1 Introduction
Lethal firearms have for decades been a hot-button political issue in the United States. There are passionate gun rights activists, but also passionate gun-control advocates who, for example, highlight that they are a leading cause of death among children in the U.S. Despite such polarization, there is limited understanding of why people hold such different views about personal lethal firearm ownership, and even less so about what might influence their views.
In this article, we design and run original large-scale surveys of both current lethal firearm owners (LFAO) and non-owners (NO) to provide an in-depth account of the motivations and beliefs behind LFA ownership or the absence of it and to experimentally study whether providing information can shift people’s views. We provide two types of information treatments, one focusing on the private costs of lethal firearms – via increased risk of legal liability and harm to others – and the other centering around an option for personal protection previously unknown or neglected by most survey respondents: a non-lethal firearm (NLFA).
Beginning with the landmark report entitled Guns in America, by Cook and Ludwig (1996), several papers have sought to describe America’s gun owners and their motivations. Using General Social Survey (GSS) data, for instance, Glaeser and Glendon (1998) note: “Gun ownership appears to become a social norm when there is mistrust of public justice or where there is a tradition of private retribution.” One limitation of surveys such as GSS and Gallup is that they typically only include a handful of questions about firearm ownership. There is also a small set of dedicated surveys on the subject that include policy views (Siegel and Boine, 2020; English, 2022). We build on these studies and the literature in five important ways.
First, we contribute through the scale of our study. We include nearly 6,000 LFAO and non-owners together in a large-scale survey. We determined lethal firearm ownership using obfuscated recruitment and verified ownership status with a validation survey that included a timed gun part naming game. After large purchasing upticks with COVID-19, in particular, it is important to update our understanding of who currently owns firearms (Lopez, 2016; Helmore, 2021; De Vis´e, 2023; Hicks et al., 2023). Second, we expand the scope of questions covered in the “whys” behind gun ownership: we include not only questions on sociodemographics, family traditions, sports, and crime, but add new modules on social networks, externalities, internalities, and – importantly – emotional responses to lethal firearm ownership and the use of deadly force. Third, by including to the fullest extent possible similar modules for non-owners, we can identify those who are on the cusp of purchasing (non-owners interested, NO-I) and those who are uninterested (NO-U). Fourth, we include real-stakes and incentive-compatible questions that provide assurance on the quality of our responses. Fifth, we identify and test the importance of differences in knowledge about the private costs of LFAO and a fairly new non-lethal alternative (NFLA) on gun-related beliefs, behaviors, and policy views. How people view and think about NLFAs has not been explored so far but offers a potentially important avenue for policy.
Our findings can be organized in two main groups: First, a documentation of what drives demand for LFAs and, second, the impacts of information on gun-related behaviors and attitudes.
Starting with the first set of findings, there is a clear asymmetry in demand for LFAs. Many non-owners (30%) are potentially interested in acquiring a firearm1 but very few LFAOs are willing to reduce the number of guns they own or give them up altogether. In fact, 60% are interested in acquiring more. When it comes to the reason for owning guns, we find that demand for safety and protection is key. Both LFAO and non-owners (NO) listed protection of family or self as the top rationale for owning (in the case of LFAO) or why they would acquire (in the case of NO) a lethal firearm. Demand for safety does correlate with objective measures of risk such as local crime rates or gun-related violence at either the county or zipcode levels. LFAO and NO-U report feeling similarly safe in their daily lives, despite the former being personally, lethally armed. In contrast, NO-I report feeling significantly less safe in their daily lives and trust the police less to keep them safe than either group. These differences persist even if we control for local risk.
If everyone demands safety, why do some people choose to own LFAs while others do not? We find large heterogeneity in the perceived costs of guns. While LFAO, NO-I, and NO-U would feel similarly upset about accidentally harming or killing another person in self defense or otherwise, there is a large gap across groups in how worried they are that a child could use their gun to hurt themselves or others, and in being arrested or sued if someone took their gun and caused harm: NO-U are the most worried and LFAO the least, with NO-I falling in the middle. There are similarly large differences between groups in their perceptions of externalities associated with lethal firearm ownership, with NO-U being most likely, NO-I the second most likely, and LFAO the least likely to believe that firearm ownership increases crime, murders, suicides, school shootings, and serious accidents. As a result, people feel very differently about owning guns: many LFAO feel safe, confident and valuable by owning a lethal firearm. They are much more likely to report that carrying a gun reduces their likelihood of being victimized or hurt. Similarly, the top emotional responses that LFAO report to imagining their guns being taken away include feeling insecure and vulnerable, alongside feeling angry and frustrated. Many non-owners, on the other hand, feel the opposite: the top emotions related to firearms for non-owners who are uninterested in acquiring a firearm are unsafe, nervous, scared, and irresponsible. We then document a lack of awareness of non-lethal alternatives, even though many LFAOs say that they would prefer to own a firearm that could incapacitate rather than kill, suggesting an unmet demand for such alternatives.
Thus, lethal firearm owners and non-owners appear to be driven by a common objective: to feel safe. However, they have different views about whether guns make them safer. Owners feel safe with lethal guns and vulnerable without them. Uninterested non-owners feel unsafe and irresponsible with LFA and are worried about private and social harms. Interested non-owners appear to fall in-between and feel especially unsafe in daily lives. These differences do not seem readily explained by other sociodemographics or the actual crime threat people face. Most gun owners have not considered purchasing non-lethal firearm alternatives and are not aware of their availability.
Given that personal safety is the most prominent rationale provided for lethal firearm ownership, yet gun owners and interested non-owners seemed to neglect certain firearm-related risks and non-lethal alternatives to lethal force, we devised treatments that targeted these information gaps. For all respondents, we randomly provided information on the private costs of ownership, including the heightened risk of suicides and accidents among households that possess a lethal firearm, as well as the legal liability that families might face if their children use a lethal firearm to commit a crime (as had been recently decided in a case in Michigan). We call this the Private Costs of Lethal Firearms treatment. For current owners, we also randomized information about a fairly new non-lethal firearm alternative commercially known as the Byrna Gun. The product has been on the market since 2019 and is the first firearm to be specifically marketed as “an alternative to lethal force” among the U.S. civilian population. All treatments included a narrated video, and the Byrna treatment was further divided into a pure information treatment (the Non-Lethal Firearm Information treatment) vs. a treatment that combined information with an endorsement from the conservative TV personality, Sean Hannity (the Non-Lethal Firearm Information and Endorsement treatment).
We find that the Private Costs of Lethal Firearms treatment increases concerns about lethal firearm usage among LFAO and Non-Owners and affects policy views on safe storage. Information about NLFA affects both the willingness to view Byrna as an alternative to a lethal firearm, and agreement with the sentiment that they prefer a firearm to incapacitate but not kill a person. We find increases in willingness to pay for a NLFA, to consider a NLFA over a lethal firearm for their next firearm acquisition, and willingness to support policies that promote non-lethal guns. In general, the effects of the information vs. information and endorsement are indistinguishable in the short-term, but the endorsement leads to greater persistence in the follow-up. Despite this, LFAO were not willing to “give up their (lethal) guns” in response to information on the NLFA. Qualitative responses probing this reluctance revealed that the treatments favorably shifted people’s assessments about the suitability of NLFA for self-defense but some continued to imagine circumstances where only lethal force could be appropriate.
These descriptive and empirical findings can be interpreted through our organizing framework in which every household has a demand for safety but differs in how they think about producing it. Specifically, households produce safety by purchasing various tools in the market – tools that vary in their net safety benefits (protective benefits minus harms). People differ in their choices about which tools to purchase for several reasons. For example, because of heterogeneity in underlying needs or preferences, they might differentially trade off protective benefits against harms. More unique to our approach, and consistent with the empirical analysis, they might have different beliefs, perceptions, and visions for which tools deliver the most net safety: they could imagine more frequent or intense encounters in which lethal force would be required, neglect harms and focus on protective benefits, and/or differ in knowledge or beliefs of readily available options such as NLFA. Such heterogeneity leads them to hold different beliefs about the safetypossibilities frontier (the highest expected protective benefit for any given amount of expected harm) and stands in contrast to fundamentally disagreeing about the correct point to be on the true frontier. Fundamental preference disagreement suggests few levers, if any, could resolve the disagreement. Heterogeneity in beliefs about the safety-possibilities frontier, on the other hand, suggests that levers that influence those beliefs by providing or drawing attention to information relevant to the frontier could influence beliefs and behavior. Our framework then provides a lens for understanding why our experimental treatments could have a non-trivial impact on firearm-related beliefs and behavior. In the penultimate section of the paper, we discuss potential welfare implications of encouraging NLFA demand through treatments like ours, which could be viewed as “harm reduction” interventions in the language of public health.
Our focus on safety contributes to a broad literature in economics. Viscusi (1983) pioneered methods to reveal people’s willingness to pay (WTP) to reduce risks to life and health. Work by Spence (1977), Shavell (1984), and others illustrate reasons (including consumer misperceptions) why unregulated markets may not produce efficient levels of product safety. But these exercises assume individuals broadly agree on what factors reduce or increase personal safety. We innovate on this literature by empirically unpacking reasons people could agree on a goal but starkly disagree on how to achieve it.
Social scientists within and outside of economics have examined America’s gun owners. Cook and Ludwig (1996) detailed the landscape of ownership and use of lethal firearms in the mid-nineties.2 Many key features of the American lethal firearm landscape were established therein, including that lethal firearms were acquired primarily for safety. The authors also noted an important tension: “Americans are ambivalent about guns: they fear them and at the same time they feel safer possessing them.” As we show, this ambivalence largelystems from different views across individuals. With increasing sophistication of text analysis, many studies confirmed Cook and Ludwig (1996) initial findings. Boine et al. (2022), using latent class analysis, described six different types of owners, with a large share concerned for safety (further subset into concerns about family, self, or general protection) and residual groups prioritizing the Second Amendment or sport. Yamane (2022) documented a gradual increase in the diversity of lethal firearm owners over the last four decades –becoming less rural, less white, and less male. Past victimization and fear of violent crime have consistently been found to be important predictors of gun ownership (Kelley and Ellison, 2021; Kleck and Kovandzic, 2009; Warner and Steidley, 2022; Kleck et al., 2011; Warner and Ratcliff, 2021), especially for women (Warner, 2020). Yamane et al. (2018) noted a rise in self-protection topics in NRA-affiliated magazines over the last century (1917-2016); interpreting this and other findings as evidence gun culture has shifted from so-called 1.0 (recreational hunters) to 2.0 (armed citizens), with an increasing emphasis on the Constitution (Burton et al., 2021). Shapira et al. (2022) discuss how experience in childhood and young adulthood influences views on lethal firearms, and Warner and Ratcliff (2021) delve into feelings associated with ownership.
Several studies assess citizen views on lethal firearm regulation, including survey experiments that randomize respondents to clips of school shooting coverage or documentaries on the subject (Dixon et al., 2020; Robbers, 2005; Parker et al., 2017a,b). Treatments targeting misconceptions regarding other gun-owners’ support for regulation has been shown to be mildly effective at increasing support for stricter policies among those harboring misperceptions (Dixon et al., 2020; Susmann et al., 2022). Luca et al. (2020) find that mass shootings appear to lead to looser gun regulations in states with Republican-controlled legislatures. Policies that lower the health costs of gun violence appear to be highly desirable. Using a contingent valuation approach, Cook et al. (2025) find U.S. households would pay on average $744 annually for a 20% reduction in gun violence (or nearly $100 billion per year) and Rosenberg (2024) provides causal estimates of the health externalities associated with lethal firearm ownership.
A smaller literature in economics has estimated the demand for lethal firearms. The experimental exercises are limited in realism since researchers are ethically unable to provide respondents with a real firearm. Despite these limitations, Moshary et al. (2025) provide a rigorous estimate of the price elasticity for lethal firearms using a stated conjoint analysis. Respondents who expressed interest in purchasing a firearm were screened in, and attributes of lethal firearms and prices were cross-randomized. The reduced-form results found that would-be new owners were more price sensitive than those in the market for another lethal firearm. An accompanying structural model that allowed for substitution across handgun and long-gun types, with a no-gun outside option, demonstrated more substitutability from the latter to the former – with implications for long gun bans. Armona and Rosenberg (2024) build a structural model of the firearms market, finding substantial heterogeneity in preferences for lethal firearm characteristics across consumers, and derive the optimal (tax) policy under political constraints. Rosenberg (2025) uses gun store entries to identify the effects of increased LFA purchases in a given area, finding increases in associated fatalities that are driven by adversely selected marginal consumers. Our approach builds on this important reduced form and structural research by making salient legal liability associated with ownership, introducing a relatively novel less-lethal firearm technology, and more broadly uncovering the “whys” behind firearm demand.
The remainder of the article proceeds as follows. Section 2 describes the survey sample including the experimental components. Section 3 provides descriptive results as to why people do and do not own lethal firearms. Section 4 presents our organizing framework for interpreting the aforementioned patterns and formalizes the potential role of drawing attention to non-lethal firearm alternatives. Section 5 describes the experimental findings, Section 6 offers a discussion of the policy implications, and Section 7 concludes.
Click to read more