In the wake of the two most recent and tragic mass killings in El Paso, Texas, and Dayton, Ohio, the news cycle has brought with it no small amount of speculation that mass killings are on the rise and that one attack tends to inspire a copycat. While these assumptions are understandable, the data simply does not support them.
It feels like every time we turn on the news, there is another report of a senseless mass killing. These attacks claim the lives of multiple victims and leave families and communities shattered. Since the two most recent events, experts in advanced analytics bring to the forefront insight that could help law enforcement agencies and organizations responsible for maintaining the safety at public and private venues prevent such tragedies.
“Mass killings” are defined as an event that leads to four or more deaths, not including the perpetrator. These events transpire over a brief period of time, hence why serial killings are excluded. They can be carried out using several possible methods, from the use of firearms to stabbings and even arson. These incidents also happen under various circumstances—public events or spaces, domestic killings, or even robberies.
A USA Today database that chronicles these attacks found that between Jan. 1, 2006. and Dec. 31, 2017, there were 361 mass-casualty events. That is just over 30 per year. More recent data from 2018 and 2019 suggest this rate has not changed.
Based on what we already know of the two most recent attacks, the venues were different, and the motives of the shooters were different. There are still many unanswered questions that when addressed, will provide further clarification.
What we do know is that law enforcement agencies and organizations presently have access to the type of data and data models that can help them better address and possibly prevent future mass killings or the proportion of collateral damage.
The good news is that we have a model to capture the timing of such events. The bad news is that this model reinforces the observation that such events are randomly occurring phenomena. In the end, the reality is that there is no easy way to predict when and where the next such event will occur.
What the model suggests is that mass killings with a firearm are uniformly distributed over time, which suggests that they have occurred at a constant rate. This observation dispels the common belief that such events are occurring more frequently. The model also suggests that the events possess what is termed a “memoryless property.” This means the occurrence of a mass killing with a firearm does not imply that another similar event is more (or less) likely to occur in the near future.
So, what actions can be taken to protect against such an event?
When compared to other crimes, public mass killings nationwide occur at a proportionately lower rate. This coupled with the large number of public venues where they can occur, the inclination to have constant police presence at each potential target is neither practical nor cost effective.
A better solution would be through infrastructure-based interventions, like buildings that can be easily locked-down to protect people until law enforcement arrive, which are likely to be more cost effective in preventing or stalling attempted mass attacks.
In addition, the ability of law enforcement to rapidly respond to such events once they become apparent is critical to minimize the loss of life and damage. The mass killing in Dayton could have been much worse had police not been on hand, trained, and able to neutralize the threat within two minutes of its commencement.
This reinforces the need for specialized training and enhanced communication so that first responders can take effective action to terminate a mass killing event before its footprint of destruction grows. Broad education is also critical to create a prepared and informed population at the onset of a mass killing event.
There is no easy solution to predict or prevent mass killings. Think back to the June 2017 shooting in Alexandria, Va., during a congressional baseball practice. Although not fulfilling the strict criteria for a mass killing, the event shows that even in the most controlled environments, with law enforcement personnel present, the potential exists for a mass casualty event to unfold.
The best guard against these situations starts with the understanding that they are not happening with increasing regularity, but that their inherent unpredictability requires constant vigilance, and a clearer idea of the actions we will take and the roles we will play should one ever occur in our presence. Data models just may be some of the most reassuring and reliable tools we can use in this effort.
Sheldon Jacobson, a professor of computer science at the University of Illinois at Urbana-Champaign, is chairman of the INFORMS National Science Foundation Liaison Committee.
Morning Consult welcomes op-ed submissions on policy, politics and business strategy in our coverage areas. Updated submission guidelines can be found here.