specific (Gahman 2015), and we cannot address more specific geographic correlates.

Conclusion

Notwithstanding these limitations, the current study illustrates the effect of group status threats on both crime/victimization fears and protective gun ownership and use. Attention to the role of these threats alongside more general crime fears (Chadee and Ng Ying 2013) may better inform sociological and criminological work on such protective behaviors; indeed, continued investigation into the role of culture in understanding gun attitudes and behaviors is warranted (Celinska 2007; Kahan and Braman 2003). Cultural anxiety, racial resentment, economic precarity, and masculine ideologies are highly interrelated, but these measures are not necessarily redundant. Our findings suggest that the ‘festering reserve’ of racially charged resentment (Drakulich et al. 2020) may also be fueled by threats to the economic and masculine status quo, and all of these can inform firearm behaviors. For some Americans – men, in particular – gun ownership is a central part of their identity, morality, and patriotism (J. Carlson 2015a; Lacombe, Howat, and Rothschild 2019; Mencken and Froese 2019; Stroud 2016; Warner and Ratcliff 2021); it is an indicator of masculinity threats in action (Cassino and Besen-Cassino 2020; Warner et al. 2021). In our revisiting and extending of the ‘fear and loathing’ hypothesis we illustrate that a wide range of gun behaviors and attitudes cannot be solely attributed to criminal victimizations fears, nor devoid from cultural or economic anxieties (nor should the connection between fears and loathing be ignored). Moving forward, criminological research on gun ownership – and the motivations of owners – might benefit from recognizing that individuals seek protection from not only threats to their personal safety, but from threats to their social standing. Awareness of the status hierarchies within which these anxieties manifest and gun behaviors are enacted is essential for crafting fully informed firearm safety efforts. Thus, while fear certainly matters for firearm behaviors, the role of loathing is also clear.

Biographies

Notes on contributors

Tara D. Warner is an associate professor of Criminal Justice at the University of Alabama at Birmingham, and a Research Affiliate at the Center for Family & Demographic Research, Bowling Green State University. As a sociological criminologist, she has been exploring how victimization fears and group status threats shape gun ownership, attitudes, and behaviors. Her second line of research examines social, emotional, and developmental consequences of violent victimization.

Trent Steidley was an assistant professor in the Department of Sociology and Criminology at the University of Denver when this article was submitted. He currently is employed by the Colorado Department of Human Services; all of his contributions to this article were completed before he transitioned from the University of Denver.