Counting Is Not Enough: Student Victimization At and Away From School in National Estimates and a Decade of Peer-Reviewed Research

Counting Is Not Enough: Student Victimization At and Away From School in National Estimates and a Decade of Peer-Reviewed Research

Author: Sherene Thomas

Date: September 17, 2026

Note: This is a working draft prepared for review and revision. It is not formatted for journal submission.

Abstract

This paper presents a secondary analysis of Victimization at and Away From School Among Students Ages 12 to 18, 2014–2024—Statistical Tables (Tinney, 2026), a Bureau of Justice Statistics (BJS) report based on the National Crime Victimization Survey (NCVS), integrated with a thematic synthesis of eight peer-reviewed journal articles published between 2016 and 2026. The BJS report estimates rates of theft and nonfatal violent victimization at and away from school for students ages 12 to 18. In 2024, the total victimization rate at school (27.0 per 1,000 students) exceeded the rate away from school (17.7 per 1,000), though neither differed significantly from 2014 (Tinney, 2026). Theft victimization declined substantially at school (14.1 to 5.2 per 1,000) and away from school (11.2 to 5.1 per 1,000) between 2014 and 2024, while nonfatal violent victimization at school showed no significant change over the decade but rose from 14.2 per 1,000 in 2023 to 21.8 per 1,000 in 2024 (Tinney, 2026). No significant male–female differences in at-school victimization were observed in 2024, but students ages 12 to 14 experienced higher at-school theft rates than students ages 15 to 18 (Tinney, 2026). The peer-reviewed literature converges with the BJS report on younger adolescents' heightened vulnerability at school (Vidourek & King, 2019; Webb et al., 2026) but qualifies it in three respects: criminal victimization and bullying follow distinct demographic patterns (Pontes et al., 2018; Webb et al., 2021); fear of victimization diverges demographically from victimization itself, with Black and Hispanic adolescents reporting disproportionate fear (Kurpiel et al., 2023); and victimization's consequences extend into attendance-mediated academic harms (Gardella et al., 2016), unsafe-feeling absenteeism (Cheung et al., 2023), and widespread feelings of being unsafe at school (Mori et al., 2021). Measurement cautions — including the noncomparability of 2016 youth estimates and the 2024 split-sample instrument redesign — are discussed (Tinney, 2026).

Keywords: school victimization, adolescent victimization, National Crime Victimization Survey, theft, violent victimization, fear of crime, school avoidance

Introduction

Schools are among the most consequential settings in adolescents' lives, structuring peer contact, adult supervision, and daily routine for the majority of waking hours during the school year. That same concentration of young people also makes schools a distinctive context for criminal victimization: incidents that occur at school implicate not only individual victims but also the institutions charged with their safety (Vidourek & King, 2019). Understanding how much victimization students experience at school relative to away from school — and how those experiences distribute by crime type, age, and sex — is therefore a core task for both criminology and education policy.

The primary federal source for such estimates is the National Crime Victimization Survey (NCVS), the nation's principal self-report survey of criminal victimization, administered annually to a nationally representative sample of U.S. households with persons ages 12 and older interviewed about crimes experienced during the prior six months (Tinney, 2026). On September 16, 2026, the Bureau of Justice Statistics (BJS) released Victimization at and Away From School Among Students Ages 12 to 18, 2014–2024—Statistical Tables (Tinney, 2026; NCJ 311950), which details rates of theft and nonfatal violent victimization at and away from school for students ages 12 to 18 across an eleven-year span, disaggregated by location of victimization, type of crime, age, sex, and other demographic characteristics. The report's headline findings are striking in combination: in 2024, students were victimized at higher rates at school than away from school; theft victimization fell by more than half over the decade in both locations; violent victimization at school was statistically unchanged from 2014 to 2024 yet rose significantly between 2023 and 2024; and male and female students did not differ significantly in at-school victimization rates in 2024 (Tinney, 2026). These patterns invite contextualization: what does the peer-reviewed literature, produced with different instruments and constructs, say about the same phenomena?

This paper pursues that contextualization through two linked aims. First, it provides a secondary analysis of the BJS report's published estimates, organizing them around five research questions concerning violent victimization by location, theft victimization by location, age and sex risk patterns, fear and avoidance correlates, and measurement considerations. Second, it synthesizes eight peer-reviewed journal articles published from 2016 to 2026 that bear on those questions, drawn from criminology, public health, community psychology, nursing, and adolescent-health journals. The synthesis is deliberately disciplined about constructs: NCVS incident rates (victimizations per 1,000 students), School Crime Supplement (SCS) prevalence estimates (percentage of students), and Youth Risk Behavior Survey (YRBS) bullying prevalence are related but not interchangeable measures, and the paper treats them as such (Tinney, 2026; Vidourek & King, 2019).

The five research questions guiding the analysis are:

1. How do rates of nonfatal violent victimization among students ages 12 to 18 compare at school versus away from school, and how have they changed from 2014 to 2024?

2. How do rates of theft victimization among students ages 12 to 18 compare at school versus away from school, and how have they changed from 2014 to 2024?

3. How do victimization rates at and away from school vary by age and sex?

4. What does peer-reviewed research reveal about fear of victimization, avoidance, and school avoidance as correlates and consequences of student victimization?

5. What methodological considerations — instrument design, self-report, comparability breaks — condition the interpretation of NCVS-based youth victimization estimates?

Literature Review

Violent Victimization At Versus Away From School

The BJS anchor report defines nonfatal violent victimization as rape or sexual assault, robbery, aggravated assault, and simple assault, and defines "at school" to include the school building, school property, the school bus, and travel to and from school (Tinney, 2026). Because the NCVS interviews household members ages 12 and older about all criminal victimizations regardless of location, it is one of few instruments capable of placing school and nonschool violent victimization on a common metric (Tinney, 2026). Peer-reviewed research rarely replicates that at-versus-away disaggregation directly; instead, it typically studies violence at school in depth, using school-based surveys.

Webb et al. (2026) examined interpersonal violence among 21,892 U.S. high school students in grades 9 through 11 using the 2017 and 2019 YRBS. Students who were young for their grade — that is, younger than typical for their grade level — had elevated odds of physical fighting generally, of being threatened or injured with a weapon specifically at school, and of both school and electronic bullying victimization, with the specific pattern of associations varying by sex (Webb et al., 2026). The study is notable for the anchor report's purposes because it isolates school-located violent experiences (weapon threats or injuries at school) and shows that developmental timing relative to peers, not only chronological age, structures risk. Its age bands (grades 9–11, roughly ages 14–17) overlap the upper portion of the BJS report's 12-to-18 range, offering a complementary lens on the older adolescents among whom the BJS report finds lower at-school theft rates (Tinney, 2026; Webb et al., 2026).

Kurpiel et al. (2023) approached school violence through the NCVS's companion instrument, the School Crime Supplement, pooling 49,782 surveys from 35,588 adolescents between 1999 and 2017 to study trends in violent victimization and fear at school. Their central finding concerns the fear–victimization gap: Black and Hispanic adolescents had higher odds of fearing victimization at school even after adjusting for violent victimization itself, and school harms, exposure to violence, and school security and discipline measures explained substantial portions of that racial and ethnic gap (Kurpiel et al., 2023). For interpreting the BJS report, the implication is that victimization rates and the experience of safety at school can move on different tracks; a flat or declining victimization trend does not guarantee a flat fear trend, particularly for demographic groups whose school environments carry heavier security and disciplinary footprints (Kurpiel et al., 2023; Tinney, 2026).

Theft Victimization Among Adolescents

Theft — defined in the BJS report as attempted and completed purse snatching, completed pocket picking, and all attempted and completed thefts except motor vehicle theft — is the only property crime the anchor report tracks, and it is the category in which the clearest decade-long improvement appears: the at-school theft rate fell from 14.1 to 5.2 per 1,000 students between 2014 and 2024, and the away-from-school rate from 11.2 to 5.1 per 1,000 (Tinney, 2026). Notably, the systematic search underlying this paper located no peer-reviewed journal article published between 2016 and 2026 that estimates theft victimization specifically among 12- to 18-year-olds at versus away from school. Peer-reviewed youth-victimization research concentrates heavily on violence, bullying, and peer aggression; property victimization among adolescents is comparatively neglected as a standalone object of journal-based inquiry. This is a genuine gap in the literature rather than a search failure, and it means the BJS statistical series stands essentially alone as the source for student theft-victimization trends. The absence is itself analytically informative: theories and interventions organized around school safety emphasize interpersonal harm, while the most common criminal victimization students experience at school — theft — receives little dedicated theoretical attention in the journals surveyed.

Bullying, Peer Victimization, and Physical Attacks: Age and Sex Patterns

If theft is undertheorized, bullying and peer aggression are among the most studied adolescent-victimization phenomena, and they provide essential context for the BJS report's age and sex findings — provided the constructs are not conflated. Bullying, as measured in the YRBS and SCS, captures repeated or power-imbalanced peer aggression; NCVS violent victimization captures criminal incidents (assault, robbery, sexual assault) regardless of repetition or relationship (Tinney, 2026; Vidourek & King, 2019). The two overlap empirically but behave differently across demographic groups.

Pontes et al. (2018) estimated linear time trends by gender in school and electronic bullying victimization across four YRBS waves (N = 61,042). School bullying prevalence increased significantly among females, from 21.2% in 2009 to 24.8% in 2015 (trend odds ratio = 1.08, 95% CI [1.04, 1.12]), while decreasing significantly among males, from 18.7% to 15.8% (OR = 0.93, 95% CI [0.89, 0.98]); electronic bullying prevalence was unchanged between 2011 and 2015 for both sexes (Pontes et al., 2018). Asian, Black, and Hispanic students reported significantly lower rates of both school and electronic bullying than White students (Pontes et al., 2018). These divergent sex-specific trends stand in instructive contrast to the BJS anchor's 2024 result of no significant male–female differences in criminal victimization at school across all crime types (Tinney, 2026): the contrast does not contradict the BJS finding but rather demonstrates that conclusions about "sex differences in school victimization" are construct-dependent.

Webb et al. (2021) extended the demographic analysis of bullying using 2015 and 2017 YRBS data (n = 15,624 and 14,765), finding that approximately 20% of high school students reported school bullying victimization and 15% electronic bullying victimization. Sexual minority youth were significantly more likely than heterosexual peers to report both forms, whereas Black and Latinx students were significantly less likely to report both; few changes occurred between 2015 and 2017 apart from a significant decrease in school bullying among White heterosexual youth (Webb et al., 2021). The study's relevance to the anchor report lies in its qualification of the BJS finding that total victimization at and away from school did not differ significantly by student race or Hispanic origin in 2024 (Tinney, 2026): when victimization is operationalized as bullying rather than criminal incidents, and when sexual identity is measured, demographic risk profiles look markedly different (Webb et al., 2021).

Vidourek and King (2019) provided the most directly NCVS-adjacent peer-reviewed evidence on school-based victimization risk factors, conducting a secondary analysis of the 2015 NCVS School Crime Supplement with 4,767 student completers ages 12 to 18. Peer victimization — operationally defined as having been threatened with harm or pushed, shoved, tripped, or spit on by another student at school — was reported by 7.2% of students (Vidourek & King, 2019). With the exception of sex, which was not significant (OR = 1.091, p = .595), all individual factors examined were associated with victimization: middle school (vs. high school) status, receiving Cs, Ds, or Fs, recent skipping of class, avoiding school or school activities (OR = 14.442), and classroom distraction elevated victimization odds, as did attending a school enrolling fewer than 1,000 students (OR = 1.693), the presence of gangs in school (OR = 3.822), and low peer connectedness (OR = 2.582) (Vidourek & King, 2019). In the final logistic regression model, grade level, grades received, classroom distraction, enrollment size, and gangs in school remained significant predictors (Vidourek & King, 2019). Two features of these results speak directly to the BJS report. First, the middle-school elevation in peer victimization parallels the anchor's finding that younger students (ages 12–14) face higher at-school theft rates than older students (ages 15–18), suggesting that early adolescence is a period of generalized vulnerability in school settings across distinct victimization constructs (Tinney, 2026; Vidourek & King, 2019). Second, the null finding for sex in peer victimization converges with the anchor's null finding for sex in 2024 at-school criminal victimization — a rare point of cross-construct agreement (Tinney, 2026; Vidourek & King, 2019).

Measurement: NCVS- and SCS-Based Research on Youth Victimization

Interpreting the BJS report requires understanding the instruments behind it, and three of the reviewed studies are themselves built on NCVS or SCS data, making their methodological discussions directly relevant. The SCS is a supplement to the NCVS, developed jointly by BJS, the National Center for Education Statistics, and the U.S. Census Bureau; eligible NCVS household members ages 12 to 18 who are enrolled in school are interviewed about school-related victimization, with the first interview conducted face to face and subsequent interviews by computer-assisted telephone (Vidourek & King, 2019). Vidourek and King (2019) detailed both the strengths of this design — national probability sampling across all 50 states — and its limits: all data are self-reported and thus subject to social-desirability pressures; the SCS provides respondents with key definitions that may lead to underreporting of some behaviors; and the cross-sectional design precludes causal inference. These cautions apply with equal force to the anchor report's NCVS estimates (Tinney, 2026; Vidourek & King, 2019).

Kurpiel et al. (2023) demonstrated the analytic power of pooling SCS cross-sections, combining 49,782 surveys from 35,588 adolescents across 1999–2017 to study trends and the fear–victimization gap. Their work illustrates that SCS-based research can track school victimization and its subjective correlates over decades — a design logic parallel to the anchor report's 2014–2024 NCVS trend figures (Kurpiel et al., 2023; Tinney, 2026). Gardella et al. (2016) used the 2011 SCS (n = 5,930 adolescents ages 12–18) to model repeated victimization, showing how SCS items can be combined to capture multiple victimization — victimization "on a regular basis" — rather than single incidents, a conceptualization closer to lived experience than incident counts (Gardella et al., 2016).

Three measurement issues flagged in the anchor report deserve emphasis because they condition every trend comparison. First, due to the methodology used to create the 2016 revised NCVS data, victimization estimates among youth in 2016 are not comparable to estimates for other years; the report omits or flags 2016 accordingly (Tinney, 2026). Second, in 2024 the NCVS transitioned to a redesigned instrument administered via a split sample — half of respondents received the legacy instrument and half the redesigned one — and the anchor report's 2024 estimates use the legacy-instrument data; splitting the sample reduces estimate precision during the transition (Tinney, 2026). Third, NCVS estimates count victimizations (incidents) per 1,000 students, whereas SCS and YRBS studies typically report the percentage of students victimized; incident rates and prevalence are not interchangeable, and a single student may contribute multiple incidents to a rate (Tinney, 2026; Vidourek & King, 2019).

Fear, Avoidance, and School Avoidance

The BJS report tabulates incidents; the peer-reviewed literature documents what incidents do — to students' sense of safety, attendance, and learning. Cheung et al. (2023) analyzed 2019 YRBS data from a nationally representative sample of 9,507 high school students and found that 7.9% had missed school because they felt unsafe at school or traveling to or from school. Students who had experienced physical or sexual dating violence had 3.69 times the adjusted odds of such unsafe-feeling absenteeism relative to non-victims, and the association was stronger for males than for females (Cheung et al., 2023). The study is a reminder that victimization occurring outside a dating relationship's school boundaries — dating violence need not occur at school — can nevertheless produce school avoidance, blurring the at-versus-away distinction that organizes the BJS tables (Cheung et al., 2023; Tinney, 2026).

Mori et al. (2021) synthesized 43 quantitative studies of feeling unsafe at school among children and adolescents, finding a mean prevalence of 19.4% with wide variation across studies — variation the authors attributed in part to inconsistent measurement. Feeling unsafe was associated with victimization experiences as well as with depressive symptoms and suicidal behavior (Mori et al., 2021). The review establishes that the subjective experience of unsafety is both common — roughly one in five students on average — and clinically consequential, even though it is not itself a crime rate (Mori et al., 2021).

Gardella et al. (2016) connected victimization to concrete academic consequences in their SCS-based path models: among 5,930 adolescents ages 12–18, multiple victimization predicted lower academic achievement and poorer attendance, absenteeism partially mediated the victimization–achievement relationship, and the presence of metal detectors and security guards moderated the link between multiple victimization and absenteeism (Gardella et al., 2016). The moderation finding is policy-relevant: school security measures, which are invisible in the BJS rate tables, appear to condition whether victimization translates into staying away from school (Gardella et al., 2016; Tinney, 2026). Complementing this, Vidourek and King (2019) found that students who avoided school or school activities had 14.442 times the odds of reporting peer victimization — an association that likely runs in both directions, with avoidance functioning as both consequence of and risk factor for victimization (Vidourek & King, 2019).

Methodology

Design

This paper employs a secondary-analysis design integrating two evidence streams: (a) published statistical estimates from a federal statistical report, and (b) a thematic synthesis of peer-reviewed journal literature. No primary data were collected, no human subjects were involved, and no statistical reanalysis of microdata was performed; all quantitative claims derive from the published tables, figures, highlights, and appendix tables of the anchor report or from the published results of the reviewed articles.

The Anchor Report

The anchor is Tinney (2026), Victimization at and Away From School Among Students Ages 12 to 18, 2014–2024—Statistical Tables (NCJ 311950), released by the Bureau of Justice Statistics in September 2026 and based on the National Crime Victimization Survey. The report presents victimization rates per 1,000 students ages 12 to 18 — defined as NCVS respondents ages 12 to 18 whose educational attainment did not exceed grade 12 — for total victimization (theft plus nonfatal violent victimization), theft, nonfatal violent victimization, and nonfatal violent victimization excluding simple assault, each at and away from school, for 2014 through 2024, with demographic breakdowns for 2024 (Tinney, 2026). Statistical significance in the report is assessed at both the 90% and 95% confidence levels, and all comparisons described in its text are statistically significant unless otherwise stated (Tinney, 2026). The Findings section below reports the anchor's estimates as published; where the report flags estimates as unstable (based on 10 or fewer sample cases or a coefficient of variation above 50%), that qualification is noted.

Literature Search and Inclusion Criteria

The peer-reviewed literature was identified through searches of PubMed, publisher journal pages, and scholarly search engines conducted on September 17, 2026. Inclusion criteria were: (a) publication in a peer-reviewed journal; (b) publication year 2016–2026 inclusive, using the journal issue year; (c) substantive relevance to at least one of the five research questions — violent victimization at versus away from school, theft victimization among youth ages 12–18, school victimization/bullying/physical attacks with age or sex risk factors, NCVS-based or methodological studies of youth victimization measurement, or fear, avoidance, and school avoidance; and (d) verifiable citation metadata (authors, year, title, journal, volume, pages, DOI) confirmed against PubMed records, publisher pages, or institutional faculty CVs. Eight articles met all criteria. Each article's abstract — and, where openly available, full text — was read and its findings recorded in dated research notes before synthesis. Government statistical reports were consulted only as supplementary context and are not counted among the eight.

Analytic Approach

Findings from the anchor report were extracted by research question and organized into trend comparisons (2014–2024, 2023–2024), location comparisons (at versus away from school), and demographic comparisons (age, sex, and secondarily race/Hispanic origin, household income, and location of residence). The peer-reviewed articles were synthesized thematically following the five literature-review themes above, with explicit attention to construct comparability: claims are not made that bullying prevalence equals violent-victimization rates, or that SCS prevalence equals NCVS incident rates. Points of convergence and divergence between the anchor estimates and the peer-reviewed findings are identified in the Discussion. Consistent with the evidentiary standard adopted for this draft, every factual claim is anchored to an in-text citation to either the BJS report or a verified peer-reviewed source, and no statistic, quotation, or finding has been fabricated; details that could not be verified from the sources at hand are omitted rather than estimated.

Findings

RQ1: Nonfatal Violent Victimization At Versus Away From School, 2014–2024

In 2024, the rate of nonfatal violent victimization at school stood at 21.8 per 1,000 students ages 12 to 18, compared with 12.6 per 1,000 away from school (Tinney, 2026). Two features of the violent-victimization trend are notable. First, over the full 2014–2024 span, there was no statistically significant difference in the at-school violent victimization rate between 2014 and 2024, and the away-from-school rate likewise did not change significantly (Tinney, 2026). Violent victimization excluding simple assault — that is, rape or sexual assault, robbery, and aggravated assault — was similarly stable: 5.8 per 1,000 at school and 3.6 per 1,000 away from school in 2024, neither significantly changed from 2014, with no significant 2023–2024 changes at either location (Tinney, 2026). Second, against that decade-long stability, the at-school violent victimization rate rose significantly between 2023 (14.2 per 1,000) and 2024 (21.8 per 1,000) — a year-over-year increase of roughly one-half that the report flags as statistically significant even as the decade comparison remains null (Tinney, 2026). Because total victimization is the sum of theft and violent victimization, and theft fell while violence was flat-to-rising, violence accounts for a growing share of the at-school total: in 2024, violent victimization (21.8) comprised the large majority of total at-school victimization (27.0 per 1,000) (Tinney, 2026).

The peer-reviewed literature does not directly replicate the at-versus-away violent-victimization comparison, but it contextualizes the school side of the ledger. Webb et al. (2026) documented that being young for one's grade elevates the odds of weapon-related threats or injuries specifically at school among high school students, identifying a developmental-risk mechanism consistent with elevated school-located violence among younger adolescents (Webb et al., 2026). Kurpiel et al. (2023), pooling SCS data from 1999 to 2017, situated violent victimization at school within a longer downward trend while showing that fear of victimization follows its own trajectory — a decoupling relevant to the anchor's 2023–2024 uptick, whose experiential meaning for students' sense of safety cannot be read off the rate alone (Kurpiel et al., 2023; Tinney, 2026).

RQ2: Theft Victimization At Versus Away From School, 2014–2024

Theft is the anchor report's clearest success story and its most construct-isolated finding. The at-school theft victimization rate decreased from 14.1 per 1,000 students in 2014 to 5.2 per 1,000 in 2024, and the away-from-school rate decreased from 11.2 to 5.1 per 1,000 over the same span; neither location showed a significant change between 2023 and 2024 (Tinney, 2026). By 2024, then, theft rates at and away from school had essentially converged at approximately 5 per 1,000 — a striking contrast with violent victimization, where the at-school rate remained nearly double the away-from-school rate (Tinney, 2026). In other words, the school setting in 2024 was disproportionately a violent-victimization setting relative to elsewhere, not a theft setting.

No peer-reviewed journal article from 2016–2026 was located that estimates student theft victimization at versus away from school, so this finding rests on the BJS series alone. Its interpretation is correspondingly constrained: the literature offers no tested account of why student theft fell by more than 60% in both locations over the decade. What can be said descriptively is that the decline predates and spans the 2020–2021 period in which total at-school victimization reached its series lows (11.2 per 1,000 in 2020; 6.8 in 2021), indicating that the theft decline is not an artifact of a single anomalous year (Tinney, 2026).

RQ3: Age and Sex Patterns in Victimization At and Away From School

Age and sex pattern differently across crime types and locations in the anchor report's 2024 estimates. At school, there were no statistically significant differences between male and female students in victimization rates across all types of crime — total (24.3 per 1,000 for males; 29.9 for females), theft (4.0; 6.4), and nonfatal violent victimization (20.3; 23.5) (Tinney, 2026). Away from school, by contrast, the total victimization rate was significantly higher for female students (23.5 per 1,000) than for male students (12.2 per 1,000) (Tinney, 2026). The sex story is thus one of convergence at school and divergence away from school in 2024.

Age gradients are crime-specific. Total victimization in 2024 did not differ significantly between students ages 12–14 and 15–18 either at school (31.9 vs. 22.4 per 1,000) or away from school (14.1 vs. 21.0) (Tinney, 2026). For theft at school, however, the younger group faced a significantly higher rate — 7.5 per 1,000 for ages 12–14 versus 3.0 per 1,000 for ages 15–18 — while away-from-school theft did not differ by age (5.1 per 1,000 for both groups) (Tinney, 2026). The peer-reviewed literature converges on younger adolescents' school vulnerability: Vidourek and King (2019) found middle school students at elevated odds of peer victimization relative to high school students in 2015 SCS data, and Webb et al. (2026) found young-for-grade high school students at elevated risk of school-located violence and bullying (Vidourek & King, 2019; Webb et al., 2026). The literature qualifies the anchor's sex nulls, however: Pontes et al. (2018) documented significantly divergent bullying trends by sex (rising among females, falling among males, 2009–2015), and Cheung et al. (2023) found dating-violence–related unsafe-feeling absenteeism more strongly associated for males — demonstrating that sex patterns in adjacent constructs do not mirror the NCVS criminal-victimization null (Cheung et al., 2023; Pontes et al., 2018; Tinney, 2026). Beyond age and sex, the anchor report found no significant differences by race or Hispanic origin in total victimization at or away from school in 2024, but significantly higher at-school total victimization for students in households earning less than $25,000 per year (60.3 per 1,000) than for those in $25,000–$49,999 households (26.7 per 1,000), and higher at-school rates in suburban (31.5) than rural (11.0) areas (Tinney, 2026).

RQ4: Fear, Avoidance, and School Avoidance

Four peer-reviewed sources address what victimization and perceived safety do to students' school engagement. Quantitatively, the most direct estimate comes from Cheung et al. (2023): 7.9% of U.S. high school students in 2019 missed school because they felt unsafe at school or in transit, and victims of physical or sexual dating violence had 3.69 times the adjusted odds of such absenteeism — evidence that feeling unsafe is a measurable, nontrivial driver of lost school time (Cheung et al., 2023). Mori et al. (2021), synthesizing 43 studies, placed the mean prevalence of feeling unsafe at school at 19.4%, while emphasizing wide cross-study variation attributable to inconsistent measurement — a caution that applies to any comparison of "fear" across instruments (Mori et al., 2021).

The mechanisms linking victimization to avoidance are elaborated in two SCS-based studies. Gardella et al. (2016) showed in path models (n = 5,930, ages 12–18) that absenteeism partially mediated the relationship between multiple victimization and academic performance, and that metal detectors and security guards moderated the victimization–absenteeism link — meaning the translation of victimization into missed school depends in part on the school's security context (Gardella et al., 2016). Vidourek and King (2019) reported the striking univariate association between avoiding school or school activities and peer victimization (OR = 14.442), while noting the association's bidirectional plausibility (Vidourek & King, 2019). Kurpiel et al. (2023) added the demographic dimension: fear at school is disproportionately borne by Black and Hispanic adolescents net of violent victimization, implicating school harms, violence exposure, and security/discipline regimes — the very measures Gardella et al. (2016) found to moderate victimization's consequences — as structural contributors to the fear gap (Gardella et al., 2016; Kurpiel et al., 2023). Taken together, these studies indicate that the BJS report's incident rates understate victimization's total burden on schooling, which includes fear, avoidance, and attendance-mediated learning losses distributed unevenly across student groups.

RQ5: Methodological Considerations in NCVS-Based Youth Estimates

Three methodological features of the anchor report condition its interpretation, each corroborated or elaborated by the peer-reviewed literature. First, the 2016 break: because of the methodology used to create the 2016 revised NCVS data, youth victimization estimates for 2016 are not comparable to other years, and the report excludes 2016 from trend comparisons (Tinney, 2026). Any decade trend statement — including the theft decline and the violent-victimization stability reported above — is therefore a 2014-to-2024 comparison with a documented hole at its midpoint. Second, the 2024 instrument transition: the NCVS moved to a redesigned instrument administered via split sample in 2024, with half of respondents receiving the legacy instrument; the anchor report uses the legacy-instrument data, and the split design reduces estimate precision during the transition (Tinney, 2026). The significant 2023–2024 increase in at-school violent victimization should be read with that precision caveat in mind. Third, the self-report and construct caveats: NCVS estimates derive from household respondents' self-reports of crimes in the prior six months (excluding the interview month), count incidents rather than victims, and rest on respondents' understanding of crime categories (Tinney, 2026). Vidourek and King (2019) noted parallel limitations for the SCS — social-desirability pressures, definition-driven underreporting, and cross-sectional designs that preclude causal claims — and Mori et al. (2021) documented how inconsistent operationalization of "feeling unsafe" produces wide prevalence variation across studies (Mori et al., 2021; Tinney, 2026; Vidourek & King, 2019). A final comparability point: the anchor report's rates (victimizations per 1,000 students) cannot be directly compared with the reviewed studies' prevalence figures (percent of students victimized), because one student may contribute multiple incidents to a rate (Tinney, 2026; Vidourek & King, 2019).

Discussion

How the Literature Aligns With the BJS Report

The most robust alignment concerns early adolescence as a period of school-located vulnerability. The anchor report's significantly higher at-school theft rate for ages 12–14 than 15–18 in 2024 converges with Vidourek and King's (2019) finding of elevated peer victimization among middle school students and Webb et al.'s (2026) finding of elevated school violence and bullying among young-for-grade students (Tinney, 2026; Vidourek & King, 2019; Webb et al., 2026). That three distinct constructs — NCVS theft incidents, SCS peer victimization, YRBS school violence — all point toward younger adolescents suggests a developmental regularity rather than a measurement artifact: the transition into secondary schooling appears to concentrate multiple forms of school-based harm. A second alignment is the null sex finding for school victimization specifically: the anchor's 2024 result of no male–female differences at school across crime types echoes Vidourek and King's (2019) nonsignificant sex effect for peer victimization (OR = 1.091, p = .595) (Tinney, 2026; Vidourek & King, 2019).

How the Literature Qualifies the BJS Report

Three qualifications are substantive. First, construct specificity: the anchor's sex and race nulls for criminal victimization at school do not generalize to bullying. Pontes et al. (2018) showed school bullying moving in opposite directions by sex between 2009 and 2015, and Webb et al. (2021) showed bullying stratifying by sexual identity and race/ethnicity in ways the NCVS crime categories do not capture (Pontes et al., 2018; Tinney, 2026; Webb et al., 2021). Policy discourse that treats "school victimization" as a single phenomenon risks misreading these divergences.

Second, the fear–victimization decoupling: Kurpiel et al. (2023) demonstrated that Black and Hispanic adolescents' fear of school victimization exceeds what their violent-victimization rates predict, with school security and discipline regimes explaining part of the excess. This matters for interpreting the anchor report because the same security measures that may suppress incident rates can simultaneously elevate fear among the students most exposed to them — a dynamic consistent with Gardella et al.'s (2016) finding that security measures moderate how victimization converts into absenteeism (Gardella et al., 2016; Kurpiel et al., 2023; Tinney, 2026). Interventions evaluated solely on incident rates may therefore miss — or worsen — the experiential outcomes students actually carry.

Third, the burden beyond incidents: the anchor report counts victimizations, but the literature shows victimization also produces avoidance and learning loss. With 7.9% of students missing school over safety concerns (Cheung et al., 2023), absenteeism mediating victimization's academic harms (Gardella et al., 2016), and roughly one in five students feeling unsafe at school on average (Mori et al., 2021), the true cost of school victimization substantially exceeds its incident rate. The significant 2023–2024 rise in at-school violent victimization (14.2 to 21.8 per 1,000) is thus concerning not only as a rate change but as a potential precursor to downstream avoidance and achievement effects documented in the SCS literature (Gardella et al., 2016; Tinney, 2026).

Implications

For practice, the synthesis points toward developmentally timed prevention: because multiple victimization constructs concentrate risk in early adolescence and middle school, prevention resources may yield the greatest return when directed at school transitions rather than spread uniformly across grades (Tinney, 2026; Vidourek & King, 2019; Webb et al., 2026). For measurement, the findings argue for routinely pairing incident-rate surveillance with fear and avoidance indicators — as the SCS already does — so that trend reports can distinguish schools that are safer from schools that merely record fewer incidents (Kurpiel et al., 2023; Tinney, 2026). For research, the theft gap is the clearest opportunity: the decade's largest documented improvement in student victimization has essentially no peer-reviewed explanatory literature, leaving open questions about whether the decline reflects changes in students' property, supervision, reporting, or the environments themselves (Tinney, 2026).

Limitations

This paper's limitations begin with its evidentiary base. As a secondary analysis, it inherits all limitations of the anchor report: self-reported data subject to recall and social-desirability biases; the noncomparability of 2016 youth estimates; reduced precision in 2024 from the split-sample instrument transition; and incident-based rates that cannot identify repeat victims (Tinney, 2026; Vidourek & King, 2019). The peer-reviewed synthesis is limited to eight articles meeting strict verification standards; relevant studies in books, dissertations, or non-indexed journals were excluded, and the 2016–2026 window excludes foundational earlier work. Construct heterogeneity across the eight studies — YRBS bullying, SCS peer victimization and fear, NCVS incidents — enriches interpretation but precludes quantitative pooling; all cross-study comparisons here are qualitative. Finally, the anchor report's detailed appendix tables (standard errors, counts, subgroup estimates) were consulted for headline figures but not exhaustively extracted; fine-grained subgroup claims should be verified against the published PDF before further use.

Conclusion

The decade from 2014 to 2024 transformed the composition of student victimization more than its overall level. Total victimization at school in 2024 (27.0 per 1,000) was statistically indistinguishable from 2014, yet beneath that stability, theft victimization collapsed by more than three-fifths in both locations while violent victimization held steady across the decade and jumped significantly between 2023 and 2024 (Tinney, 2026). School remained, in 2024, the higher-rate location overall — driven by violence, not theft — with younger adolescents bearing disproportionate theft risk at school, no sex differences in at-school criminal victimization, and pronounced income gradients (Tinney, 2026). The peer-reviewed literature affirms younger adolescents' school vulnerability and the at-school sex null while insisting on three refinements: bullying and criminal victimization follow different demographic scripts (Pontes et al., 2018; Webb et al., 2021); fear obeys a different demographic logic than victimization, to Black and Hispanic students' disadvantage (Kurpiel et al., 2023); and victimization's full burden includes fear, avoidance, and attendance-mediated academic harm (Cheung et al., 2023; Gardella et al., 2016; Mori et al., 2021). Together, the BJS estimates and the journal literature describe a school-victimization landscape in which counting incidents is necessary but insufficient — and in which the most improved crime category, theft, remains the least understood.

References

Cheung, W. W., Caduff, A., & Raj, A. (2023). The association between dating violence and student absenteeism among a representative sample of U.S. high school students: Findings from the 2019 Youth Risk Behavior Survey. Journal of Interpersonal Violence, 38(1–2), NP2218–NP2233. https://doi.org/10.1177/08862605221090564

Gardella, J. H., Tanner-Smith, E. E., & Fisher, B. W. (2016). Academic consequences of multiple victimization and the role of school security measures. American Journal of Community Psychology, 58(1–2), 36–46. https://doi.org/10.1002/ajcp.12075

Kurpiel, A., Hullenaar, K. L., & Ruback, R. B. (2023). Racial and ethnic differences in the fear-victimization gap at school: An examination of school context and trends over time. Journal of Interpersonal Violence, 38(3–4), 2534–2565. https://doi.org/10.1177/08862605221101196

Mori, Y., Tiiri, E., Khanal, P., Khakurel, J., Mishina, K., & Sourander, A. (2021). Feeling unsafe at school and associated mental health difficulties among children and adolescents: A systematic review. Children, 8(3), 232. https://doi.org/10.3390/children8030232

Pontes, N. M. H., Ayres, C. G., Lewandowski, C., & Pontes, M. C. F. (2018). Trends in bullying victimization by gender among U.S. high school students. Research in Nursing & Health, 41(3), 243–251. https://doi.org/10.1002/nur.21868

Tinney, E. (2026). Victimization at and away from school among students ages 12 to 18, 2014–2024—Statistical tables (NCJ 311950). Bureau of Justice Statistics, U.S. Department of Justice. https://bjs.ojp.gov/document/vaafs1424.pdf

Vidourek, R. A., & King, K. A. (2019). Risk factors for peer victimization among middle and high school students. Children, 6(1), 11. https://doi.org/10.3390/children6010011

Webb, L., Clary, L. K., Johnson, R. M., & Mendelson, T. (2021). Electronic and school bullying victimization by race/ethnicity and sexual minority status in a nationally representative adolescent sample. Journal of Adolescent Health, 68(2), 378–384. https://doi.org/10.1016/j.jadohealth.2020.05.042

Webb, L., Ma, M., Clary, L. K., Brooks-Russell, A., & Johnson, R. M. (2026). Associations between age-for-grade, sex, and interpersonal violence among U.S. high school students. Journal of Interpersonal Violence, 41(3–4), 497–517. https://doi.org/10.1177/08862605241311615