Impact of School Shootings on Adolescent School Safety, 2009–2019

Abstract

Objectives. To examine the impact of school shootings on indicators of adolescent school safety in the United States.

Methods. We linked 2009–2019 Youth Risk Behavior Survey data on 211 236 adolescents aged 14 to 18 years from 24 school districts with data on high school shootings from the Center for Homeland Defense and Security. We conducted 2-way fixed-effects logistic regression models to assess the impact of shootings on self-report of 3 indicators of school safety: avoiding school because of feeling unsafe, carrying a weapon at school, and being threatened or injured with a weapon at school.

Results. High school shootings were associated with adolescents having 20% greater odds of avoiding school because of feeling unsafe (adjusted odd ratio [AOR] = 1.20; 95% confidence interval [CI] = 1.11, 1.29) than those who had not. Findings were slightly attenuated in sensitivity analyses that tested exposure to shootings at any school in the district or state. High school shootings were associated with a statistically nonsignificant (P = .08) elevated risk of carrying a weapon at school (AOR = 1.11; 95% CI = 0.99, 1.25).

Conclusions. The negative ramifications of school shootings extend far beyond the event itself to adolescents’ concerns about school safety. (Am J Public Health. 2023;113(4):438–441. https://doi.org/10.2105/AJPH.2022.307206)

In 2020, firearm-related deaths were the leading cause of mortality for children and adolescents, with more than 3500 youths dying by gun violence.^1,2 School shootings have also increased, peaking in 2018 and 2019, with 75 shootings occurring annually.³ Over recent years, almost one third of states have weakened their gun laws,⁴ which has implications for adolescent health.

Ghiani et al. found that states that adopted stricter gun laws decreased the likelihood of adolescents avoiding school because of safety concerns or of carrying or being threatened by weapons at school.⁵ However, whether exposure to school shootings is associated with these behaviors remains unknown. Using representative samples across 24 school districts, we examined the impact of school shootings on 3 indicators of adolescent school safety: avoiding school because of feeling unsafe, carrying a weapon at school, and being threatened by a weapon at school.

METHODS

We used repeated cross-sectional data from the 2009–2019 Youth Risk Behavior Survey (YRBS) biennial school district surveys, which collect self-reported health-related behaviors on 9th through 12th graders in public and private schools.⁶ Among the 31 districts that participated across study years with at least 2 years of data collection, we excluded 7 because the Centers for Disease Control and Prevention did not have permission to release the data. Our analytic sample included 211 236 adolescents aged 14 to 18 years from 24 districts that collected information on 3 outcomes: (1) number of days not attending school because of feeling unsafe at school or on the way to or from school during the past month, (2) number of days carrying a weapon on school property during the past month, and (3) number of times threatened or injured with a weapon on school property during the past year. Each outcome was dichotomized (0 vs ≥ 1), and analytic samples were based on available responses for each outcome measure.

We obtained school shootings data from the Center for Homeland Defense and Security from April 2007 to March 2019, which we defined as shootings resulting in an injury or death that occurred at any high school in the district.⁷ We linked the shootings to each adolescent based on whether the event occurred between April 1 of the previous survey year to March 31 of the current survey year, as the YRBS is administered biennially in the spring.⁶ We dichotomized school shootings (0 vs ≥ 1) for each survey period.

We first examined the associations between demographic characteristics (age, gender, race/ethnicity) and each outcome using adjusted logistic regression models with year and district fixed effects. Next, we conducted 2-way fixed-effects logistic regression models to assess the impact of school shootings on each indicator of school safety, controlling for demographics with year and district fixed effects. We tested interactions between school shootings and age, gender, and race/ethnicity using the Wald test. Only 1 interaction was jointly significant (P < .05), but none of the stratum-specific estimates were significant (results not shown).

We conducted 2 sensitivity analyses: (1) shootings that occurred at any elementary, middle, or high school in the district; (2) shootings that occurred at any school in the state. We also conducted 2 robustness checks with the main model (results not shown). We first included county-level indicators of percentage Black, percentage Hispanic, and ratio of 90th to 10th percentile of income,⁸ but none of the factors had any effect. Second, we included an indicator of any school shooting in the prior 2 to 4 years (i.e., in the prior wave), but it did not alter the effect size or the significance of exposure to recent shootings.

We conducted analyses using Stata statistical software version 17.0 (StataCorp LP, College Station, TX), with robust standard errors and survey weights to account for the cluster sampling design.⁶

RESULTS

Over the study period, 17 of 24 school districts experienced at least 1 shooting, with a total of 56 shootings across districts (Table A, available as a supplement to the online version of this article at http://www.ajph.org). Overall, 9.2% of adolescents reported avoiding school because of feeling unsafe, 3.5% reported carrying a weapon at school, and 7.3% reported being threatened with a weapon at school. Age was positively associated with each outcome, and males were less likely to avoid school because of feeling unsafe than females but more likely to have weapon exposure (Table B, available as a supplement to the online version of this article at http://www.ajph.org). Black and Hispanic adolescents were more likely to report all outcomes than White adolescents.

We found that being exposed to high school shootings was associated with adolescents having 20% greater odds of avoiding school because of feeling unsafe (adjusted odds ratio [AOR] = 1.20; 95% confidence interval [CI] = 1.11, 1.29) than those who had not (Table 1). Findings were consistent, but slightly attenuated, in sensitivity analyses that tested exposure to shootings at any school in the district or state.

TABLE 1—

Impact of School Shootings on Adolescent School Safety and Weapon Exposure: United States, Youth Risk Behavior Survey, 2009–2019

	Avoiding School Because of Feeling Unsafe, AORa (95% CI)	Carrying Weapon at School, AORa (95% CI)	Being Threatened by Weapon at School, AORa (95% CI)
School shooting at any high school in the district	1.20 (1.11, 1.29)	1.11 (0.99, 1.25)	1.03 (0.95, 1.12)
School shooting at any school in the districtb	1.15 (1.07, 1.23)	1.08 (0.98, 1.21)	1.04 (0.97, 1.13)
Any school shooting in the stateb	1.18 (1.08, 1.29)	1.00 (0.89, 1.13)	0.96 (0.88, 1.04)

Note. AOR = adjusted odds ratio; CI = confidence interval. For more details, see Table A (available as a supplement to the online version of this article at http://www.ajph.org).

^a Model adjusted for age, gender, race/ethnicity, year, and district fixed effects.

^b Includes elementary, middle, and high schools.

We found some evidence that high school shootings were associated with an elevated risk of carrying a weapon at school (AOR = 1.11; 95% CI = 0.99, 1.25), but at a P level of .08 (Table 1). There were no effects of shootings on being threatened with a weapon at school. The alternative specifications were not significant for either outcome.

DISCUSSION

Adolescents exposed to school shootings in their district or state were more likely to avoid at least 1 day of school during the past month because of feeling unsafe compared with adolescents who were not. We found some evidence, albeit marginally significant, that adolescents were more likely to carry weapons at school in response to shootings. Although school safety outcomes varied by age, gender, and race/ethnicity, the effects of shootings did not vary across these characteristics. These results extend previous work⁵ by demonstrating that simply being exposed to school shootings increases adolescents’ school avoidance because of feeling unsafe and, possibly, increases weapon carrying at school.

Using large, representative samples of adolescents linked with government data, our work highlights that the negative ramifications of school shootings extend far beyond the event itself.⁹ These effects are evident for exposure to shootings that occur not only in high schools within the district that adolescents attend school, but in any type of school in their district or state. Because of ever-expanding news coverage and social media, exposure to such events beyond school districts will likely increase.

Despite these strengths, limitations remain. Outcomes were self-reported and subject to reporting bias. Because our analysis included only 24 school districts, the findings may not be generalizable to all high school students. Although the YRBS collects limited demographic information, we found that county-level indicators of race/ethnicity and income were not associated with any outcomes other than individual-level factors. The YRBS also does not release school identifiers, so it was not possible to isolate the effects of attending the actual school where a shooting occurred, or whether adolescents moved into the district after the shooting took place. We could not control for community-based violence, which may affect school safety. Because the YRBS is cross-sectional, we were also not able to follow adolescents over time. The YRBS was administered in school, and students who avoided school because of feeling unsafe may be underrepresented in the data set if they were absent on the day of the survey. This suggests that our findings may underestimate the true association between experiencing a shooting firsthand and adolescent concerns about school safety.

PUBLIC HEALTH IMPLICATIONS

School shootings are disturbingly common, occurring in more than 70% of the included districts. Our results highlight that, with firearm violence on the rise,¹ health care providers need to screen for the educational and psychosocial sequelae of these events. Furthermore, screening should occur for adolescents who attend neighboring schools or adjacent districts with a shooting, in addition to those in the immediate vicinity, as the adverse effects appear to extend well beyond the schools where the shooting occurs. For these districts, a universal, trauma-informed approach to providing psychosocial support to adolescents appears warranted. Our findings also underscore the need for continued advocacy to implement policies that reduce school shootings and gun-related violence, as they have previously been found to reduce adverse outcomes.⁵ Minimizing the number of adolescents affected by these horrific events remains a public health imperative.