Weapon Carrying Among Boys in US Schools by Race and/or Ethnicity: 1993–2019

We used data from the 1993–2019 YRBS surveys. Methods for recruitment and administration of the YRBS have been previously published.²²  In brief, eligible YRBS participants include a representative sample of students among US public and private high schools. Students complete the surveys during the school day, administered by teachers and school administrators. All data are self-reported and remain confidential. Annual and the combined 1991–2019 data are publicly available, including questions on weapon carrying at schools added in 1993. We used the publicly available combined 1993–2019 national YRBS data set with responses from 205 068 students, of which 88 393 were male, NH Black/AA, Hispanic, or NH white, and with completed data on weapon-carrying at school.

We categorized race and/or ethnicity as NH white versus NH Black/AA versus Hispanic (the latter including the YRBS categories “Hispanic/Latino” and “multiple-Hispanic”). Small sample sizes prohibited quantitative analyses of other racial and/or ethnic groups.

Our main outcome was self-reported weapon carrying at school in the past 30 days (“During the past 30 days, on how many days did you carry a weapon, such as a gun, knife, or club, on school property?”). We used the dichotomous version: never (0 days) versus ever bringing a weapon into school (1 day, 2 or 3 days, 4 or 5 days, 6 or more days).

We stratified by age, perceived safety at school, and school violence. Age was dichotomized as ≤16 years versus ≥17 years. Participants’ safety and violence experience at school (termed “school climate” in this article) was derived from the following: (1) perceived school safety (“During the past 30 days, on how many days did you not go to school because you felt you would be unsafe at school or on your way to or from school?”), (2) having been threatened and/or injured (“During the past 12 months, how many times has someone threatened or injured you with a weapon, such as a gun, knife, or club, on school property?”), and (3) getting into physical fights (“During the past 12 months, how many times were you in a physical fight on school property?”). We categorized each of these variables as never (0 times) versus ever (≥1 time) versus more often (≥2 times).

We conducted a time trend analysis using weighted frequencies of carrying a weapon at school in the past 30 days (1993–2019) by survey year, race and/or ethnicity, and age group. We depicted joint probabilities of race and/or ethnicity in each age group with self-reported weapon carrying at school, aggregated over all schools and differentiating between schools according to students’ self-reported school climate (safety, injuries or threats, and physical fights). Rather than pooling the data, disaggregating by survey year, racial and/or ethnic group, and school climate captures variations in time trends that may be traced to policy and other structural changes over time. We visualized these time series using the R ggplot2 and tidyverse packages, showing smoothed curves, weighted on the basis of the inverse variance of the point estimates.^23,24  We visualized school climates by age and racial and/or ethnic group from 1993 to 2019 in the same manner. We used logistic regression by survey year to determine linear time trends in weapon carrying at school. All quantitative analyses accounted for the YRBS survey weights, which make the sample representative of US high schools, by using the SURVEYFREQ and SURVEYLOGISTIC procedures in SAS 9.4, Cary, NC (SAS Institute, Inc, Cary, NC).²⁵ 

Overall prevalence of weapon carrying in schools dropped substantially between 1993 and 2019. These trends were consistent across all racial and/or ethnic and age groups. Stratified analyses by students’ experience of safety and violence in schools also revealed consistent declines in weapon carrying (linear time trend, P < .0001 for all groups) (Figs 1–4). In 2017 and 2019, 4.6% (95% confidence interval [CI]: 3.9%–5.4%) of all boys reported bringing a weapon into school (across all schools), without evidence for significant differences in weapon carrying by race and/or ethnicity (Table 1).

In 2017 and 2019, 93.3% (95% CI: 92.5%–94.1%) of all boys reported that they never missed school in the past 30 days because they felt unsafe, 92.3% (95% CI: 91.6%–92.9%) reported that they were never threatened or injured at school in the past 12 months, and 88.6% (95% CI: 87.6%–89.5%) reported that they were never in a physical fight at school in the past 12 months (Table 1).

Higher levels of negative school experience were associated with more frequent weapon carrying in all racial and/or ethnic and age groups. Prevalence of weapon carrying was at least 2 times higher among students who reported ever feeling unsafe or ever experiencing violence at school compared with those who did not (Table 1). Such negative experiences were more common among NH Black/AA and Hispanic boys compared with NH white boys. For example, 10.2% (95% CI: 7.5%–12.8%) of NH Black/AA and 9.7% (95% CI: 8.2%–11.1%) of Hispanic boys missed school at least once because of safety concerns compared with 4.5% (95% CI: 3.8%–5.3%) of NH white boys, and 10.2% (95% CI: 8.3%–12.0%) of NH Black/AA boys compared with 7.1% (95% CI: 6.3%–7.8%) of NH white boys were threatened or injured at school at least once in the past 12 months (Table 1). Experiences of perceived unsafety and threats and/or injuries at school have not substantially changed over time, especially for NH Black/AA and Hispanic boys (Supplemental Figs 5–7).

We found that at schools where students reported never missing school because of safety concerns, never being threatened or injured, or never being in physical fights (meaning in the vast majority of schools), NH white boys have been more likely to bring weapons into school compared with NH Black/AA and Hispanic boys since the late 1990s, especially among boys aged ≥17 years (Figs 2–4). In 2017 and 2019, there were no significant differences in weapon carrying at schools where students reported never missing school because of safety concerns (Table 1), but NH white boys were more likely than NH Black/AA or Hispanic boys to report bringing weapons into schools where they had never been threatened and/or injured (NH white 4.0% [95% CI: 3.0%–5.1%] versus NH Black/AA 2.9% [95% CI: 1.7%–4.1%] versus Hispanic 2.4% [95% CI: 1.7%–3.1%]), as well as in schools where they reported never having been in physical fights (NH white 3.8% [95% CI: 2.7%–4.8%] versus NH Black/AA 2.6% [95% CI: 1.5%–3.7%] versus Hispanic, 2.4% [95% CI: 1.7%–3.1%]) (Table 1).

At schools where students reported safety concerns or being in physical fights, differences in weapon carrying by race and/or ethnicity were not significant in 2017 and 2019 (Table 1). In schools where students reported having been threatened or injured, NH Black/AA and Hispanic boys more often reported bringing weapons into schools than NH white boys (2017 and 2019; NH Black/AA 28.7% [95% CI: 18.8%–38.5%] versus Hispanic 24.5% [95% CI: 17.6%–31.4%] versus NH white 13.2% [95% CI: 8.7%–17.7%]) (Table 1). In further examination of these results, we found that within the category of ever threatened, NH white boys more often reported being threatened and/or injured only 1 time compared with NH Black/AA and Hispanic boys, who more often reported being threatened and/or injured ≥10 times in the past 12 months, and more frequent threats and/or injuries were associated with greater odds of weapon carrying (Supplemental Tables 2 and 3).

As hypothesized, we found significant declines in weapon carrying in schools among high school–aged boys between 1993 and 2019 across all racial and/or ethnic and age groups and no differences in weapon carrying by race and/or ethnicity when comparing all schools in 2017 and 2019. Experiences of unsafety were linked to more frequent weapon carrying, and NH Black/AA and Hispanic boys more often reported such negative school experiences than NH white boys. We found that in schools perceived as safer, NH white boys have been more likely to carry weapons than NH Black/AA or Hispanic boys since the late 1990s, with the opposite pattern in schools where students experienced threats and/or injuries.

Racist prejudice continues to shape our perceptions of male individuals of color as more threatening and prone to criminality than white male individuals. Contrary to such stereotypes, when looking at weapon-carrying behaviors among high school–aged boys, our findings point to the opposite: aggregated over all schools, we could not find significant differences in weapon carrying by race and/or ethnicity in recent years; more compellingly, we found that in more favorable school climates, NH white boys have been more likely than NH Black/AA or Hispanic boys to bring weapons into school in the past 20 years. NH white boys, especially aged ≥17 years, and not NH Black/AA or Hispanic boys, have posed the greater threat with regard to bringing weapons into schools that were perceived as safer.

Our findings are consistent with a previous study on 2005 YRBS data²¹  but do not allow us to draw causal conclusions. However, we can provide some context. Violence and crime rates peaked in the United States in the early 1990s. Since then, crime and violence have declined dramatically in all spheres of life in the United States.²⁶  In 1994, the Gun-Free Schools Act (GFSA) was passed, with a federal mandate that students who bring weapons into schools be suspended for a year.²⁷  However, neither the GFSA nor general declines in crime explain why NH white boys have been more likely than boys of color to bring weapons into safer school environments in the past 20 years. In conjunction with the GFSA, zero-tolerance policies were implemented in US schools. These policies are characterized by severe punishment for small and serious infractions and are unevenly distributed across schools and student populations, with BIPOC students often experiencing harsher punishments than white students.^28,29  Not only have these policies been criticized as racist and counterproductive,^30,31  but our study also suggests that these policies may be ineffective in identifying a small group of white boys who have been bringing weapons into relatively safer schools at higher rates than students of color.

We found that in schools where students experienced unsafety and violence, weapon carrying was more common, and NH Black/AA and Hispanic boys were more likely to bring weapons into schools where they were threatened and/or injured. We contextualized the latter finding by revealing that the category of ever threatened and/or injured disguised, on average, more violent school experiences among NH Black/AA and Hispanic boys than among NH white boys. Together, these findings suggest that addressing school climates may be effective at reducing weapon carrying at schools.

Do our observations suggest that weapon carrying has been more restricted among boys of color than among NH white boys? This hypothesis seems plausible given persistent racism, the policing of persons of color, and evidence that white individuals are often allowed to openly carry guns, whereas individuals of color merely suspected of carrying any weapon have to fear for their lives.^12,13,32  Furthermore, Black students are more often affected by school shootings; however, school shootings in primarily white schools tend to result in a greater number of casualties per incident,⁴  suggesting different intent or more dangerous weapons. School shootings that received wide media attention, such as Sandy Hook and Stoneman Douglas, were conducted by white perpetrators, but we were unable to find published summary data on perpetrator demographics or weapons used of all shootings in the United States.

Interventions to improve school climates are necessary at multiple levels, not just at schools, and must acknowledge that US schools are racialized institutions.³³  Interventions include federal and state policies, such as permit-to-purchase and extreme-risk laws, raising the minimum age for gun purchases, secure firearm storage, and background checks^34–36 ; investment into low-income neighborhoods¹⁹ ; prosecuting hate crimes; and holding police accountable for the unnecessary force against and over-policing of BIPOC communities.³⁷  At a societal level, we must root out white supremacy in military and law enforcement institutions³⁷ ; stop pitting different racial and/or ethnic population groups against each other³⁸ ; close opportunity gaps through access to educational resources^18,39 ; and expand mental health services, trauma-informed and positive youth development,⁴⁰  conflict resolution programs, and antiracist curricula, which normalize and center the life experiences of BIPOC communities. Often, we can draw from what already works: nationwide, community initiatives have successfully reduced local violence and addressed the impact of white supremacy.^41,42 

Our knowledge on gun violence is limited even after decades of mass shootings in the United States. Until 2019, research on gun violence was impeded by the Dickey amendment,⁴³  and state laws continue to prohibit research that could help prevent future violence. For example, the Minnesota statute 144.05 blocks the state commissioner of health “from collecting data on individuals regarding lawful firearm ownership in the state or data related to an individual’s right to carry a weapon.”⁴⁴  Given unabated suffering from gun violence in the United States, we are left to wonder for whose benefit such laws are in place.

Our study has limitations. These cross-sectional data allow for examining time trends in weapon carrying but do not allow causal conclusions. Instead, we provided historical context and highlighted unanswered questions and gaps in the literature, which will have to be further examined in future studies. We examined perceptions of safety and violence in school, but other potential confounders (for example socioeconomic status) were unavailable, as were measures of experiences of racism. There is a growing recognition of racism as a priority public health issue.⁴⁵  The YRBS measures describing safety and violence at school (eg, never having missed school in the past 30 days because of safety concerns) do not adequately describe safe schools: many factors can contribute to students feeling unsafe before they miss a day of school. Furthermore, implications of safety and violence measures may vary by race and/or ethnicity: our supplemental analysis revealed that having ever been threatened disguised, on average, worse school experiences for students of color. Future research should probe the types of threats experienced by different racial and/or ethnic groups, which may have important implications for school programs and policies.

This systematic assessment of weapon carrying in US high schools over the past 26 years highlights important areas for future research. The patterns of weapon carrying we found confirmed that unsafe school environments are linked to increased weapon carrying and contradict prevailing racist stereotypes: in schools perceived as safer, NH white boys have been most likely to carry weapons in the past 20 years. We found it disturbing how few reliable data we were able to find to contextualize our findings (eg, aggregated demographic data on gun perpetrators and gun sales). The lack of accessibility to such basic information in the United States where guns are used to kill almost 40 000 individuals each year⁴⁶  demonstrates the vast knowledge gaps left behind by 25 years of suppressed public health research on gun violence. Neither general declines in crime rates nor the GFSA or zero-tolerance approaches have prevented an epidemic of mass shootings and everyday gun violence, in schools and elsewhere. We believe this work offers an important foundation for much needed research to disentangle the intertwined phenomena of racism, toxic environments of violence, and gun- and weapon-culture in the United States.

AA

African American

BIPOC

Black, Indigenous, and People of Color

CI

confidence interval

GFSA

Gun-Free Schools Act

NH

non-Hispanic

YRBS

Youth Risk Behavior Surveillance System