In American politics, myriad issues compete for the public’s attention. Some succeed in limited ways, gaining salience in communities that have particular reasons to care about them; insulin pricing, for example, is likely to be a top-tier issue for diabetics, and federal flood insurance programs are important to people in areas especially vulnerable to climate change. However, issues perceived as mostly affecting small, narrowly defined communities are rarely the subject of broad mobilization and play little role in structuring political conflict. Other issues become broadly salient following catastrophes, alarming shifts in troubling indicators (e.g., inflation), or other news-generating events (Kingdon Reference Kingdon1984). But often these issues are subject to an “issue-attention cycle”—rising quickly as matters of public attention but cratering once the problem subsides, fades from memory, or gains a superficial response from policymakers (Downs Reference Downs1972; Edelman Reference Edelman1964). These issues are sometimes consequential, but attention to them is fleeting in ways that reduce their long-term political importance and the likelihood that policymakers will meaningfully address the problems associated with them. Finally, some issues constitute what we term “durable attention items.” They receive sustained public engagement and may play a central role in organizing partisan contestation over many years. Surprisingly, we know less about the dynamics surrounding the emergence of such issues—those that come to the public’s attention and continue to capture it over long periods of time. Focusing on gun control, we explore how previously ephemeral issues might become hearty perennials of American politics.
Gun regulation is a useful case for exploring this question. Although salient to many gun owners—for whom the perceived threat of new laws constantly looms—gun regulation has been subject to an issue-attention cycle for the majority of Americans who support it. In Spitzer’s now-classic formulation, gun issues have followed a cycle of “outrage, action, and reaction,” in which especially salient events draw an emotional public response and demands for stricter gun laws, arousing opposition from the well-organized gun lobby and an accompanying retreat by pro-regulation lawmakers (Spitzer Reference Spitzer2018, 24-26). The ability of pro-gun groups to sustain activism as energy fades among their opponents helps explain why popular gun regulations fail to be enacted into law.
As we show, however, issue-attention cycles may no longer define the gun debate to the extent they once did. Instead, supporters of gun regulation have recently demonstrated increased and sustained political participation. Durable engagement in favor of gun control appears to have been pivotal to enactment of the Bipartisan Safer Communities Act of 2022, the first significant new federal gun law in nearly thirty years (Murphy Reference Murphy2022). The pro-regulation shift in political engagement constitutes both an empirical puzzle and an opportunity to further theorize about how narrowly concentrated or fleeting public problems become broad, durable issues. Many issues are perennially concerning to the public; we are interested in how they become so.
We situate our question at the intersection of three distinct literatures: on mass participation in politics, agenda setting, and issue evolution. We suggest that several contextual factors touched on in these literatures—namely policy- and event-motivated threats and partisan polarization—shape which issues matter to the mass public in a sustained way. We then introduce our primary dataset: a multi-decade corpus of thousands of letters to newspaper editors that we use as a novel measure of the public agenda—“the set of policy issues to which the public attends,” which we typically refer to as public attention (Jones and Baumgartner Reference Jones and Baumgartner2004, 3). Using these data, we demonstrate that gun policy has become a durable issue of public concern (a durable public issue, for short) and that its status as such is due to increased mobilization among pro-regulation Americans—a group whose participation was historically episodic and tepid relative to that of pro-gun people (Goss Reference Goss2006). By “durable,” we mean an issue that reliably captures public attention over a period of years, not merely news cycles. By “of concern to the mass public,” we mean that attentive citizens, particularly those predisposed to care about public affairs, are engaged with the issue not only psychologically, but also as a matter of political behavior. Having established the empirical puzzle that motivates the theoretical question, we demonstrate how contextual factors (more than individual and organizational factors) not only can bring an issue to public attention, but also hold citizens’ attention over time.
Setting the Public Agenda: When Do Issues Stick Around?
Most agenda-setting literature focuses on the governmental agenda (sometimes called the formal, or elite, agenda), meaning the set of issues to which public officials are attending. From this literature we know that public problems can quickly grab policymakers’ attention—that is, sometimes an issue’s “time has come,” per Kingdon (Reference Kingdon1984)—but then fall off their agenda (Kingdon Reference Kingdon1984; Baumgartner and Jones Reference Baumgartner and Jones1993; Downs Reference Downs1972). We don’t take issue with this finding but note that less scholarly attention has been paid to why and how problems might rise to prominence, but then, rather than receding, continue to demand attention. When does a political issue “whose time has come” (Kingdon Reference Kingdon1984) become an issue “that sticks around”? We examine this question with a focus not on elite policymakers, but rather on everyday people.
A massive body of scholarship has identified the factors that drive mass-level political participation, a particularly important manifestation of issue attention. We can place these factors into three overlapping categories: individual, organizational, and contextual. Scholars have paid far more attention to individual and organizational factors than to contextual ones. For example, the influential “civic resources” model developed by Verba and colleagues focuses on how individual-level endowments—e.g., time, money, civic skills—facilitate people’s engagement (e.g., Verba, Schlozman, and Brady Reference Verba, Schlozman and Brady1995). Meanwhile, another body of work emphasizes how organizations train (Verba et al. Reference Verba, Schlozman, Brady and Nie1993), mobilize (Skocpol Reference Skocpol2003), recruit (Rosenstone and Hansen Reference Rosenstone and Hansen1993), and empower (Barnes Reference Barnes2020) people to participate in politics.
While individual and organizational perspectives have birthed powerful explanatory theories relevant to the case at hand, we focus on a third perspective—contextual factors—to understand shifts in public engagement on the gun issue. Contextual factors pop up in the agenda-setting literature under the guise of focusing events and critical indicators, but work on this topic focuses less on mass behavior than on the behavior of political elites (Kingdon Reference Kingdon1984). To the extent that the agenda-setting literature does consider mass publics, it focuses on what grabs citizens’ attention more than what sustains their engagement on any given issue. We look at events and indicators—i.e., contextual factors—as spurs for durable participation.
Contextual factors also crop up in studies of “policy feedback,” the proposition that “new policies create new politics” (Schattschneider Reference Schattschneider1935, 288). We conceive of policies (and proposed policies) as a key type of contextual factor that drives durable public engagement. While policy is ultimately a contextual factor, most scholarship on policy feedback focuses on how laws and regulations spur participation via individual and organizational factors. Policy shifts, for example, can influence individuals’ capacity, resources, and psychological dispositions for participation (e.g., Pierson Reference Pierson1993; Mettler and Soss Reference Mettler and Soss2004; Mettler Reference Mettler2005; SoRelle Reference SoRelle2020), as well as the ability of organizations to mobilize them into action (e.g., Skocpol Reference Skocpol1992; Campbell Reference Campbell2003; Goss Reference Goss2010; Rose Reference Rose2018; Goss, Barnes, and Rose Reference Goss, Barnes and Rose2019; Barnes Reference Barnes2020). However, with important exceptions (e.g., Patashnik Reference Patashnik2023), policy feedback scholarship has paid less attention to how laws, rules, and guidelines can generate public engagement by provoking backlash by the policy’s opponents. We pick up from those who have pointed out that backlash—and the policy threat inspiring it—is an important dynamic (Arnold 1992; Campbell Reference Campbell2003; Patashnik Reference Patashnik2019; Patashnik Reference Patashnik2023).
In addition to playing a role in agenda-setting and policy feedback processes, contextual factors are important to the “issue evolution” process, which refers to how issues reconfigure party cleavages in ways that then impact the public and elite agendas moving forward (Carmines and Stimson Reference Carmines and Stimson1989). This perspective holds that issues remain salient because they are baked into the party system and thus constitute important lines of political conflict. Some scholars have argued that gun regulation has become such an issue (Lindaman and Haider-Markel Reference Kara and Haider-Markel2002; Joslyn Reference Joslyn2020), at least for gun-rights supporters (Conley Reference Conley2019). Indeed, Conley (Reference Conley2019) argues that gun regulation meets the criteria that Adams (Reference Adams1997) sees as critical to whether an issue reconfigures partisan conflict, including that it remains a matter of public attention year after year. However, the issue-evolution framework is more interested in voting cleavages than in participation broadly understood. Additionally, while valuable in recognizing how focusing events “shake up” politics, the framework centers the role of elites and parties in ways that may hold more for some issues than others.
In sum, the participation literature explains the individual and organizational mechanisms that help determine the likelihood of citizens’ political engagement and observed variation in participation across differently situated people. But the participation literature provides few well-developed insights into the role of contextual factors. Other work helps fill this gap: the agenda-setting literature documents the role of events and indicators in galvanizing political leaders’ attention; the policy feedback literature shows how policy threat can galvanize public backlash; and the issue-evolution literature traces long-term changes in party structures that may result from mass attention to given issues. These theories provide important insights, from which we draw. However, none directly addresses the question before us: How previously fleeting problems become durable matters of public engagement.
To explore this dynamic, we first lay out the core argument that contextual factors are primarily responsible for initiating the reconfiguration of gun politics such that firearm policy has become a persistent object of public attention. We posit that two contextual factors, working in conjunction, are crucial: recurring threats (caused by focusing events and actual or proposed policy change) and partisan polarization. Our data suggest that threats light the match, while polarization provides the tinder that allows the fire to spread. We then revisit two plausible counterarguments—one centered on individual-level explanations (that newly energized, pro-regulation mothers have driven the shifts) and the other on organizational explanations (that pro-regulation forces have built a massively resourced social movement). We show that while both explanations are relevant, the developments underlying them were—at least initially—more the result of contextual changes than drivers of mass participation. In other words, while the efforts of mothers and the rise of gun-safety organizations have surely come to play a key role in sustaining mobilization, they responded to trends that had already been catalyzed by contextual factors.
Threat and Agenda Durability
Focusing events—e.g., crises or disasters (Kingdon Reference Kingdon1984)—dramatize losses in ways that mobilize participation. These events are powerful because, among other things, they tend to be sudden and create visible harms while also illustrating the potential for additional, future harms (Birkland Reference Birkland1997). As Birkland (Reference Birkland1998, 55) notes, such events cause “members of the public to identify new problems, or to pay greater attention to existing but dormant problems, potentially leading to a search for solutions in the wake of apparent policy failure” and a desire to act on behalf of those solutions. Mass shootings, for instance, can be followed by mobilization on behalf of gun control (Goss Reference Goss2006; Dees-Thomases Reference Dees-Thomases and Hendrie2004). Importantly, despite their well-documented impact on the public agenda and issue-specific political participation, these events are commonly characterized by a flurry of activity that proves to be relatively brief (Spitzer Reference Spitzer2018). As a result, they are important for understanding shifts in an issue’s place on the agenda but are not by themselves sufficient for explaining why an issue becomes a durable agenda item.
A second form of threat—rooted in proposed or enacted public policy—also can spur mobilization. In her study of how Social Security has fostered high levels of participation among older Americans, for example, Campbell (Reference Campbell2003, 100) argues that policy changes can influence an “individual’s tangible well-being or her values or principles” and thereby increase “the benefits of collective action by highlighting the potential for loss.” If policy-based threats are persistent, we argue, they can contribute to an issue’s status as a hardy perennial on the public agenda. Policies may spur participation when they are proposed, by creating a threat (Campbell Reference Campbell2003; Cho, Gimpel, and Wu Reference Cho, Gimpel and Wu2006), or once they are enacted, by imposing tangible or symbolic costs on diffuse or target groups (Pierson Reference Pierson1993; Arnold Reference Arnold1990). When policies shift in ways that would-be activists oppose and cause them to feel threatened, they can create a sustained backlash—what Patashnik (Reference Patashnik2019, 48) describes as “a strong adverse reaction against a line of policy development.”Footnote 1
To summarize, event- and policy-based threats can spur reactive political pressure in the form of increased participation, thus bringing issues onto the public agenda and, as long as the threat looms, keeping them there. One way that either type of threat can spur countermobilization and thus influence the agenda is by personalizing an issue; when events or policies create a situation in which the stakes of a given issue are believed to be more directly relevant or threatening to individuals, they may then be more motivated to take action. Prior work shows, for example, how gun control policies were perceived as targeted threats to gun owners, thereby spurring their participation on behalf of gun rights (Lacombe Reference Lacombe2021, Reference Lacombe2022). Meanwhile, gun regulation advocates historically struggled to frame their issue in terms that made the threat immediate to large swaths of Americans—or, in the words of Goss (Reference Goss2006), these groups struggled to “personalize the costs” of gun violence. Because feelings of threat motivate action, events or policies that personalize an issue in threatening ways might increase that issue’s prominence on the public agenda (e.g., Klar Reference Klar2013; Huddy, Mason, and Aarøe Reference Huddy, Mason and Aarøe2015; Nauroth et al. Reference Nauroth, Gollwitzer, Bender and Rothmund2015).
Polarization and Negative Partisanship
Another relevant factor is polarization and the negative partisanship associated with it. Polarization has not often been connected to agenda setting, but it has been shown to increase participation in ways that are relevant to our focus. Indeed, prior work demonstrates that activism, on the macro-level, has increased as polarization and identity-based partisan sorting also have increased. This is a product of polarization in both affective and issue-based forms; the former generates emotions that drive participation, while the latter highlights differences between parties, thereby making the stakes of politics seem higher and increasing negative partisanship (Mason Reference Mason2018; Hetherington Reference Hetherington, Brady and Nivola2008; Abramowitz and Saunders Reference Abramowitz and Saunders2008). These findings align with other research showing that polarization and identity-based partisan sorting are predictors of activism on the individual level, as well (Simas and Ozer Reference Simas and Ozer2021; Wagner Reference Wagner2021; Mason Reference Mason2018). Typically, this work has examined polarization and activism in general, as opposed to issue-specific, terms. Here, we extend these insights and connect them to the agenda-setting literature by examining what happens when particular issues become more polarized along partisan lines. Based on the aforementioned work on polarization and activism, we contend that issue-based polarization is likely to spur issue-based participation, and that—when other conditions are present—can contribute to issues becoming durable agenda items.
When an issue is highly polarized, new threats are more likely to be viewed through a partisan lens and thus connected to individuals’ partisan identities. As a result, rather than mostly spurring participation among individuals who have particular reason to care about an issue, such threats encourage broader participation among partisans who attribute harms associated with them to the opposing party and are motivated by negative partisanship to act. These polarized reactions to new threats can be the product of two categories of contextual factors: 1) focusing events (especially when they occur in rapid succession) and critical indicators, and 2) dissonant policy changes. Both kinds of factors can personalize an issue, leading to sustained activism and causing the issue to become a durable agenda item.
We illustrate this process through the gun debate. As we show later, the salience of gun regulation has waxed and waned in response to focusing events (e.g., the assassinations of Martin Luther King, Jr., and Robert Kennedy in 1968 and the Columbine shooting in 1999) and relevant indicators (e.g., the gun violence epidemic of the late 1980s and early 1990s). From the agenda-setting literature, this pattern is predictable and, from the perspective of advancing social science theory, not of particular interest. But as we show, the predictable spikes in attention (and participation) that surrounded particular developments masked an underlying trend of growing and durable agenda status for gun policy. It is this underlying trend, not the predictable spikes, that is puzzling, theoretically intriguing, and in search of an explanation.
Data and Methods
To examine trends in the prominence of gun control on the public agenda, we analyze letters to the editors of four geographically diverse newspapers: The Arizona Republic, The Atlanta Journal-Constitution, the Chicago Tribune, and The New York Times. Footnote 2 We collected all letters to the editor published in these newspapers between 1930 and 2019 (n=5,179) that contained the word “gun” or “firearm” and the word “law” or “legislation,” a process that returned both individual and batches of letters meeting these criteria (refer to n. 3). Most of our analysis centers on the 1980 to 2019 period (n=4,441). The volume of letters during this latter period is sufficient to enable analyses of various subsamples of letter writers—pro-gun versus pro-regulation people, males versus females, etc.—and, moreover, the shifts that interest us took place between these years (as we document later).
We use letters to the editor to inform a mixed-method analysis featuring quantitative and qualitative data. Following seminal studies of agenda-setting (see, e.g., Kingdon Reference Kingdon1984; Baumgartner and Jones Reference Baumgartner and Jones1993, Reference Baumgartner and Jones2015; and Jones and Baumgarner 2005), we produce descriptive statistics, including trend lines, drawn from “large n” longitudinal data. These data pose both the empirical and theoretical puzzles motivating the study and provide hints about how to unravel them. The qualitative data—what the letters were substantively “about”—allow us to evaluate mechanisms that may be inferred from the quantitative data. As other scholars have noted (e.g., Goss Reference Goss2020), it is typically not feasible to develop simple causal accounts anchored in elegant statistical models to explain major shifts in political dynamics over long periods of time. However, highly plausible and persuasive explanations can be derived by considering quantitative and qualitative data in tandem and evaluating one against the other.
Letters to the editor provide insights into the issues that are both drawing citizens’ attention and spurring them to commit a public act of political participation (Cooper, Knotts, and Haspel Reference Cooper, Knotts and Haspel2009). From a methodological perspective, studying letters to the editor has several advantages. For one, such letters constitute a mode of public participation that has remained broadly accessible in the same form over time. Letter writing requires certain civic resources (e.g., political interest, communication skills) associated with educated publics (Verba, Schlozman, and Brady Reference Verba, Schlozman and Brady1995)—the group to whom policymakers pay disproportionate attention (Gilens and Page Reference Gilens and Page2014; Bartels Reference Bartels2016).
Moreover, as Lee (2002, 98-101) notes while discussing his use of constituent mail as a measure of public opinion, these sorts of data represent a “proactive form of political expression” that inform “elite actors about the preferences of an active and attentive public” and preview individuals’ “likely actions” in ways that surveys do not. Compared to surveys, letters are better equipped to capture which issues have, in Lee’s (2002, 102-103) words, “high political awareness, high intensity of preferences, and the impetus to link thoughts and action.” This difference is important, as the gun debate has historically been characterized by a pattern in which a majority of Americans express support for additional gun regulations in surveys, but are out-mobilized by an active, intense minority (Goss Reference Goss2006; Lacombe Reference Lacombe2021)—a pattern that aligns closely with Arnold’s (Reference Arnold1990) broader, foundational argument that members of Congress fashion legislation based on the reactions they anticipate from mobilizable voters.
Further, letters allow for both quantitative counts and more nuanced qualitative analysis of rhetorical themes; unlike close-ended survey responses, letters’ open-ended nature sheds light on the underlying motivations and beliefs of those who write them. In short, letters to the editor—while limited in some ways relative to surveys—also have notable advantages, which we exploit to learn more about changes in the gun debate.
To collect the full population of letters from these newspapers, we primarily relied on ProQuest’s databases, which offer digitized archives of each paper covering most of the years included in our study; to cover the missing years, we turned to the newspapers’ own online archives, as needed.Footnote 3 We then coded each letter along a number of dimensions, which we describe throughout the text and in the notes. Detailed coding rules can be found in the online appendix, which also notes coding checks we performed to ensure a high level of reliability. As described later, we also use automated topic modeling to analyze the letters.
Reliability and Validity
Letters to the editor are not commonly used to study political participation over time, so we address up front questions that might arise about their reliability and validity when deployed for such a purpose. First, one might worry that changes in the volume of letters on a given topic (here, guns) might simply reflect changes in the volume of letters overall. While we cannot readily compile a definitive, systematic measure of “gun letters” as a share of all letters, we doubt this alternative hypothesis. For one thing, a spot check of the four newspapers’ archives suggested no change in the number of letters published over time.Footnote 4 Likewise, scholars using other data sources have charted the persistence of gun regulation as an agenda item; indicators include consistent levels of state bill introductions (Luca, Malhotra, and Paloquin Reference Luca, Malhotra and Poliquin2020); increases in gun-related legislative enactments (Siegel et al. Reference Siegel, Pahn, Xuan, Ross, Galea, Kalesan, Fleegler and Goss2017); and pollsters’ frequent inclusion of gun-related questions on surveys (Conley Reference Conley2019).
Another possible objection to letters as an indicator of issue expansion centers on gatekeeper preferences. If gun-related letters are growing—and, as we also show later, increasingly advancing pro-regulation perspectives—perhaps these trends don’t represent underlying changes in gun politics but instead editors’ discretion to elevate the issue generally and certain voices specifically. We doubt this explanation both because other scholars have shown that editors seek to include a representative sample of submissions (Hill Reference Hill1981; Hynds Reference Hynds1991; da Silva Reference da Silva2012) and because our observed trends are robust across all four newspapers.Footnote 5 If editorial idiosyncrasies were at play, we would not expect such consistency. In addition, as we show later, trends in letters to the editor parallel another salient form of gun-related participation—contacting lawmakers—as well as two other measures of public interest (Google searches and public sentiment that guns are an exceptionally important problem).
Guns: Public Engagement and Democratic Voice
Political conflict over firearms regulation dates back at least to the mid-nineteenth century, when individuals legally challenged state laws limiting the carrying of concealed weapons in public places.Footnote 6 Broader, if sporadic, engagement on gun issues followed the establishment in the late nineteenth century of federated mass membership groups, such as the National Rifle Association (1871) and the General Federation of Women’s Clubs (1890), as well the emergence of urban social reform movements in the early twentieth century. Noteworthy spurts in public participation accompanied legislative consideration of gun control measures at the local level (e.g., passage of New York City’s handgun ban in 1911; see Kennett and Anderson Reference Kennett and Anderson1975, 173-75), the state level (e.g., around the Uniform Firearms Act in the 1920s; see Kennett and Anderson Reference Kennett and Anderson1975, 192-197), and the national level, particularly during the 1930s (e.g., around the National Firearms Act of 1934; see Goss Reference Goss2006, 109; Lacombe Reference Lacombe2021, 129-136); the 1960s (e.g., around the Gun Control Act of 1968; see Goss Reference Goss2006, 37-39; Lacombe Reference Lacombe2021, 136-148); and the 1990s (e.g., around the Brady Handgun Violence Prevention Act of 1993; see Goss Reference Goss2006, 46-47).
However, these bursts of activity were just that: bursts. They followed the issue-attention cycle observed by Downs (Reference Downs1972) and, more precisely, the cycle of “outrage, action, and reaction” that Spitzer (Reference Spitzer2018, 24-26) identified in his study of gun politics. These cycles have emerged at what Goss (Reference Goss2006) called “movement moments,” periods when a fundamental shift in gun politics looked likely, yet the policy ecosystem nevertheless returned to the status quo ante, wherein pro-gun groups dominated pro-regulation forces in resources, mobilizing capacity, and political clout.
Nearly two decades after Goss’s analysis, we suggest that the fundamental shift that seemed never to materialize in fact may now be underway. While attention to gun issues continues to flare up after particularly newsworthy shootings, we argue that the cumulative effect of these regularized focusing events, coupled with broader political developments, has been to shift gun policy from an episodically salient issue to a durably prominent feature of the political agenda. These patterns are of interest both to lay people concerned about gun violence and to social scientists interested in theories of agenda access and democratic voice.
Letters to the Editor, 1930–2019
Gun issues generated at least some attention in virtually all of the ninety years for which we have data, but attention didn’t begin to grow notably until the early 1960s, when John F. Kennedy’s assassination and a wave of juvenile violence put gun control under consideration in Congress. Thereafter, letters to the editor illustrate the familiar issue-attention cycle, with spikes accompanying particularly salient focusing events, such as mass shootings (conventionally defined as shootings in one place resulting in at least four deaths, excluding the assailant). Note that relatively few mass shootings generate public mobilization of the sort we observe in the letters. Those that do spur mobilization typically occur in publicly accessible spaces widely expected to be “safe,” such as schools and workplaces; involve victims who are socially constructed to be innocent (such as schoolchildren); or claim especially large numbers of victims. In short, they have characteristics that are likely to generate media attention. Figure 1 shows the pattern.
Figure 1 Number of gun-related letters to the editor in four major newspapers
Viewed in a time-bound context, the spikes appear to be followed by a return to the status quo ante. And indeed, this conclusion seems valid for the 1960s through at least the mid-1980s. However, starting as early as the late 1980s, we begin to see a slightly different pattern—one noticeable only when looking at the totality of the time series, as opposed to the limited “up-down” spike associated with a given event. As the graph shows, the spikes become more frequent and the baseline level of engagement from which they emerge becomes higher. Thus, while public engagement around gun policy continues to depend on newsworthy events, these event-specific spikes are accompanied by a less visible but pronounced underlying trend in broadened engagement. Consider the average number of letters during the three least active years of each decade: During the 1960s, those “doldrums” years averaged four letters per year; during the 1970s and 1980s, the average was nineteen letters annually; in the 1990s, the average was sixty letters; in the 2000s (when 9/11 and security fears put gun control off the agenda), the average was still 46. By the 2010s, the “doldrums” average was 104 letters per year. In other words, the least engaged years in the 2010s were about five times more active than the least engaged years in the 1970s and 1980s and about twice as active as the 1990s and 2000s. Put another way, the least active years of the 2010s produced more letters per year than the most active years of the 1980s.
In sum, on the surface, gun letters appear to demonstrate the familiar agenda-setting dynamic—one featuring “punctuated equilibria,” “issue-attention cycles,” and “movement moments” of “outrage, action, and reaction.” But viewed over the longer term, we see a second process at work, one masked by the attention-grabbing spikes: an increase in attention without a subsequent return to the status quo ante.
Letters to the Editor, 1980–2019: Shifts in Democratic Discourse
The steep rise in letters about the gun issue tells us that more citizens are engaging with it—that it has become a durable feature of the public agenda. These trends raise important questions that letters to the editor allow us to address: Which broad categories of citizens are paying attention? What are they saying? What is motivating them?
The gun debate has been marked by asymmetrical organization and participation favoring gun rights advocates over gun regulation supporters (Schuman and Presser Reference Schuman and Pressser1981; Cook and Goss Reference Cook and Goss2020; Goss Reference Goss2006, Reference Goss2017; Lacombe Reference Lacombe2021). Interestingly, however, the letters data suggest that the historic participation gap appears not only to have closed, but also to have reversed. Figure 2 shows that this reversal is present in the percentage of all letters taking a pro-regulation, as opposed to a pro-gun, position.Footnote 7 (The online appendix depicts the absolute number of letters on each side and tells a very similar story.)
Figure 2 Percentage of letters on each side of the gun debate
The key finding is a stark one and runs contrary to the conventional wisdom. By at least one robust measure of public discourse, pro-gun people have lost their dominance of the gun debate. Until the mid-2000s, pro-gun letters outnumbered pro-regulation letters by 2:1 or even 3:1. By 2010 or so, the ratios had generally reversed. As figure 2 shows, the change point occurred between 2009 and 2010. During this period, pro-gun forces went from representing the majority of letter-writers to a minority of letter-writers; by contrast, pro-regulation voices increased from about one-third to one-half of all letters on guns.
Robustness Checks
We are confident the trends shown in figures 1 and 2 are substantively meaningful and not the result of alternative explanations. For example, we considered whether pro-gun people might be turning away from newspapers as they become associated with liberalism and alternative forms of news become available. We doubt this explanation, as data suggest that newspaper readership remains concentrated among political conservatives and moderates.Footnote 8
We also investigated whether these trends might be an artifact of changing preferences among editors—for example, perhaps they have become increasingly inclined to publish gun-related letters or prioritize pro-regulation over pro-gun submissions. Four robustness checks lead us to doubt these hypotheses. As noted earlier, we probed our data for observable differences among newspapers that might reveal the independent, idiosyncratic role of editors, but found that the trends were broadly consistent across the four geographically diverse newspapers. Second, we searched for “gun” in Google Trends – which plots the relative frequency of search terms over time (going back to 2004)—and found an upward trend with localized spikes around mass shootings, similar to what is in our letters data. Third, we examined survey data from 1979–2022 on public perceptions of the “most important problem facing the country today” and found that guns became a regular public concern beginning around 2010.Footnote 9 Finally, we examined patterns in contacting lawmakers, another popular form of political expression. These survey data (although collected sporadically) show trends similar to those in our letters: Pro-regulation people have overtaken pro-gun people both in rates of contacting and in the composition of all “contactors.” (Refer to table 1.) This new development defies the iron law of gun politics that pro-gun people are more likely than anti-gun people to turn preferences into political action (Schuman and Presser Reference Schuman and Pressser1981; Goss Reference Goss2006).
Table 1 Elected official contact regarding gun policy
Note: Respondents were sorted into “pro-gun” and “pro-control” categories based on their reported views on concealed carry laws (which each survey asked about in similar forms). Each survey also asked similarly phrased questions about whether respondents hand contacted elected officials about gun policy; however, the 1996 survey asked whether they had done this in the past 5 years, whereas the 2017 and 2018 surveys asked about the prior 12 months. Survey data for 1996 is from the Johns Hopkins Center for Gun Policy Research/Joyce Foundation (1996) conducted by NORC; 2017 from the Pew Research Center for the People & the Press: American Trends Panel Waves 25 and 26 (2017); 2018 from Pew Research Center: American Trends Panel Wave 38 (2018).
From our primary analysis and robustness checks, we reach two conclusions. First, gun regulation has become a durable item on the public agenda. While waxing and waning as agenda-setting theories would predict, periods of growing engagement result not in a total retreat to the baseline, but rather a “step function” increase in baseline interest. Unlike in the pre-1970s period, public involvement on gun issues never returns to (functionally) zero. Second, all of the increase in public engagement on gun issues, at least in the last decade, comes from pro-regulation people. Gun rights supporters have lost their dominance of the gun debate, rather dramatically so in recent years. The next section explores what might be driving these shifts.
What Made Gun Regulation a Growing Feature of Mass Political Engagement?
Here we begin to untangle the empirical puzzle surrounding gun politics and, in the process, offer an account of how ephemeral issues come to occupy a permanent place on the public agenda. A caveat is in order: Any long-term trend, particularly in complex socio-political phenomena, will inevitably be driven (or at least facilitated) by multiple forces that may or may not be visible or measurable. The story we develop here, therefore, is based on the most plausible inferences emerging from multiple lines of evidence and theoretical stress-testing. We look at a forty-year pattern and seek to explain it by exploring theoretically informed hypotheses that train our eye on individual-, organizational-, and contextual-level explanations. We look for inflection points in the empirical trends and turn to the qualitative data to understand what might be causing these shifts.
Event-Based Threat: “You Can Keep Your Thoughts and Prayers” Footnote 10
In figure 1, we showed that public expressions of concern about guns from 1930–2019 spiked after highly salient shootings but, in recent years tended not to slide back to the status quo ante. We also showed that gun-regulation advocates, not the “gun lobby,” are responsible for the increase in gun-related public expression since the mid-2000s. The threat environment created by mass shootings, which have become deadlier and more frequent (Lankford and Silver Reference Lankford and Silver2020), appears to be driving the overall increase in gun-related participation. To probe this proposition, we seek clues in the content of letters to the editor during the 1980–2019 period using two methods: concept coding and automated topic modeling.
Are mass shootings driving the increase in public participation around guns? Figure 3 shows the percentage of letters—based on our coding analysis—that mention mass shootings explicitly or, read in context, implicitly allude to such events (with a similar figure in the online appendix that shows these letters in absolute numbers).Footnote 11 We show both the see-saw patterns—responding to events, as predicted—and the smoothed, five-year averages (dotted lines) to facilitate interpretation of the underlying trends. As the figure shows, the increase in mass shooting-related letters that begins in earnest in the mid-2000s coincides temporally with the increase in pro-regulation letters overall.
Figure 3 Percentage of “anti-gun” letters written in response to mass shootings
*Dotted line is five-year moving average.
The role of mass shootings in driving engagement also emerges in automated topic models of letter discourse. Topic modeling is a method for inferentially extracting core themes from qualitative data—using algorithms to identify underlying “topics” discussed in the letters based on patterns in the words they contain. Topic models allow ideas to emerge “automatically” based on word usage rather than from categories we impose on the data.Footnote 12 Figure 4 depicts the prominence of each topicFootnote 13 over time, showing a pronounced rise in letters centered on mass public shootings, including in schools, while demonstrating fairly steady (if “wavy”) trends in other themes. Notably, we find evidence that this trend has been driven by gun control supporters. In a separate model, we estimated which topics are more or less associated with letter writers on each side of the debate and find that the mass-shooting trend—the “big mover” in the data—is driven by pro-regulation writers, who are statistically significantly more likely to discuss it.Footnote 14 This finding provides evidence that the overall pro-regulation shift in letters to the editor has been driven by people concerned with mass shootings.
Figure 4 Relative prominence of topics in gun-related letters over time
Overall, this section presents evidence consistent with the notion that mass shootings are event-based threats that help explain the growing number of letters to the editor about guns in recent decades, as well as the new dominance of pro-regulation voices. The lack of a national policy response in the form of new gun regulation may serve to intensify these threats, increasing people’s willingness to use their voice to influence lawmakers. This hypothesis is consistent with Baumgartner and Jones’s argument (1993, 117-118) that the same issue (here, gun violence) can generate two kinds of mobilization: a “Downsian” mobilization consisting of a focusing event followed by a spurt of activity that fades once the futility of the problem’s solution become apparent (see Downs Reference Downs1972); and a Schattschneider mobilization, wherein opponents of the status quo seek to expand the scope of conflict (see Schattschneider Reference Schattschneider1960). In both cases, governmental action (or inaction) can shape the intensity and durability of public attention (for an account of how government institutions process information signals, including from the public, see Jones and Baumgartner Reference Jones and Baumgartner2005).
Policy Threat: “The Appalling Idea of Arming Teachers in the Classroom” Footnote 15
Beyond mass shootings, we investigate a second category of contextual drivers: policy-based threats. Have rightward shifts in proposed or enacted gun policy contributed to increasing activism on the issue among pro-regulation Americans? As noted earlier, legislation can spur activism by generating backlash among individuals who oppose and feel threatened by it. In this context, state-level gun policy changes (and proposed changes)—which have loosened regulations on guns around the country—may have a played a role in mobilizing greater activism on behalf of gun control.
Suggestive evidence of this hypothesis comes from the timing of the patterns in our data. Our period of study coincides with a widespread, relatively rapid pro-gun shift in state-level gun laws around the country. In the 1980s and 1990s, many states repealed bans on concealed carry or relaxed the requirements for obtaining a concealed-carry permit. Starting in the mid-to-late 2000s and continuing throughout the 2010s, twenty-three states have gone further by abandoning permit requirements altogether (Bellware Reference Bellware2022; Siegel et al. Reference Siegel, Pahn, Xuan, Ross, Galea, Kalesan, Fleegler and Goss2017). Additionally, thirty states have adopted stand-your-ground laws, allowing individuals to use deadly force to defend themselves outside the home if they feel threatened (Cook and Goss Reference Cook and Goss2020, 151; Siegel et al. Reference Siegel, Pahn, Xuan, Ross, Galea, Kalesan, Fleegler and Goss2017). The proliferation of “guns everywhere” policies coincides with the increased number of letters and increased share of pro-regulation letters we documented earlier, suggesting that some letter writers may have been motivated by threatening policy shifts.
To assess this hypothesis, we look at the content of the letters. Figure 5 depicts the percentage and number of letters written in support of gun control that both discuss pro-gun policies and attribute negative outcomes to them.Footnote 16 The number and percentage of such letters have risen dramatically. (Note that the high percentages early in our period of study—coinciding with debate over the Firearm Owners Protection Act of 1986—are mostly driven by a small number of total letters.) These patterns lend credence to the notion that a shift toward greater pro-regulation activism has been driven at least partially by backlash against enacted and proposed pro-gun policies. While the proportion of total pro-regulation letters that discuss “guns everywhere” policies noticeably slid toward the end of our period of study—at a point in time when the threat generated by mass shootings came to dominate the debate to a greater extent than in the past (refer to figure 3)—it seems clear that these policy changes played a role in catalyzing the durable mobilization of gun control supporters that we observe over the past two decades. Indeed, as figure 2 shows, it was a spike in these types of letters that temporally coincided with the number of pro-regulation letters finally coming to outnumber the number pro-gun letters, a trend that then continued through the end of our data series.Footnote 17
Figure 5 Percentage and number of “anti-gun” letters written in response to pro-gun policies
*Dotted lines are five-year moving averages.
To gain additional leverage on the impact of pro-gun policies, we look at the topic modeling findings. Figure 4 shows that the concealed carry topic bumps up a bit around the time we see an uptick in letters, and a separate model (refer to the online appendix) shows that this uptick is associated with anti-gun writers (albeit not quite at a statistically significant level). To validate that a pro-gun policy threat was associated with anti-gun participation, we also leverage variation across newspapers. We find that a disproportionate share of backlash letters in the 2004–2012 period—when such letters peaked in our data—come from the Arizona Republic and Atlanta Journal-Constitution. This finding is notable because Arizona and Georgia during this period both made substantial changes to their concealed carry laws and seriously considered additional, more controversial policies. Both states, for example, enacted laws allowing guns in bars (Gay Reference Gay2010), and Arizona’s legislature passed a bill—vetoed by the governor—that would have allowed gun carrying on college campuses (KNAU 2011). These patterns lend further credence to our argument about the impact of pro-gun policies on gun control activism. The findings also suggest that, given the important role of the states in setting firearm policy, studies of gun activism should distinguish between gun-related threats experienced nationallyFootnote 18 (such as the Sandy Hook shooting) and those that are state specific.Footnote 19
Taking the last two sections together, the dip in attention to pro-gun policies at the end of our period of study suggests not that gun-related threats became irrelevant, but instead that the threat posed by guns everywhere policies diminished relative to the threat posed by mass shootings. Together, these two types of threats have encouraged sustained pro-regulation activism at a level that exceeds pro-gun activism. Figure 6 charts the two major sources of threat—mass shootings (event-based threat) and pro-gun legislation (policy threat) separately and alongside the share of letters that reference at least one of these two types of threat. As the figure shows, threat went from occasionally motivating letters (e.g., 1984 and 1989) to consistently and increasingly doing so for the last two decades—to the point where threat is a motivating factor in most letters. Threat is sustaining public attention to gun issues.
Figure 6 Percentage of “anti-gun” letters written in response to threats
Why Does Threat Mobilize? Contextual, Individual, and Organizational Linkages
Having presented evidence suggesting that threat—a contextual as opposed to individual or organizational factor—has been a key driver of growth in pro-regulation participation, we can probe our data for hints about potential mechanisms underlying our contextual explanation; we can explore, in other words, why and how threat spurs activism.
One potential explanation has to do with personalization: These event- and policy-centered developments may bring the stakes of the gun debate closer to home for individuals, causing them to feel a greater sense that “this could be me” or “this could be my child.” We coded letters for these types of personalization and find suggestive evidence for each.
Personal threat: “After Pittsburgh and Poway, I feel nervous walking into my synagogue”. Footnote 20 We coded the letters for signals of personal threat—for example, that an individual feels uneasy going into public spaces and fears for their own safety or that of their loved ones.Footnote 21 Starting in the mid-to-late 2000s, the letters demonstrate an upward trend (with some recent downward movement), in both the number and percentage of pro-regulation letters “personalizing” the issue of guns in society. Figure 7 shows these trends, which generally coincide with the expansion of pro-gun policies we discussed earlier.
Figure 7 Percentage and number of “anti-gun” letters expressing personal threat
*Dotted lines are five-year moving averages.
Threat to children: “‘Mommy, is someone shooting people at my school?’” Footnote 22 As figure 8 shows, increasing percentages and numbers of pro-regulation letters mention guns in connection with threats to children. Since the mid-1980s, around 20%–40% of letters have invoked children as a rationale for tightening gun laws. However, the five-year moving average of child-framed letters (represented by the dotted line) ebbs and flows predictably, with spikes around high-profile school shootings. Thus, the “child frame” did not become a dominant driver of sustained pro-regulation engagement, suggesting the gun issue has not been redefined as primarily about children’s safety and well-being.
Figure 8 Percentage and number of “anti-gun” letters discussing threat to children
*Dotted lines are five-year moving averages.
In short, our findings hint that the personalization of threat—people’s fears about their own well-being and that of their children—partially explain why contextual threats drive mobilization. Our somewhat inconclusive results could be due to mechanisms we have not measured, or it may be that letters to the editor are an imperfect medium for measuring psychological processes. Further data could shed additional light on the extent to which pro-regulation people personalize the threat of guns.
Partisanship: “The Vigilante Mentality … Condoned by Many Republican … Politicians” Footnote 23
We now turn to the role of partisan polarization. As noted earlier, polarization and the negative partisanship associated with it serve as facilitating factors that alter how individuals interpret and respond to threats. Polarization on guns was catalyzed by the NRA’s alignment with the Republican Party in the early 1980s and has grown since then, with the parties moving further apart on the issue and mass-level partisans increasingly adopting their parties’ stances on gun regulation (Lacombe Reference Lacombe2021; Pew Research Center 2018). We argue that this development has played a role in encouraging greater pro-regulation activism. With support for gun control linked to Democratic partisan identification and opposition to gun control more closely associated with the GOP, threatening events are now more likely to be viewed through a partisan lens, which can spur participation among individuals who are upset by those events and feel animosity toward Republicans. Existing research lends support to this notion, demonstrating, for example, that Republican candidates have started to suffer electorally in places that have experienced school shootings (García-Montoya, Arjona, and Lacombe Reference García-Montoya, Arjona and Lacombe2022). These patterns are perhaps unsurprising in light of research linking generalized (as opposed to issue-specific) polarization to activism (Mason Reference Mason2018; Hetherington Reference Hetherington, Brady and Nivola2008; Abramowitz and Saunders Reference Abramowitz and Saunders2008; Simas and Ozer Reference Simas and Ozer2021; Wagner Reference Wagner2021).
We find evidence suggestive of these dynamics within the letters to the editor written in support of gun regulation. As figure 9 shows, the number and proportion of such letters expressing animosity toward RepublicansFootnote 24 has increased in recent years. Notably, we do not find changes in a similar variable capturing out-group negativity (i.e., animus toward gun rights supporters), which has remained high but flat over the course of our period of study. Individuals motivated to write pro-regulation letters, in other words, have durably disliked the NRA and its supporters. The key change—coinciding with a greater number of such letters—relates to individuals’ linkage of that animosity to the Republican Party.
Figure 9 Percentage and number of “anti-gun” letters displaying negative partisanship
*Dotted lines are five-year moving averages.
This section has presented evidence consistent with our claim that contextual factors have a played a key role in encouraging pro-regulation activism and helping to produce a situation in which pro-regulation activists—after years of being at a disadvantage—now outnumber pro-gun activists. The next section examines some potential alternative explanations that seemed promising but did not pan out.
Accessories after the Fact: Individual and Organizational Drivers
One individual-level factor we initially thought might be relevant is growth in pro-regulation activism among mothers. The political dynamics of issues change when broad understandings become reframed (Baumgartner et al. Reference Baumgartner, Berry, Hojnacki, Kimball and Leech2009) and when the scope of conflict expands to include sympathetic bystanders (Schattschneider Reference Schattschneider1960). Within the gun debate, gender is critical to both processes. Women have been under-mobilized bystanders (Goss Reference Goss2006) and targets of recruitment by both pro-gun and pro-regulation forces (Goss Reference Goss2017). Of particular interest here have been periodic efforts to reframe gun violence as a threat to children and gun regulation (or deregulation) as a means of protecting them. This “maternal” reframing informed two well-financed organizational pushes on the gun-regulation side—the Million Mom March, beginning in 1999, and Moms Demand Action, beginning in 2012. Figure 10 shows the number of letters over time, broken down by the gender and position of the writer, allowing us to examine whether pro-regulation women are responsible for the big shift in the public debate over guns.
Figure 10 Number of letters by gender & gun control stance, 1980–2019
As the figure shows, many forces were at work in shaking up public discourse on guns. The patterns are gendered in both hypothesized and unexpected ways and yield several key findings. First, consistent with existing work, we find that pro-gun men led all other groups in voicing public opinions on firearm policy for the first three decades of the series. But they lost their dominant place around 2010 and have continued to slide. Second, contrary to expectation, pro-gun men’s declining engagement is not the mathematical result of displacement by newly energized “gun control moms.” While pro-regulation women’s engagement has increased, nearing and in recent years exceeding participation by pro-gun men, pro-regulation men’s engagement has increased even more. To be sure, women were entrepreneurial in developing organizations and mobilizing frameworks, and these efforts no doubt bore fruit. But whatever forces have made gun regulation an enduring part of the public agenda would appear to transcend single-issue organizing efforts and gendered frameworks. The shifting politics of guns appears to be a story about ideology more than gender, at least on the pro-regulation side. On the gun rights side, however, gender does come into play. Pro-gun men’s voices were becoming quieter in the pages of America’s newspapers, while pro-gun women—never prominent to begin with—were not speaking more loudly. Instead, they continued to be a faint echo. Even as pro-regulation women increased their participation, their pro-gun sisters did not.
Another alternative explanation consists of an organizational factor: Perhaps the patterns we explore can be explained by the ascension of pro-regulation interest groups. However, we find that sustained participation on behalf of gun control, which began in the mid-2000s and was pronounced by 2010, preceded the growth of gun control advocacy groups. In 2010, there were five significant national gun violence prevention organizations—the Brady Campaign, the Coalition to Stop Gun Violence, Mayors Against Illegal Guns, the Legal Community Against Violence, and the Violence Policy Center. The combined budgets of these organizations totaled approximately $12 million, which had been essentially flat for at least five years. However, after the Sandy Hook school shooting at the end of 2012, new organizations were established, notably Everytown for Gun Safety and Moms Demand Action (which merged in 2013); Sandy Hook Promise; and Americans for Responsible Solutions (now Giffords). By 2019, the national movement’s combined budgets had soared to $162 million, more than thirteen times the 2010 figure.Footnote 25
Although political scientists typically think of mass participation as a top-down process—fueled by the recruitment efforts of social movement organizations, interest groups, and political parties—our data suggest that the process was more bottom-up, at least initially. For one, pro-control letter writing occurred during a “doldrums” period for gun control organizations, preceding the post-2010 push to build a broad-based grassroots movement. Older gun control groups might have served as “abeyance structures” (Rupp and Taylor Reference Rupp and Taylor1987) that sustained communities of interest in anticipation of political opportunities for expanded mobilization. Likewise, and consistent with a bottom-up mobilization process, our qualitative analysis of letters to the editor finds little evidence of consistently repeated interest group talking points. The newly formed groups no doubt played a role in mobilizing their supporters—indeed, after they arrived, our data show continued increases and even acceleration in the number of letters written in support of gun control.Footnote 26 However, in terms of catalyzing these trends, our data show that organizations played a lagging role.
Conclusion
We started with an empirical puzzle: When does an issue “whose time has come” stick around? The case of gun regulation suggests an answer: When events and policy choices interact so frequently and synergistically that they pose an ongoing experience of threat among everyday individuals.
Gun violence is a serious problem in the United States. More than 45,000 people died by gunshot in 2020 (U.S. Centers for Disease Control and Prevention 2022). Studies, including ours, suggest that Americans are concerned about gun violence and, starting around 2010, have become increasingly supportive of stricter firearm regulations (Gallup Organization 2022). This study documents a fundamental shift in public engagement on the gun issue, marking a potential turning point in a policy subsystem historically dominated by participation on the pro-gun side. Led by the NRA, gun rights supporters—despite comprising a minority of the public—have been louder and more durably engaged than their opponents, which has often impacted gun policy outcomes (Lacombe Reference Lacombe2021; Goss Reference Goss2006). However, our evidence suggests that gun regulation has become a mobilizing issue for progressives, as well. This shift may portend changes in other forms of participation, the legislative agenda, and policy outcomes—something that passage of the Bipartisan Safer Communities Act of 2022 suggests may have begun.
Besides being substantively important, the gun issue provides a window into theories of public agenda-setting and mass political engagement. As we argue, the vast literatures on these processes touch on common intellectual questions but don’t speak to each other in ways that could answer fundamental questions of American politics. The case of gun politics helps connect the theoretical dots. We contend that features of the political context—such as focusing events, policy proposals and enactments, and polarization and negative partisanship—play a role in mass political engagement that has been heretofore underappreciated.
This observation, however, does not negate the important role of individual- and organization-level factors in creating the capacity, efficacy, opportunity, and collective space for people to exercise their public voice or otherwise engage in politics. Indeed, one insight from this study is that contextual factors no doubt operate not only at the “meta” level, but also at the individual level (through emotions, identities, and other psychological dynamics) and organizational level (through the mobilization of resources and recruitment).
This study calls our attention to the role of threat, particularly when coupled with negative partisanship, in structuring not just episodic participation, but also the sort of continuous engagement that makes some issues enduring features of the public agenda. We offer a perspective on how once-cyclical issues might become so commonplace, and so concerning, that they become “baked into” the structure of politics. Our findings suggest that polarization—while producing some well-documented negative effects—can also help drive civic engagement in response to unresolved problems, such as gun violence. Given that our study is based on just one case (albeit a high-profile one), future work might test and refine the theory we have offered here.
Supplementary Material
To view supplementary material for this article, please visit http://doi.org/10.1017/S1537592723001007.
Acknowledgement
The authors are grateful for the careful and incisive comments of the reviewers. For excellent research assistance, they thank Krysten Pringle. They are also grateful for useful feedback provided by participants at the 2022 APSA meeting, especially their discussant, Matt Grossmann.
Open access
