School crime and school shootings: Understanding the numbers

On Tuesday May 31, the Bureau of Justice Statistics (BJS) and the National Center for Education Statistics (NCES) published their latest report in a long-standing series on school crime and safety. The recent report focuses on 2020 youth victimizations rates at school and away from school. The findings weren’t incredibly notewothy, though there is one point worth noting: they found that the at-school victimization rate declined about 60% from 2019 to 2020, but the out-of-school victimization rate remained fairly stable.

This report builds from the more comprehensive findings released in a July 2021 report, which provided information on the prevalence of school shootings, which has been a primary concern in recent months and especially so in the wake of the recent elementary school shooting in Uvalde, Texas. The data from these reports was collected from the National Crime Victimization Survey (NCVS) and the School Crime Supplement. In July 2022, the BJS will release a more comprehensive publication (more akin to the July 2021 report) with updated data for 2021.

The numbers regarding school shootings are very important to understand, but unfortunately are misleading in terms of how they are presented in the BJS/NCES reports. To better understand this issue, I took a deeper dive into the raw data from Homeland Security’s K-12 School Shooting Database along with two other sources: the Violence Project’s database on mass shooters and the Washington Post’s database on school shootings.

As stated above, the latest BJS report compared youth at-school and out-of-school victimization rates for 2019 and 2020. They found that, from 2019 to 2020, the at-school victimization rate declined about 60%, from a rate of 30 per 1,000 to a rate of 11 per 1,000. The decline from 2019 to 2020 is likely attributable, at least in part, to COVID-19 school closures that reduced in-person contact at schools. Though, the at-school victimization rate has actually been declining since before the pandemic, with 50 victimizations per 1,000 recorded in 2009. When looking back to 2009, the out-of-school victimization rate also declined before stabilizing from 2019 to 2020. In 2009, the out-of-school victimization rate was 33 per 1,000, which decreased to a rate of 20 per 1,000 by 2019. But in 2020, the rate was not statistically significantly different from 2019, at 15 per 1,000. Regarding the overall decline since 2009, it’s important to note that these numbers did fluctuate from year-to-year and there was no discernable pattern regarding these fluctuations.

What About School Shootings?

The fact that at-school victimizations declined might be surprising to some, considering the uptick in mass shootings in recent years and increased fear over school shootings. As stated above, a July 2021 report shed some light on the prevalence of school shootings from 2000 through 2020. That report cited data collected as part of the K-12 School Shooting Database (K-12 SSDB), which is used to form the basis of several findings. On its face, it looks like the number of school shootings has increased by a wide margin. But the definitions used, coupled with how data are presented, are not entirely reflective of what we think of in terms of a “typical” school shooting.

The K-12 SSDB compiles information on school shootings from publicly available sources into a comprehensive database. Their definition of a school shooting isn’t necessarily reflective of the “typical” school shooting though, and includes “all incidents in which a gun is brandished, fired, or a bullet hits school property for any reason, regardless of the number of victims (including zero), time, day of week, or reason (e.g., planned attack, accidental, domestic violence, gang-related).” During the coronavirus pandemic, shootings that occurred on school property during remote instruction were also included. Unfortunately this definition casts a wide net in terms of what is considered a school shooting. This measure is perplexing and in some ways seems futile, as most people think of school shootings as horrific acts that are committed purposefully (i.e., not accidental) during school hours, where there is at least one victim.

The data presented by the NCES on this (who partnered with BJS on this project) are broken down into a few categories that provide a little more detail: shootings with injuries and shootings with deaths. These data show that, from 2000 to 2021, the number of annual school shootings with injuries ranged from a low of 11 in the 2009-10 school year (5 of which caused deaths) to a high of 93 in the 2020-21 school year (43 of which caused deaths). Still, these data are not parsed out by whether or not they occurred during school hours, nor are they parsed out by reason (e.g., accidental, drive-by, targeted attack). This is a little misleading, considering that many of these shootings occurred outside school hours and are not considered a “typical” school shooting.

A Deeper Dive

To better understand the prevalence of “typical” school shootings, I took a deeper dive into the raw data from the K-12 SSDB. I also examined the raw data from two other sources: the Violence Project’s database on mass shooters and the Washington Post’s database on school shootings.

From 2000 to 2021, the K-12 SSDB database counted 1,238 school shootings overall, ranging from a low of 15 in 2010 to a high of 251 in 2021. But the reasons for these shootings is wide-ranging — for example, 129 were accidental, 93 were suicides, and 93 were random drive-by shootings. Further, only 673 actually occurred during school hours. When looking at the overall number of school shootings that resulted in injuries (which are conveniently reported by the NCES here, but are not revealed in the actual data), numbers drop dramatically, with annual numbers ranging from 11 to 93. When looking at the number of school shootings that caused deaths, this drops even more, with annual numbers fluctuating between 5 and 43.

When looking at the SSDB raw data in more detail, there were 96 “active shooter incidents” that occurred during school hours, about half of which were pre-planned. In contrast to the SSDB’s definition of a school shooting, active shooter incidents are defined by the FBI as an individual who is “actively engaged in killing or attempting to kill in a populated area.” Arguably, active shooter incidents that occur during school hours are more indicative of what we think of as far as a “typical” school shooting. Most commonly, these incidents occured about once per year (this was evidenced in 2020, 2011, 2009, 2008, and 2004), though some years experienced higher numbers (e.g., 2018 and 2001 saw 11 and 9 of these incidents, respectively). The numbers seemed to ebb and flow quite a bit, with no real discernable trends upward or downward. In 2019, 2020, and 2021, there were 6, 1, and 5 active shooter incidents during school hours (respectively).

Interestingly, when looking at the data in a different way, the results appear much different than how NCES presented it in their report. Notably, the latter makes school shootings appear increasingly prevalent, especially in recent years. Though when looking at active shootings occurring during school hours, these numbers are much lower. Because the data on “shots fired” is incomplete and there was no variable related to fatalities, I was unable to parse these out further in terms of active shooter incidents during school hours that resulted in injuries and fatalities.

When it comes to mass shootings (i.e., generally those with four or more fatalities) at schools, the numbers are even lower. The Violence Project collected data on mass shootings through February 2020. When looking at the last 20 years, this database identified 95 mass shooters, 6 of which were school shooters. Importantly, there can be more than one shooter per incident, but this can be parsed out by the incident level to identify how many mass school shootings occurred. Per this database, there were many years that did not see any school mass shootings at all, including the year 2019. 2018 saw two school mass shootings, which was the maximum in any year from 2000-2020.

The Washington Post’s database on school shootings indicates that “more than 311,000” students have been exposed to gun violence since the 1999 Columbine shooting. However, this is worded very carefully. Because it represents students who are exposed to gun violence (not numbers of fatalities or injuries), this calculation sums the entire student capacity from each school affected by a school shooting from 2000 through 2021. Their interpretation of the data also contends that 2021 saw 42 school shootings, supposedly the highest of any year since 1999. They also comment that, so far in 2022 there have been at least 24 acts of gun violence on school campuses occuring during school hours. Of course, I took a deeper dive into the raw data, which unfortunately are only available through April 2019. In this dataset, certain shootings were excluded from the counts, including shootings at after-hours events, accidental discharges that caused no injuries to anyone, and suicides that posed no threat to others.

The raw Washington Post data showed that a total of 231 incidents occurred between 2000 through 2019, with annual numbers ranging from a low of 7 in 2015 to a high of 25 in 2018. Regarding the number of shootings resulting in injuries or death, the data tabulated 168 total incidents, which ranged from a low of 2 in 2015 to a high of 16 in 2018. 59 incidents resulted in deaths; 47 resulted in only one death and 6 met the criteria for a mass shooting (i.e., 4 or more fatalities). Regarding school shootings resulting deaths, annual numbers by year ranged from a low of 1 (which was seen in 2009, 2007, and 2002) to a high of 7 in 2018.

Conclusion

Looking at these sources together, there have been a total of 6 mass shootings in schools in the last 20 years (per the Washington Post and the Violence Project). 2018 was a particularly bad year for school shootings, which saw the highest number of active shooter incidents (n=11, per the K-12 SSDB), shootings resulting in injuries (n=16, per the Washington Post), shootings resulting in deaths (n=7, per the Washington Post), and mass shootings (n=2, per both the Violence Project and the Washington Post). The two mass shootings experienced in 2018 also resulted in higher than average death counts; these included the shootings in Parkland, Florida that killed 17, and Santa Fe, Texas that killed 10. 2012 was also a bad year for school shootings — there was only one mass shooting that year, but it was the infamous massacre at Sandy Hook Elementary School in Connecticut, which resulted in 26 deaths.

Looking at these sources alongside each other, it seems that the latter two are better in assessing the prevalence of “typical” school shootings. This is likely more useful when talking about gun violence, rather than citing inflated numbers based on loose definitions. This is not to say that school shootings are not concerning, however. But the number of school shootings might be much lower than what is commonly purported, which has important policy implications.

An important takeaway: the shocking numbers making headlines are often far different from what the data actually show. It is unclear why the BJS are not more forthcoming about this, though it is certainly disappointing. Relatedly, the other two datasets are not entirely forthcoming about results either, and it took a closer look at the actual data to come up with the breakdowns presented here. Nonetheless, this should help you understand how numbers can differ so vastly based on definitions used. I also look forward to breaking down the 2021 data when it becomes available.