On January 21, a man opened fire at a dance hall in Monterey Park, California, killing 11 people and injuring nine others. Two days later, on January 23, another man attacked two farms in Half Moon Bay, California, shooting and killing seven and injuring one.
Both shooters were seniors — 72 and 66 years old respectively — and both were Asian men. Both used semiautomatic handguns. Both tried to shoot multiple people in multiple places. And both incidents happened in public spaces, a few hundred miles from one another.
The similarities prompted a reader to ask: “Is there any evidence that the mass shooting contagion effect is getting stronger? Has it become more common for a mass shooting that receives national media coverage to be followed by others?”
There’s no clear evidence that the Half Moon Bay shooter was inspired by the Monterey Park one, but the question hits upon some confusing evidence regarding the behavior of mass shooters.
First, let’s break down some basics.
What is a contagion effect?
In social science, “contagion” generally refers to the spread of a phenomenon, often a behavior, through a population. When a person perceives a behavior as commonplace, they may be subconsciously or consciously inspired to act in a similar way. Researchers typically use statistical methods to identify a contagion effect by analyzing the rate at which a phenomenon normally occurs and whether specific events increase the probability of similar occurrences in the near future.
If a mass shooting contagion effect exists, one shooting would increase the probability of more shootings to come.
It’s not uncommon for mass shootings to cluster in time. In 2019, a mass shooting that targeted Latin American immigrants in El Paso, Texas, killed 23 and injured 23. Just a day later, a shooter in Dayton, Ohio, killed nine and injured 27. And the May 2022 shooting in Buffalo, New York, in which 10 people were killed and one wounded, was followed by the school shooting at Robb Elementary School in Uvalde, Texas, less than two weeks later, which killed 21 and injured 17.
While none of those shootings appeared to be directly linked, earlier examples of clustering led some researchers, advocates, and commentators to hypothesize that there is a contagion effect increasing the rate and frequency of mass shootings.
So are mass shooters typically copycats?
Not necessarily. Contagion differs in significant ways from the “copycat effect,” which some social scientists describe as a more specific mechanism of contagion. Copycatting is a more direct and often conscious form of imitation, occurring when a person takes specific inspiration from the behavior of another and seeks to replicate it. In other words, contagion refers to the more general spread of a phenomenon and the rate at which it happens, while a copycat is identified by the motivations of an individual perpetrator. And while copycatting may play a role in broader contagion, it’s not necessarily the main contributor.
Indeed, many mass shooters have been directly inspired by their predecessors, seeing them as role models. In the case of Sandy Hook, the perpetrator who killed 20 school children and six teachers took inspiration from the Columbine and Virginia Tech shooters years before. More recently, the FBI warned that the widespread distribution of the 2022 Buffalo mass shooter’s racist screeds and plans could lead to copycats. The Buffalo shooter himself was influenced by earlier mass murders.
What does the evidence say about the contagion effect?
In 2015, data scientist Sherry Towers was one of the first to examine contagion in mass shootings. She and her co-authors found that mass shootings were temporarily contagious and increased the probability of future shootings for up to 13 days, with each mass shooting inciting on average 0.2 to 0.3 future attacks.
“Our model showed that there was this unusual bunching together in time, but that doesn’t mean that all of them were necessarily due to a potential contagion effect, just some fraction of them, and our analysis found that was 20 to 30 percent,” Towers said.
Towers and her co-authors did not identify a specific cause or mediator. The authors hypothesized that increased news coverage of a mass shooting could serve as a mediator for contagion to make future shootings more likely, but they did not test that theory in their study.
Other studies have gone further in pinpointing what drives the contagion effect. In a 2022 study published in the European Economic Review, researchers Michael Jetter from the University of Western Australia and Jay K. Walker from Old Dominion University in Virginia found that increases in news coverage of mass shootings predicted an increase in mass shootings for up to a month after.
Jetter and Walker attempted to quantify media coverage and add that as a variable in their model. In other words, they tried to determine whether an increase in news coverage of mass shootings led to more shootings, or whether shootings simply increased news coverage.
To do that, they measured the number of segments devoted to shootings on “ABC’s World News Tonight” and used natural disasters as what’s known as an “instrumental variable.” They were able to estimate whether elevated media coverage actually influenced future shootings by comparing those against instances in which the natural disasters dampened shooting coverage.
“The missing coverage actually seemed to be associated with fewer events in the future,” Walker told The Trace.
In 2018, Towers published a follow-up to her 2015 paper, in response to some studies that used other statistical methods to attempt to replicate her findings but failed to identify a contagion effect. In the follow-up, she and her coauthors compared their model’s predictions to several years of new data and found that it predicted the patterns in the new data well.
So the consensus is that the contagion effect exists?
Several other studies have found little or no evidence of contagion and little or no evidence of contagion from news coverage. The differences could come down to whether and how the researchers quantified news coverage, what shootings they included in their samples, and the statistical methods they used.
Some later studies disaggregated attacks by type, separating mass shootings in public spaces with high death tolls — like Sandy Hook and Virginia Tech — from other types, like a man killing his family, home invasions, and gang- or drug-motivated shootings. Their reasoning was that similar attacks would be more likely to influence one another — and not all mass shootings receive much, if any, news coverage outside local news.
Indeed, the majority of mass shootings are not media spectacles. Many occur in private spaces and are targeted with specific motivations aside from indiscriminate killing or terrorism. Though these multivictim shootings are far more common, other studies have shown that they receive little to no national media attention. And without news coverage, there’s limited means for contagion.
James Alan Fox, a criminologist at Northeastern University, argues that prior studies failed to measure the quantity of media attention, skewing their results in favor of finding evidence of contagion. That’s because the incidents may be clustered in time but not necessarily a result of contagion. “There are external forces, like economic conditions, that could influence shootings,” Fox said. “School shootings certainly only happen when school is in session.”
In a 2021 article published in the journal Statistics and Public Policy, Fox and four co-authors found no evidence of a contagion effect and no evidence that media coverage caused a contagion effect. They quantified news coverage by measuring stories in more than a dozen major newspapers, segments on several television news programs, and stories on the Associated Press wire service. They also separated their data on mass public shootings from shootings that typically receive little or no national media attention.
Using a statistical method known as a multivariate point process model, which looks at the interplay between phenomena that occur over time, they found that mass public shootings do result in significant spikes in news coverage, especially in the first three days after a shooting, while other types of mass shootings don’t. But they did not find that the spikes led to additional shootings.
Fox’s study also has some limitations. By limiting the data to only mass public shootings, the sample size was smaller. Gun violence — and mass shootings in particular — is difficult to study. There is a dearth of reliable data and no universally agreed upon definition of mass shootings. And mass public shootings, though tragic and deadly, are still infrequent compared to other phenomena that social scientists research. That means that sample sizes are inherently small.
“We have different samples. And we’re using different methods,” Walker said of the differences in findings. “It’s not totally surprising that you would have perhaps different findings.”
Has it become more common for a mass shooting that receives national media coverage to be followed by others?
Mass shootings themselves have become more frequent and deadly in recent years. Of the 10 deadliest mass public shootings in the United States, eight have happened since 2010, and six of those have happened since 2015. USA TODAY‘s mass killing database — which includes mass murders by other means besides shootings — shows that 2019 and 2022 saw the most mass killings and mass shootings on record, but 2018 saw the most mass shootings when only including those that occurred in public spaces. The Gun Violence Archive recorded the most mass shootings in 2021 and the most mass murders in 2022, but the GVA uses a lower threshold to define mass shootings and only began tracking them in 2014.
Assuming for a moment that a contagion effect exists, one would expect that an increase in mass shootings, especially high-profile public mass shootings, could lead to a vicious cycle in which more shootings lead to more shootings. That doesn’t necessarily mean that the contagion effect itself is getting stronger, but that there is a higher baseline of shootings that could lead to contagion. The baseline could be higher because of a number of factors, including economic conditions, changes to gun laws or the number of guns in circulation, and other factors.
“I don’t know that we can claim that the effect has gotten stronger or that necessarily [the effect] has become more common,” Walker said. “But if the shootings are coming more frequently and there is this empirical link, then you would anticipate that, as they become more frequent, this would also encourage more, more frequent shootings in the future.”