Content uploaded by Wolfgang Stroebe
Author content
All content in this area was uploaded by Wolfgang Stroebe on Apr 21, 2023
Content may be subject to copyright.
When Mass Shootings Fail to Change Minds About the Causes of Violence:
How Gun Beliefs Shape Causal Attributions
Wolfgang Stroebe
1
, Maximilian Agostini
1
, Jannis Kreienkamp
1
, and N. Pontus Leander
1, 2
1
Department of Social and Organizational Psychology, University of Groningen
2
Department of Psychology, Wayne State University
Objective: We developed and tested a gun-blame attribution model to explain why mass shootings do not
change attitudes towardgun control among gun owners and/or conservatives. For a mass shooting to increase
gun control support, individuals mustattributethe shooting at least partly to gun availability. Suchattributions
are unlikelyfor individuals who believe that there would be less crime if more people had guns. Method: After
two mass shootings, we assessed political orientation, gun ownership, the belief that widespread gun
ownership reduces crime, causal attributions about the mass shootings, and attitudes toward gun control
(Orlando, N=1756; El Paso, N=910). Data were analyzed using multiple regression (Study 1) and path
analyses (Study 2). Demographic information is reported in the Supplemental Material.Results: Across both
shootings, political conservatism and gun ownership positively predicted a belief that widespread gun
ownership reduces crime, which subsequently predicted less blaming of gun availability for mass shootings
and less support for stricter gun laws. Conclusions: Findings support our gun-blame attribution model. Mass
shootings predict people’s attitude toward stricter gun laws if they attribute the mass shooting to gun
availability. Such attributions are unlikely for U.S. gun owners and/or conservatives, who are more likely to
believe that widespread gun ownership reduces crime. To the extent that this belief is ideological, persuasion-
based psychological interventions are unlikely to be as effective as political intervention.
Keywords: belief gun ownership reduces crime, political orientation, mass shooting, support for stricter gun
laws, gun-blame attributions
Supplemental materials: https://doi.org/10.1037/vio0000431.supp
A predictable reaction to any mass shooting is a call by survivors,
family members, and politicians for stricter gun control laws (e.g.,
Shabad, 2016). In this article, we propose a psychological explanation
for why such calls have been unsuccessful: if mass shootings are to
change attitudes toward gun control, people need to believe that ease
of access to firearms is a risk factor for mass shootings. Unless one
makes this attribution, increased gun control is unlikely to be
perceived as a solution. Given that a majority of Republicans and
gun owners subscribe to the belief that, if more people had guns, there
would be less crime (and thus also fewer mass shootings, PEW,
2017a), mass shootings will not persuade them of the need for stricter
gun control laws. After a literature review that supports the associa-
tion between gun ownership/political orientation and the belief that
widespread gun ownership reduces crime, we report data from studies
of two U.S. mass shootings, which support our hypothesis.
After the mass shooting in Boulder, Colorado (March 22, 2021),
the second mass shooting in 1 week, Steve Chaggaris at Aljazeera
(2021) commented. “The political rhetoric in the wake of these
shootings has become all-too familiar—a postmassacre cycle
featuring Democratic calls for stricter controls followed by Republi-
can outrage atthe idea and/or alternative proposals that do not involve
stricter controls ::: .”A similar pattern emerges at the state level. A
study on the impact of mass shootings that occurred between 1989
and 2014, on state-level gun policy, observed that a single mass
shooting led to a 15% increase in the number of firearm bills
introduced within 1 year following the mass shooting (Luca et al.,
2020). However, this legislative action did not result in stricter gun
control laws. Rather, in states controlled by Republicans, there was a
doubling in laws that loosened gun control in the year after mass
shootings, with no significant change in Democratic-controlled states.
Mass shootings appear to have only slight effects in public
opinion polls. Gallup (2020) reported that, in the last three decades,
Americans’preferences for stricter gun control typically peaked
after mass shootings and then decreased as the memory of the
shooting had faded. Thus, during the COVID-19 related lull in mass
shootings, support for stricter gun laws declined to 57%, its lowest
level since 2016 (Gallup, 2020). However, this decline has mainly
been due to a 14-point drop for Republicans, of whom only 22%
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This article was published Online First May 12, 2022.
Wolfgang Stroebe https://orcid.org/0000-0003-2767-2838
Maximilian Agostini https://orcid.org/0000-0001-6435-7621
Jannis Kreienkamp https://orcid.org/0000-0002-1831-5604
N. Pontus Leander https://orcid.org/0000-0002-3073-5038
The authors declare that they have no conflict of interest. The authors
received no specific funding for these studies. All authors are members of the
Center for Psychological Gun Research (https://gunpsychology.org). Proto-
cols, materials, analysis data, and code are available at https://osf.io/v8ndy/
(Agostini et al., 2021).
Correspondence concerning this article should be addressed to
Wolfgang Stroebe, Department of Social and Organizational
Psychology, University of Groningen, Grote Kruisstraat 2/1, Groningen
9712 TS, The Netherlands. Email: wolfgang.stroebe@gmail.com
Psychology of Violence
© 2022 American Psychological Association 2022, Vol. 12, No. 5, 305–313
ISSN: 2152-0828 https://doi.org/10.1037/vio0000431
305
favored stricter gun laws in November 2020. In contrast, 85% of
Democrats favored stricter gun laws.
That mass shootings do not increase gun owners’support for
gun control was previously reported in a cross-sectional study,
which measured support for gun control immediately before and
after the 2016 Orlando nightclub shooting (Stroebe et al., 2017a).
Whereas people who did not own guns were more supportive of
stricter gun laws after the shooting, gun owners showed no change.
However, this analysis mainly focused on gun owners versus
nonowners and did not assess political orientation as possible
moderator.
Why Should Mass Shootings Change Attitudes
Toward Gun Control?
The expectation that mass shootings should persuade Americans of
the need for stricter gun control laws assumes that everyone blames the
shootings (at least partly) on ease of access to firearms. Although it
might seem obvious—to liberals and gun control proponents—that a
terrible mass shooting should result in attitude change toward support
for stricter gun control laws, such attitude change is neither obvious
nor logical to people who believe there would be less crime, if more
people had guns. Such change may require individuals to make a gun-
blame attribution; that is, they would have to believe that stricter gun
control laws would have made a difference. People are only likely to
make this attribution if they believe stricter gun control laws reduce the
risk of gun violence. However, blaming gun availability is inconsistent
with the pro-gun belief that an armed citizenry reduces violent crime.
As stated by Wayne LaPierre, the executive vice president of the
National Rifle Association (NRA), “The only thing that stops a bad
guy with a gun is a good guy with a gun”(Columbia Broadcasting
System [CBS], 2012).
1
This opinion seems to be shared by many Republicans and gun
owners. In a 2017 survey by the PEW Research Center, 56% of
Republicans (vs. 15% of Democrats) agreed there would be less
crime, if more Americans owned guns (PEW, 2017b). The agree-
ment among Republican leaning gun owners was even higher,
namely 71% compared to only 24% among Democrat leaning
gun owners (PEW, 2017b). Similarly, only 22% of Republicans
and 24% of gun owners (but 68% of Democrats and 58% of gun
nonowners) agreed there would be fewer mass shootings if it were
harder to legally obtain guns (PEW, 2019).
There is also evidence, from earlier opinion polls, that is consis-
tent with the assumption that political party affiliation polarizes
attitudes, around gun control, after mass shootings. In the year 2000,
when asked whether protecting the right of Americans to own guns
was more important than controlling gun ownership, 38% of Re-
publicans (vs. 20% of Democrats) considered the right to own a gun
more important; in 2017—despite several high-profile mass
shootings—agreement among Republicans was up to 76% (vs.
22% among Democrats; PEW, 2017b). By Fall 2020, there was
a record-high 63-point gap in attitudes toward stricter gun control
laws between Republicans and Democrats (Gallup, 2020). Whereas
Democrats maintained high levels of support for gun control after a
period of mass shootings, support decreased among Republicans.
A study by Barney and Schaffner (2019) found support for a
polarization due to ideology. These researchers were initially con-
cerned with a reanalysis of findings, reported by Newman and
Hartman (2019), that people’s geographical proximity to a mass
shooting determined whether their support for gun control
increased. Barney and Schaffner (2019) instead observed an inter-
action of geographic distance and political orientation, indicating
“that mass shooting events may cause Democrats to become more
supportive of increasing gun control regulations while Republicans
become less supportive of doing so”(p. 1563).
Although Rogowski and Tucker (2019) failed to find evidence of
such polarization after the Sandy Hook shooting, this could have
been due to asking whether federal law should ban the possession of
handguns, except by law enforcement personnel. Given that the
Sandy Hook shooter used a semiautomatic rifle and not a handgun,
this question was not necessarily relevant to that shooting. Further-
more, there is much less support for a ban on handguns and, at 31%,
the gap between Democrats and Republicans on handguns is smaller
than the gap in support for stricter gun laws (Gallup, 2020).
It is interesting to note that the polarizing effect, of party affilia-
tion, could be moderated by the level of anxiety aroused by a mass
shooting, as well as the framing of questions in opinion polls. Joslyn
and Haider-Markel (2018) argued that anxiety inhibits people’s
reliance on habitual information processing and increases support
for institutions perceived as protective. In support of their hypothe-
ses, they found that anxiety substantially decreased the difference
between conservatives and liberals in blaming mass shootings on
gun availability, with anxious conservatives increasing their blame
on gun availability. Importantly, anxiety also increased support for
stricter gun laws among conservatives.
Haider-Markel and Joslyn (2001) pointed out that the response
to mass shootings could be moderated by the answer alternatives
offered in opinion polls. Participants asked to rate potential reasons
for a recent school shooting were told that “many people blame
violence on Television”or “blame weak gun control laws”for the
shooting. Not surprisingly—given the suggestive formulation,
which is likely to act as a prime—they found that the first
alternative elicited more blame of TV violence and the second
more blame of gun control laws. Given that professional opinion
polls would have simply offered both alternatives as potential
reasons, they should have avoided the kind of priming response
reported by these authors. However, most interesting was the
finding that the priming effect interacted with political orientation.
Whereas Republicans were only influenced by the TV violence
prime, Democrats responded only to the gun control law prime.
Since primes are assumed only to increase the cognitive accessi-
bility of information stored in memory, this finding provided
interesting information about the relevant memory content of
Republicans and Democrats.
Next to the liberal-conservative split, gun owners may also resist
attributing mass shootings to a lack of stricter gun control laws,
because owning a gun is fundamentally important to their identity
and sense of freedom. In an opinion poll, half of all gun owners said
guns are important for their social identity and 74% perceived the
right to own guns as essential to their personal sense of freedom
(PEW, 2017a). Thus, gun ownership provides a social identity that
involves significant psychological, social, and political attachments.
Indeed, Joslyn and Haider-Markel (2017) argued, that in view of this
strong identification with other gun owners as their in-group, gun
owners should tend to make self-serving attributions when asked
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
1
This slogan became so popular that the NRA sold “Good Guy with a
Gun”T-shirts—now on sale for $2.95 at the official NRA store.
306 STROEBE, AGOSTINI, KREIENKAMP, AND LEANDER
about the causes of mass shootings. “For the case of violence
involving guns, gun owners should be less likely to blame guns
or gun culture, but instead deflect blame toward the other influences
acting upon the shooter”(p. 433).
A Gun-Blame Attribution Model
Building on these claims, as well as the attribution model devel-
oped by Joslyn and Haider-Markel (2017), we propose that the links
between gun ownership and conservatism, to lower support for
stricter gun control laws, can be partially explained by the causal
attributions gun owners and conservatives make about mass shoot-
ings, particularly whether they blame (or downplay) gun availability
as a likely cause. Causal attributions are the process whereby
perceivers arrive at conclusions about the causes of actions or
events (Heider, 1958;Kelley, 1967). With its focus on differences
in the belief about the association of gun ownership and crime, as the
central variable that determines whether mass shootings change
people’s attitudes toward stricter gun control, our model comple-
ments the social identity propositions of Joslyn and Haider-Markel
(2017) by focusing on whether certain individuals believe that guns
are means to safety rather than a threat to safety. Whereas the social
identity implications, of being a gun owner, may be a starting point
for how individuals prefer to respond to a mass shooting, our model
considers how a specific belief mediates that response—namely
whether people believe that there would be less crime if more people
owned guns. Thus, our model focuses on different psychological
processes and can apply to different groups of people who may be
receptive to such a belief. Although gun owners are more likely than
nonowners to believe there would be less crime, if more people had
guns, the overlap is not complete. Fifty six percent of Republicans
hold this belief, but only 41% own guns (PEW, 2017a). And 29% of
those Republican gun owners do not believe that more gun owner-
ship would reduce crime in the USA (PEW, 2017a). Among the 15%
of Democrats, who own guns, only 24% believe in the value of
armed citizenship (i.e., 24%; PEW, 2017a).
Figure 1 presents our gun-blame attribution model. The model
makes three predictions:
Hypothesis 1: The belief that widespread gun ownership re-
duces crime will be more strongly associated with support for
stricter gun control laws after a mass shooting. Mass shootings
are unlikely to increase support for stricter gun control laws
among individuals who believe that widespread gun ownership
reduces crime.
Hypothesis 2: The belief that widespread gun ownership re-
duces crime will mediate the link between gun ownership and
political orientation on support for stricter gun control laws.
Hypothesis 3: The effects of this belief will be itself be mediated
by tendencies to blame or attribute mass shootings to gun
availability.
The Present Studies
Two psychological surveys were conducted in the context of the
2016 Orlando nightclub mass shooting and the 2019 El Paso Walmart
mass shooting. Study 1 took place immediately before and after the
Orlando nightclub shooting. Study 1 tests whether the occurrence of
the Orlando mass shooting moderates the relationship between a
belief that widespread gun ownership reduces crime and attitudes
toward stricter gun control laws (Hypothesis 1). Study 2 tests the
mediator hypotheses (Hypotheses 2 and 3), using both the survey
conducted after the Orlando mass shooting and a second survey
conducted after the El Paso shooting. Study 2 tests the full gun-blame
attribution model for both shootings, predicting that conservatism (or
gun ownership) predicts agreement with the belief that widespread
gun ownership reduces crime, which in turn predicts lower gun-
availability attributions for the respective mass shootings, which
ultimately predicts (lower) support for stricter gun control laws.
For all studies, the protocols, materials, data analysis, and repro-
ducible analysis code are available via our OSF repository (Agostini
et al., 2021).
Study 1: The Orlando Mass Shooting and Support
for Stricter Gun Control Laws
In June 2016, a 29-year-old security guard with an immigration
background killed 49 people and wounded 53 others in a nightclub in
Orlando, Florida. At the time, it was the deadliest mass shooting by a
lone gunman in U.S. history. Due to the terrible coincidence that some
of us were finishing a survey of male U.S. gun owners and nonowners
just before this mass shooting occurred, we happened to collect data on
support for stricter gun control laws just before and immediately after
theshooting(Stroebe et al., 2017a,2017b).
2
In both samples, we
measured respondents’belief that widespread gun ownership would
reduce crime as well as their support for stricter gun control laws. The
primary aim of Study 1 was to test whether the belief, that gun
ownership reduces crime, more strongly predicted support for stricter
gun laws after the Orlando shooting (Hypothesis 1).
Method
Participants
Participants were 1,877 U.S. men, recruited through the online
research firm Qualtrics Panels. One hundred twenty-one participants
were excluded due to filling out the survey on the day of the shooting
(n=81), missing data (n=9), or low-quality responses (e.g.,
straight lining, n=31). The final sample size was n=1,756. Of the
total sample, n=835 completed the study before the mass shooting
(“pre-Orlando”) and n=921 completed it afterward (“post-
Orlando”). Participants were prescreened to include equal numbers
of gun owners and nonowners (prior to exclusions), with maximum
quotas for region of country, age, education, and income. Full pre–
post Orlando demographics are available in the Supplemental
Material (Table S1). The only difference was that, unlike the pre-
Orlando sample, the post-Orlando sample had no age difference
between gun owners and nonowners (explored further in the Online
Appendix). A sensitivity analysis using a 2 ×2 between-subjects
design with one numerator degree of freedom and four groups,
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2
The data of this study have already been used in two publications
(Stroebe et al., 2017a,2017b). The findings of Stroebe et al. (2017a) reported
earlier in this article, focused on the impact of this mass shootings on gun
owners’support for stricter gun laws (i.e., a main effect). Stroebe et al.
(2017b) Used the pre–post Orlando data to test our model of defensive gun
ownership. The data collected after the mass shooting were simply used to
show that our model could be replicated in a second data set.
IMPACT OF MASS SHOOTINGS ON GUN CONTROL SUPPORT 307
showed that this sample size is able to detect even small two-way
interactions (partial-eta squared ∼.003) with 80% power (Faul et
al., 2007).
Procedure
The first part of the study was conducted online between May 31
and June 11, 2016. Just as data collection was nearing completion, the
Orlando nightclub shooting happened on June 12. We decided to
collect a second sample rather than recontacting the original sample.
The main reason was concerns that asking respondents the same set of
questions only days after the first survey would result in response bias.
Measures
Gun ownership was assessed at the beginning of the questionnaire
by asking participants whether they owned a gun (yes/no). Political
orientation was assessed by asking respondents about their political
orientation and to locate themselves on a dimension from “1:
extremely liberal”to “9: extremely conservative”,
3
M=5.09, SD
=2.19. Belief in the crime-reducing effect of greater gun availability
was assessed with the question: “In general, if more people had
guns, there would be less crime,”M=3.85, SD =2.14. Responses
ranged from “1: strongly disagree”to “7: strongly agree.”Support
for stricter gun control laws was measured with three items (a) “In
general, do you believe the laws covering the sale of firearms should
be made more strict, less strict, or kept as they are now?”,(“1: much
less strict”to “7: much more strict”), (b), “Do you support or oppose
some kind of registry of all guns, at least at the state-government
level?,”(“1: strongly oppose a gun registry”to “7: strongly support a
gun registry”), (c) “Do you support or oppose laws that create ‘gun-
free zones’at schools and other public places?,”(“1: strongly
oppose ‘gun-free zones’” to “7: strongly support ‘gun-free zones’”),
α=.75, M=4.88, SD =1.58.
Results
We tested whether the association between the belief that gun
ownership reduces crime and support for stricter gun control laws
was indeed stronger after the Orlando mass shooting. Any between-
groups differences, pre-versus-post Orlando, would be consistent
with attitude change rather than a stable belief system. A multiple
regression analysis predicted support for stricter gun control laws
from the belief that widespread gun ownership reduces societal
crime (“belief guns reduce crime,”standardized), pre–post Orlando
(coded −0.5, 0.5), and the interaction of that belief and pre–post
Orlando. The model explained a significant proportion of variance in
support for gun control laws, R2
adj =.23, F(3, 1752) =174, p<.001.
There was no direct effect of pre–post Orlando per se,β=.11, 95%
CI [−0.03, 0.24], t(1752) =1.58, p=.114, η2
part =.001; but, as
expected, there was a strong direct effect of the belief that guns reduce
crime, β=−.75, 95% CI [−0.81, −0.68], t(1752) =−22.58, p<.001,
η2
part =.23, and a theoretically consistent interaction between that
belief and pre–post Orlando, β=−.13, 95% CI [−0.26, −0.002],
t(1752) =−1.99, p=.047, η2
part =.002.
The interaction suggests a pre–post Orlando difference in the link
between the belief that guns reduce crime and support for stricter
gun control. Indeed, simple slopes analyses indicated that the
already-strong negative association that was observed pre-Orlando
(β=−.68, 95% CI [−0.78, −0.59], t(1752) =−14.22, p<.001), was
even stronger post-Orlando, β=−.82, 95% CI [−0.90, −0.73],
t(1752) =−17.82, p<.001. As illustrated in Figure 2, this was
mainly driven by people who disagreed that beliefs that guns reduce
crime (i.e., −1 SD belief guns reduce crime) becoming even more
supportive of stricter gun control laws post-Orlando, β=.24, 95%
CI [0.01, 0.42], t(1752) =2.52, p=.012. Participants who believed
that guns reduce crime (+1SD) did not differ pre–post Orlando; they
remained opposed to gun control laws to the same extent, β=−.03,
95% CI [−0.21, −0.16], t(1752) =−0.29, p=.773. This supports
our model prediction that mass shootings only increase support for
stricter gun control laws among people who do not believe that more
widespread gun ownership would reduce crime in society. People
who endorse this belief are not moved by mass shootings to support
stricter gun control laws.
Discussion
This study supported the hypothesis that a mass shooting mod-
erates the link between a belief that guns reduce crime and support
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Figure 1
Attribution Model
3
Political orientation was measured in Study 1, but not part of any
hypothesis tests until Study 2. We also assessed party membership. However,
we think that these political beliefs are part of a conservative political
orientation that is not restricted to party membership. In opinion polls
this is typically reflected by the term Republican or Democrat “leaning”
individuals. Political orientation has the additional advantage of being a
continuous rather than a dichotomous variable.
308 STROEBE, AGOSTINI, KREIENKAMP, AND LEANDER
for stricter gun control laws. The interaction was not strong, but the
simple slopes were consistent with the assumption that the mass
shooting resulted in attitude change mainly among individuals who
did not believe that widespread gun ownership is a means to reduce
crime. This finding is intuitively plausible: Why should people, who
believe that greater accessibility of guns would reduce crime, be
persuaded by a mass shooting that stricter gun control laws are
needed? If conservatives and gun owners are more likely to endorse
such a belief, mass shootings are unlikely to increase their support
for stricter gun control laws.
Study 2: Testing the Gun-Blame Attribution Model
After U.S. Mass Shootings
In Study 2, we used data collected after the Orlando and the El
Paso mass shootings, to test the full gun-blame attribution model. In
the context of both shootings, we tested whether a belief that
widespread gun ownership reduces crime mediates the link between
gun ownership and political orientation on support for stricter gun
control laws (Hypothesis 2), and whether the effect of this belief can
be itself be explained or mediated by tendencies to attribute mass
shootings to gun availability (Hypothesis 3).
Study 2a: Testing the Full Model After
the Orlando Mass Shooting
Method
Participants. Participants were 910 U.S. men (gun owners and
nonowners) recruited through the online research firm Qualtrics
Panels. Of the 910 participants, we removed 16 due to missing data
on key variables. Participants were prescreened to include both gun
owners (n=435) and nonowners (n=459), with maximum quotas
for region and country, age, education, and income. Demographics
are reported in the Supplemental Material (Table S1).
Procedure. Most data were collected within the first week after
the shooting.
Measures. The measures for gun ownership, political orienta-
tion, and belief that guns reduce crime are the same as described in
Study 1.
Gun-Availability Attributions. Near the beginning of the
survey, participants were asked for their perceptions of what might
have motivated the gunman and their thoughts about what could
have prevented the mass shooting. The items were based on media
accounts at the time (13 items in total). These attributions were
formulated for the specific mass shooting. For example, after the
Orlando shooting, respondents were asked whether the mass shoot-
ing in the Orlando nightclub might have been prevented if stricter
gun control laws had been in place and/or if better health care
existed. With regard to motivating factors, respondents were asked,
“What might have motivated the gunman to commit the mass
shooting in Orlando”and then given alternatives, such as “ease
of access to firearms.”All items were consistently rated on a scale
from: (not applicable), −3: very doubtful, 0: neutral, 3: very
possible). We were specifically interested in two gun-availability
attributions: whether the mass shooting might have been pre-
vented if “::: stricter gun control laws were in place”(Orlando:
M=−0.22, SD =2.41) and whether the gunman was motivated by
“::: ease of access of firearms”(Orlando: M=0.52, SD =2.22).
4
The gun-availability attributions were embedded among various
other, sociocultural and psychological attributions, which we
would use as covariates. Our model should apply to the gun-blame
attributions irrespective of the other attributions.
5
Results
Path analyses were conducted using R(R Core Team, 2018),
RStudio (RStudio Team, 2016), lavaan (Rosseel, 2012), and semPlot
(Epskamp, 2015). These analyses have no fit statistics as all possible
paths are modeled. All direct paths were significant and almost
identical for both gun-availability attributions (see Figure 3 for
direct paths). Importantly, there were reliable indirect paths from
gun ownership/political orientation →belief guns reduce crime →
gun-availability attributions →support for stricter gun control laws
(Table 1, columns three and four). Effects remained stable when
controlling for the various other mass shooting attributions. Full
results are in the Supplemental Material.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Figure 2
Interaction Effect of Pre–Post Orlando With Belief Guns Reduce
Crime (−1 SD, Gray Line; +1 SD Black Line) Predicting Support
for Gun Laws
4
We also asked whether the mass shooting could have been prevented, “if
people at the location had been armed.”When trying to combine the three
items, scale reliability was not established. We, therefore, in line with our
original plan report “motivated by ease”and “prevented if stricter gun laws”
in the main text. The Supplemental Materials contain the “prevented if
armed”attribution. In short, all results replicate when using the “prevented if
armed”attribution.
5
The items tapped into perceptions of motivation and potential preven-
tion. For example, whether the shooting was motivated by “ideology/ISIS”
and “religion,”“compensation for inadequacy,”“desire for power, signifi-
cance, or attention,”“mental illness,”“cultural exposure to violence,”
“motivated by hatred of others; prejudice”; or whether the shooting could
have been prevented if “there was more surveillance of suspected radicals,”
“society was more cautious of immigrants,”“better mental health care
existed.”
IMPACT OF MASS SHOOTINGS ON GUN CONTROL SUPPORT 309
Study 2b: Testing the Full Model After the El Paso
Walmart Shooting
We sought to replicate the model in the context of the El Paso
mass shooting, which occurred on August 19, 2019 in a Walmart
superstore. A white, American gunman killed 23 people and injured
23 others. The victims were mostly Hispanic. The FBI investigated
the shooting as a hate crime.
6
Method
Participants. Participants in the El Paso study were 433 male
and female gun owners and 437 nongun owners (51% male)
recruited by Qualtrics Panels. Demographics are reported in Table
S3 in the online Supplemental Material.
Procedure. As with the Orlando study, Qualtrics Panels con-
ducted this survey online within 3 weeks of the mass shooting.
Measures. As per the Orlando study, we measured political
orientation (M=5.46, SD =2.07), gun ownership and belief that
guns reduce crime, (M=3.43, SD =1.30). The support for stricter
gun laws scale was reduced to two items, r=.65, 958% CI [.62, .69].
Since in contrast to the Orlando nightclub, Walmarts are not gun-
free zones, we omitted the item about gun-free zones from the scale.
Gun-blame attributions were measured the same way as in the
Orlando study. Again, we were specifically interested in the two
gun-availability attributions: whether the mass shooting might have
been prevented if “::: stricter gun control laws were in place”(El
Paso M=0.26, SD =2.41) and whether the gunman was motivated
by “::: ease of access of firearms”(El Paso M=0.79, SD =2.18).
Results
Results fully replicated the gun-blame attribution model observed
after Orlando. All direct effects were reliable (Figure 4); impor-
tantly, there was also a reliable indirect effect of political conserva-
tism →belief guns reduce crime →(lower) gun-availability
attributions →(lower) support for gun control laws (Table 1,
columns three and four). As per the Orlando study, gun ownership
and political conservatism were both linked to (less) support for
stricter gun laws via the belief that guns reduce crime (Hypothesis 2).
Second, attributions related to gun availability at least partially
mediated the effect between this belief and support for gun control
laws (Hypothesis 3). Effects remained stable when controlling for all
the other attributions. Full results are in the Supplemental Material.
Thus, the gun-blame attribution model replicated in a second mass-
shooting context.
We did not find evidence for gender meaningfully altering the
theoretical conclusions in El Paso. We did observe a small differ-
ence between the gender groups in the model for “prevented if
stricter gun laws”,χ
2
(9, N=863) =18.9, p=.026. Further assessing
path differences, the path from “belief guns reduce crime”to
“support for gun laws”differed for men and women, χ
2
(1, N=
863) =6.57, p=.01. Men had a slightly stronger effect, b=−.32,
SE =.04, than women, b=−.19, SE =.05. However, for both
subsamples the paths remained reliable and most importantly, all
indirect effects were reliable for both men and women (for more
information see the Supplemental Materials).
Discussion
Two independent tests of the mediator hypotheses supported our
gun-blame attribution model, using samples collected after two
high-profile mass shootings in the U.S. Conservatism and gun
ownership were positively associated with the belief that widespread
gun ownership would reduce crime in society, which in turn
predicted lower support for stricter gun laws (Hypothesis 2). The
indirect effect was further partly mediated by tendencies to blame or
attribute mass shootings to gun availability (Hypothesis 3). Conser-
vatism (and/or gun ownership) predicted a belief that more guns
would reduce crime, which in turn predicted lower gun-availability
attributions for mass shootings, which ultimately predicted (less)
support for stricter gun laws.
General Discussion
This article presented and tested a psychological attribution
model of the impact of mass shootings on support for stricter gun
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Figure 3
Attributions Model for Orlando
Note. We report effects for both attributions: that the mass shooting could have been prevented if stricter gun laws were in place (bold) and that ease of access
to firearms contributed to the gunman’s motivation to perpetrate the shooting (underlined).
6
We also collected data for the Dayton mass shooting within this study. In
short, all results replicate and are presented in the Supplemental Materials.
310 STROEBE, AGOSTINI, KREIENKAMP, AND LEANDER
control laws. According to this model, for mass shootings to increase
support for stricter gun control laws, individuals must blame guns as one
of the causes of a mass shooting. Such gun-blame attributions are
unlikely for individuals who believe that gun ownership is a means to
reduce crime. Given that conservatives and gun owners are more likely
to endorse that belief, they are less likely than liberals to be persuaded by
mass shooting of the need for stricter gun control laws.
Using data collected before and after the Orlando night club
shooting, we observed that the occurence of the mass shooting
moderated the link between the belief that guns reduce crime and
support for stricter gun control laws—but the effect was driven by
those who did not believe that guns reduce crime. There was no
difference in support for stricter gun control laws among those who
endorsed armed citizenship. This finding has commonsense plausi-
bility: if conservatives and gun owners believe there would be less
crime and also fewer mass shootings if more people owned guns, a
mass shooting will not persuade them of a need for stricter gun
control laws.
We assessed the mediator hypotheses, of the gun-blame attribu-
tion model, following both the Orlando and the El Paso mass
shootings. The belief that guns reduce crime predicted whether
people attributed the mass shooting to the availability of guns, which
in turn predicted (lack of) support for stricter gun control laws. This
belief also mediated the association between conservatism/gun
ownership and (lack of) support for stricter gun laws. Thus, because
gun owners and conservatives are likely to believe in armed
citizenship, they are less likely to blame guns for mass shootings
and, consequently, less likely to support stricter gun control laws.
Limitations
A challenge of mass shooting research is that pre–post studies
cannot be planned in advance, so we could never test our full model.
Although we happened to collect data both before and after the
Orlando mass shooting, we could only ask causal attributions
afterward. One could ask such attributions in general, but it is
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Figure 4
Attributions Model for El Paso
Note. We report effects for both attributions: that the mass shooting could have been prevented if stricter gun laws were in place (bold) and that ease of access
to firearms contributed to the gunman’s motivation to perpetrate the shooting.
Table 1
Indirect Paths
Mass shooting Indirect paths Attributions
Orlando prevGunlaw motEase
Gun →Belief →Attrib −0.07 (0.01)*** −0.07 (0.01)***
Gun →Belief →Laws −0.06 (0.01)*** −0.07 (0.01)***
Gun →Belief →Attrib →Laws −0.03 (0.01)*** −0.03 (0.01)***
Pol →Belief →Attrib −0.08 (0.01)*** −0.08 (0.01)***
Pol →Belief →Laws −0.08 (0.01)*** −0.08 (0.01)***
Pol →Belief →Attrib →Laws −0.03 (0.01)*** −0.03 (0.01)***
El Paso
Gun →Belief →Attrib −0.08 (0.01)*** −0.10 (0.02)***
Gun →Belief →Laws −0.06 (0.01)*** −0.06 (0.01)***
Gun →Belief →Attrib →Laws −0.04 (0.01)*** −0.05 (0.01)***
Pol →Belief →Attrib −0.13 (0.02)*** −0.15 (0.02)***
Pol →Belief →Laws −0.10 (0.01)*** −0.09 (0.01)***
Pol →Belief →Attrib →Laws −0.07 (0.01)*** −0.07 (0.01)***
Note. Gun =Gun Owner; Pol =political conservativism; Belief =belief guns prevent crime; prevGunlaw =shooting could have been prevented if stricter gun
laws were in place; motEase =shooting was motivated by ease of access to firearms; Attrib =Gun-blame attributions; Laws =support for stricter gun laws.
*** p<.001.
IMPACT OF MASS SHOOTINGS ON GUN CONTROL SUPPORT 311
questionable whether this is comparable to responses to a specific
shooting. Another limitation of the Orlando study is that we did not
have a “no mass shooting”control group, assessed at the same point
in time. However, given that the Orlando shooting became national
news within hours of occurring, such a control would not have been
possible.
A skeptic could rightly argue that the path analyses are merely
measures of established belief systems unaffected by specific mass
shootings: Gun owners and conservatives, generally, believe that
greater gun availability reduces crime, do not attribute mass shoot-
ings to gun availability, and oppose stricter gun control laws. In
other words, the gun-blame attributions may not be independent
mediating mechanisms that can be targeted for intervention, but all
part of a belief system. It was, therefore, important that the pre–post
Orlando study at least showed a significant pre–post difference in
support for stricter gun control laws, at least among individuals who
do not endorse the pro-gun belief. For gun owners and conserva-
tives, a belief system that combines believing in the value of armed
citizenship, (denial of) gun-availability-related causes, and (non)
support of stricter gun control laws, might inoculate them from
blaming guns for mass-shootings and, presumably, other gun-
related deaths (Joslyn & Haider-Markel, 2017).
Future Research Directions
To adequately test the full model, a longitudinal study is likely
required—such as an interrupted time-series design that happens to
overlap with multiple mass shooting events. The proposed process
could then be tested over time, within subjects, in a manner that
allows for a simultaneous test of the full model wherein temporal
precedence can be established and stable belief systems can be
separated from situational fluctuations in specific beliefs.
The present study helps to establish items for the development of
scales that apply across mass shooting contexts: Given our rapid
response approach, our measures were brief or not fully validated
prior to our studies, and we relied on fast online samples. The use of
single-item measures and convenience samples could increase error
variance in our samples. Nevertheless, we were able to replicate the
basic model across multiple mass shooting events (a third is pre-
sented in the Supplemental Materials), while exchanging items that
were conceptually related. Based on these developments, the model
could be replicated with related or extended constructs.
Prevention and Policy Implications
The implications of our results are rather bleak, at least with
respect to potential interventions. The need for safety and security is
a fundamental need, which, at least for U.S. gun owners, appears to
be linked to guns. The belief that widespread gun ownership reduces
crime could be embedded in a belief system that links a diffuse belief
in a dangerous world to the need to own a gun for self-defense
(Altemeyer, 1988;Kreienkamp et al., 2021;Stroebe et al., 2017b).
Nearly 70% of handgun owners report that they own their gun for
self-defense (PEW, 2017b;Stroebe et al., 2017b). Any intervention
aimed at reducing opposition to stricter gun control laws will thus
have to change multiple, interrelated beliefs about the risks and
causes of gun violence and the efficacy of armed citizenship to
promote public safety. According to social psychological
consistency theories, beliefs that are embedded in such extensive
belief systems are difficult to change (e.g., McGuire, 1969,1981).
References
Agostini, M., Kreienkamp, J., Leander, N. P., & Stroebe, W. (2021). When
mass shootings fail to change minds about the causes of violence: How
gun beliefs shape causal attributions [OSF project: protocol, materials,
data, software]. https://osf.io/v8ndy/
Aljazeera. (2021). Will the Colorado shooting prompt gun reform? https://
www.aljazeera.com/news/2021/3/24/will-colorado-shooting-prompt-us-gun-
reform
Altemeyer, B. (1988). Enemies of freedom: Understanding right-wing
authoritarianism. Jossey-Bass.
Barney, D. J., & Schaffner, B. F. (2019). Reexamining the effect of mass
shootings on public support for gun control. British Journal of Political
Science,49(4), 1555–1565. https://doi.org/10.1017/S0007123418000352
Columbia Broadcasting System. (2012). NRA: “Only way to stop a bad guy
with a gun is a good guy with a gun.”https://washington.cbslocal.com/
2012/12/21/nra-only-way-to-stop-a-bad-guy-with-a-gun-is-with-a-good-
guy-with-a-gun/
Epskamp, S. (2015). semPlot: Unified visualizations of structural equation
models. Structural Equation Modeling: A Multidisciplinary Journal,
22(5), 474–483. https://doi.org/10.1080/10705511.2014.937847
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A
flexible statistical power analysis program for the social, behavioral, and
biomedical sciences. Behavior Research Methods,39(2), 175–191. https://
doi.org/10.3758/BF03193146
Gallup. (2020, November 16). Support for stricter U.S. gun laws at lowest
level since 2016.https://news.gallup.com/poll/325004/support-stricter-
gun-laws-lowest-level-2016.aspx
Haider-Markel, D. P., & Joslyn, M. R. (2001). Gun policy, opinion, tragedy,
and blame attribution: The conditional influence of issueframes. The Journal
of Politics,63(2), 520–543. https://doi.org/10.1111/0022-3816.00077
Heider, F. (1958). The psychology of interpersonal relations. Wiley. https://
doi.org/10.1037/10628-000
Joslyn, M. R., & Haider-Markel, D. P. (2017). Gun ownership and self-
serving attributions for mass shooting tragedies. Social Science Quarterly,
98(2), 429–442. https://doi.org/10.1111/ssqu.12420
Joslyn, M. R., & Haider-Markel, D. P. (2018). The direct and moderating
effects of mass shooting anxiety on political and policy attitudes. Research
& Politics,5(3), Article 2053168018794060. https://doi.org/10.1177/
2053168018794060
Kelley, H. H. (1967). Attribution theory in social psychology. In D. Lincoln
(Ed.), Nebraska symposium on motivation (Vol. 15, pp. 431–439).
University of Nebraska Press.
Kreienkamp, J., Agostini, M., Leander, N. P., & Stroebe, W. (2021). How
News Exposure and Trust in Law Enforcement Relate to Defensive Gun
Ownership. Psychology of Violence,11(4), 417–427. https://doi.org/10
.1037/vio0000375
Luca, M., Malhotra, D., & Poliquin, C. (2020). The impact of mass shootings
on gun policy. Journal of Public Economics,181, Article 104083. https://
doi.org/10.1016/j.jpubeco.2019.104083
McGuire, W. J. (1969). The nature of attitudes and attitude change. In G. E.
Lindzey & E. E. Aronson (Eds.), Handbook of social psychology (2nd ed.,
Vol. 3, pp. 136–314). Addison-Wesley.
McGuire, W. J. (1981). The probabilogical model of cognitive structure and
attitude change. In R. E. Petty, T. M. Ostrom, & T. C. Brock (Eds.),
Cognitive responses to persuasion (pp. 291–308). Erlbaum.
Newman, B., & Hartman, T. (2019). Mass shootings and public support for
gun control. British Journal of Political Science,49(4), 1527–1553.
https://doi.org/10.1017/S0007123417000333
PEW. (2017a). The demographic of gun ownership.https://www.pewsocia
ltrends.org/2017/06/22/the-demographics-of-gun-ownership/
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
312 STROEBE, AGOSTINI, KREIENKAMP, AND LEANDER
PEW. (2017b). Views of guns and gun violence.https://www.pewsocialtre
nds.org/2017/06/22/views-of-guns-and-gun-violence/
PEW. (2019). 7 facts about guns in the USA.https://www.pewresearch.org/
fact-tank/2019/10/22/facts-about-guns-in-united-states/
R Core Team. (2018). R: A language and environment for statistical
computing. R Foundation for Statistical Computing. https://www.r-
project.org/
Rogowski, J., & Tucker, P. (2019). Critical events and attitude change:
Support for gun control after mass shootings. Political Science Research
and Methods,7(4), 903–911. https://doi.org/10.1017/psrm.2018.21
Rosseel, Y. (2012). lavaan: An R package for structural equation modeling.
Journal of Statistical Software,48(2), 1–21. https://doi.org/10.18637/jss
.v048.i02
RStudio Team. (2016). RStudio: Integrated development environment for R.
RStudio. http://www.rstudio.com/
Shabad, R. (2016). Obama renews his call for gun control after Orlando.
http://www.cbsnews.com/news/obama-reiterates-his-call-for-gun-control-
action-after-orlando/
Stroebe, W., Leander, N. P., & Kruglanski, A. W. (2017a). The impact of the
Orlando mass shooting on fear of victimization and gun-purchasing
intentions: Not what one might expect. PLOS ONE,12(8), Article
e0182408. https://doi.org/10.1371/journal.pone.0182408
Stroebe, W., Leander, N. P., & Kruglanski, A. W. (2017b). Is it a dangerous
world out there? The motivational bases of American gun ownership.
Personality and Social Psychology Bulletin,43(8), 1071–1085. https://
doi.org/10.1177/0146167217703952
Received December 19, 2019
Revision received March 21, 2022
Accepted March 30, 2022 ▪
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
E-Mail Notification of Your Latest Issue Online!
Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at https://my.apa.org/portal/alerts/ and you will
be notified by e-mail when issues of interest to you become available!
IMPACT OF MASS SHOOTINGS ON GUN CONTROL SUPPORT 313
