The Nature, Trends, Correlates, and Prevention of Mass Public Shootings in America, 1976-2020, [United States] (ICPSR 38331)

Time Period(s) View help for Time Period(s)

Date of Collection View help for Date of Collection

Data Collection Notes View help for Data Collection Notes

Hide

Study Purpose View help for Study Purpose

Criminologists have increasingly turned their attention to studying deadly mass shootings--incidents in which four or more victims are killed by gunfire in a 24-hour period--especially those occurring in public settings. Until fairly recently, however, there was little interest among academics, and criminologists in particular, in conducting research on mass shootings. Although there were several mass public shootings during the 1980s and beyond, including a few with relatively large victim counts, many considered such events to be so rare and idiosyncratic that mass shooting was not a suitable focus for empirical analysis.

Prior to the last few decades, mass shootings did not have the same draw for the mass media as did serial homicides, a topic which has been a longtime focus of criminology scholarship. Until recently, the technology to provide viewers with live video coverage of unfolding tragedies did not exist, nor was there an array of cable news channels that could devote hour-upon-hour to these shootings and their aftermaths without having to preempt regularly scheduled programming. Additionally, social media, which can rapidly and widely relay information and images surrounding rare and shocking events like mass shootings, had not yet been developed. However, the year 2012 became a watershed when three large-scale shootings--at a California university, a Colorado movie theater, and especially a Connecticut elementary school--had criminologists, other social scientists, and the general public starting to take notice.

With interest in the topic growing, many academics were frustrated by the lack of official data on mass shootings. Some researchers looked to the FBI's Supplementary Homicide Reports (SHR) as a resource; however, many incidents, including some with large body counts, are missing from the SHR. Also, in many cases, a police department will improperly include data on injured victims in a homicide incident record, making it appear as if it were a mass killing. In the absence of a reliable resource on cases, several news organizations and academic groups attempted to build their own databases, some attempting to backfill cases retrospectively. Over a dozen databases were created, differing in the timeframe covered and definition of what constitutes a mass shooting in terms of victim threshold, whether the threshold was based on all victims shot or just those killed, motivation, location, and victim-offender relationship.

Because there was no consensus on definition, the competing databases tell very different stories about incidence and trend, including the number of incidents and victims. Some databases, such as the Mother Jones initiative, focused on the narrowest set of cases (indiscriminate fatal shootings in a single public setting generally by a lone gunman) while others, such as the project launched by Everytown for Gun Safety, included cases regardless of location, motivation, or victim-offender relationship. Besides differences in defining characteristics, there was also no agreement as to the minimum victim count, with thresholds for the number of victims killed ranging from three up to six. Even more problematic, there remains disagreement as to whether the victim threshold should include all those shot or just the fatalities. Imprecise reporting on these cases can easily deceive the public, inadvertently creating panic. News stories often conflate active shooter events with mass shootings. However, most of the wannabe mass killers fail to realize their goal. Nearly half of these events result in at most one victim fatality. One-quarter involve no deaths, and some result in no one even being injured.

Several datasets were created and analyzed as part of a three-year project on mass public shootings in the United States, funded by the National Institute of Justice. Notwithstanding the date range contained in the project title, the researchers updated the data as the work progressed and used the most up-to-date data for analyses and associated publications as they became available. They performed the following analyses:

• An in-depth description of mass public shooting incidents, offenders, and victims with a comparison to general homicide patterns
• An evaluation of whether state-level gun legislation, such as concealed carry laws and prohibitions on large capacity magazines, affect the incidence and severity of mass public shootings
• An examination of the incident, offender, and victim characteristics that impact the newsworthiness of mass public shootings
• An assessment of the extent to which timing of mass public shootings suggest a contagion effect based on media coverage
• A comparison between completed mass shootings and thwarted plots
• A forecast of the severity of mass public shootings over the next couple decades
• An estimate of the global prevalence of mass public shootings accounting for missing incidents

Study Design View help for Study Design

As a central part of this research initiative, a database was created of mass public shooting incidents, offenders, and victims that occurred in the United States from 1976 through 2020. With this as the stating point, the researchers then carried out a series of analyses using these and other data.

A triangulation approach was utilized to create the first, second, and third datasets (the characteristics of incidents, offenders, and victims, respectively), incorporating the FBI's Supplementary Homicide Reports and media reports to provide a baseline for inclusion of cases. Several other relevant datasets, both published and unpublished, were consulted as well. Three members of the research team then reviewed cases in order to arrive at consensus for inclusion. The final dataset covering the years 1976-2020 includes 165 mass public shooting incidents, 169 offenders, and 1,208 victims (including 41 who died but not from gunshots). Detailed information on each case was drawn from open-source news archives and various other publicly available documents.

For the fourth dataset (the study of media contagion of mass shootings), day-by-day tallies from January 1, 2000 through December 31, 2018 of the number and lengths of news stories on the general topic of mass shootings were generated from 16 major newspapers weighted by their circulations (MP), from the Associated Press national wire (AP), and from network television news broadcasts (TV). These daily counts of media content were then merged with a binary indicator of whether there was a mass public shooting on each date (there was never more than one) and, if so, the number of victims killed, as well as with the count of all mass shooting incidents on each date (16 had two incidents) and the total number of victims.

To create the fifth dataset (the newsworthiness of mass public shootings), detailed information on each case was drawn from open-source news archives and various other publicly available documents. For each of the included news sources (Associated Press State and Local Wire, Associated Press National Wire, and The New York Times), the number of stories that referenced the assailant's name during the 15 days following the incident, including the day of the incident, was obtained through Nexis Uni.

Data for the sixth and seventh datasets (characteristics of averted mass shooting incidents and offenders, respectively) were collected from available reports from news media stories as well as available correctional agencies or courts; the researchers relied on databases provided by other researchers and those found online as well as open sources such as media accounts. Cases were included if they met the following criteria: the perpetrator(s) had access to or plans to gain access to a firearm and made a threat of a mass shooting; if the perpetrator(s) did not have access to a firearm, the plan had to be specific and detailed (e.g., map of a building, specific targets). In all cases, the researchers required some threat or plan of a mass shooting to be present, rather than inferring intention from the perpetrator(s) action (e.g., shooting at people randomly may or may not be a planned mass public shooting).

Sample View help for Sample

A mass public shooting was defined as any event in which four or more individuals, not including the assailant(s), were killed by gunfire in a public setting within a 24-hour period. Mass shootings involving both private and public settings were included if at least half of the victims were killed in a public place and the death toll was under seven, or if at least four of the victims were killed in a public place and the death toll was seven or more. Mass shootings that were associated with another form of criminal activity--such as gang conflict, illegal drug trade, and robbery--were excluded regardless of location.

Averted or thwarted mass public shootings were defined to be as similar as possible to the above. Specifically, averted mass public shootings were defined as as "any plot, plan, or threat to shoot four or more individuals in a public location within a 24-hour period, absent of other criminal activity (e.g., robberies, drug deals, gang conflict, etc.), in which no injuries (other than to the perpetrator) occurred."

Time Method View help for Time Method

Universe View help for Universe

All mass public shootings in the United States from 1976-2020, defined as any event in which four or more individuals, not including the assailant(s), were killed by gunfire in a public setting within a 24-hour period, absent any associated criminal activity (such as robbery, gang conflict, or illicit drug trade).

Unit(s) of Observation View help for Unit(s) of Observation

Data Type(s) View help for Data Type(s)

Description of Variables View help for Description of Variables

The first, second, and third datasets contain relational data files with information about the mass shooting incidents, the offenders, and their victims. In total, there are 72 variables: 28 in the incident file, 32 in the offender file, and 12 in the victim file. The datasets cover the years 1976-2020 includes 165 mass public shooting incidents, 169 offenders, and 1,208 victims (including 41 who died but not from gunshots). The first dataset contains information about the dates and locations of each incident, numbers deaths and injuries, and numbers and types of weapons used. The second dataset contains information about the offender in each incident, including demographics, criminal history, history of mental illness, whether or not they owned (or legally acquired) the guns used in the incident, whether they shared information about intent prior to the incident, and the outcome for the offender. The third dataset contains demographic information about the victims in each incident as well as their relationship to the offender.

The fourth dataset contains 6,940 records (one for each day between January 1, 2000 and December 31, 2018) and 12 variables: besides the date and the U.S. population (interpolated based on yearly figures), there are four pertaining to mass shootings and six concerning media coverage.

The fifth dataset contains one record per offender and includes information about the incident, that offender, the victims (in the aggregate), and related news coverage--for a total of 24 variables. It covers the years 2000-2019 and includes 97 mass public shootings incidents, 99 offenders, and 763 victims.

The sixth and seventh datasets contain information regarding averted or thwarted mass public shootings; the former pertains to the incident and the latter to the perpetrator. With respect to the incident, the researchers collected data on location, whether there were planned targets, whether the threat was in progress, who reported the threat and to whom, the length of the threat and plan, response to the threat, motivation and weapon information (whether firearms were available and how acquired). For the perpetrator file, the researchers collected demographic information (age, race, gender), whether they were employed or a student, mental illness and arrest information. There are 21 variables for the incident file and 18 for the offender file.

Response Rates View help for Response Rates

Not applicable

Presence of Common Scales View help for Presence of Common Scales

None

Hide

Original Release Date View help for Original Release Date

2022-09-14

Version History View help for Version History

Hide

Weight View help for Weight

Not applicable

Hide