Q&A: Sherry Towers on the contagion effect of mass shootings

When the media are calling Sherry Towers, it's often on a sad day.

Towers, a physicist at Arizona State University, has become a regular source for journalists since her paper on how the media coverage of mass shootings can inspire future mass shootings was published earlier this year. After last week's tragic shooting at Umpqua Community College in Roseburg, Oregon, that left 10 people dead, Towers' phone was ringing once again.

The statistician, modeler and research professor spoke to about the recent tragedy and her study on the contagious nature of mass shootings.

Question: Does this recent incident seem to fall within the findings of your research?

Answer: Our research examined whether or not there was evidence that mass killings appear to inspire copycat killings. We found evidence that killings that receive national or international media attention do indeed inspire similar events a significant fraction of the time. In the case of this particular tragedy in Oregon, there have been reports that the killer apparently had a blog where he praised Vester Flanagan, the killer who shot two news reporters and a bystander on camera in August. If the reports are true, then indeed this recent killing may be an example of the contagion we have seen evident in so many other killings.

Q: So is conventional wisdom correct that some mass shootings are copycats?

A: Yes, we believe so. In fact, during the trial of the Aurora theater shooter, the father of one of the victims asked the media not to cover the trial, because he feared that the coverage would inspire copycat killings. Unfortunately, his prediction came true. A gunman opened fire in a Louisiana movie theater, and in a Tennessee movie theater a man attacked people with a hatchet. All within two weeks.
 
Q: How does just looking at numbers prove that?

A: The hallmark of contagion is seeing events unusually bunched together in time. The details of our analysis, where we fit a mathematical model of contagion to the data to quantify the level of contagion, are quite technical. But really, what it essentially amounts to is seeing if there are unusual groupings of events. In mass killings (four or more people killed), where the tragedies usually get national or international media attention, we saw significant evidence of this kind of unusual bunching. In mass shootings — with less than four people killed, but at least three people shot — we didn't see any evidence of unusual bunching. Interestingly, those events are so common in the U.S., happening once every few days, that they don't even make it past the local news. Because we saw evidence of contagion in high-profile events, and no evidence of contagion in events that mostly just got local news, we hypothesize that media attention may be the driver of the patterns we see. This kind of contagion has been suspected for a long time; our study is the first to quantify it.

Q: How does this compare to the probability of, say, a disease spreading, since we’re talking about a contagion phenomenon?

A: With a disease, you usually need close contact to spread it to someone else. In this case, the news media act as a "vector" that can transmit the infection across a very large area. The people who are susceptible to ideation to commit these terrible acts are quite rare in the population ... that's why it appears that it takes a lot of media coverage over a wide geographic area for this kind of contagion to take place.
 
Q: What is the news media’s role in this? Do they push up the numbers?

A: It appears that yes, national media coverage does end up increasing the frequency of these tragedies. However, the U.S. Constitution ensures freedom of the press ... we cannot legislate restrictions on the press to avoid this. It has to be a voluntary move. In fact, most press agencies will not report on suicides for exactly this reason ... suicides have been shown to be contagious. The sheriff in Oregon made the decision not to mention the killer's name. Perhaps his choice will be the beginning of a larger national conversation on how we can choose (or choose not) to cover these events.

Q: What is the next step in this research? What can you answer by taking it further?

A: It needs to be pointed out that we did this research without funding, because there has been a Congressional moratorium since the 1990s on funding for research into firearm violence. We had to do this study unpaid, in our spare time. This lack of funding is a huge barrier to better understanding of the dynamics that underlie these tragic events. No other developed country in the world expects its scientists to work for free, spending their evenings and weekends studying public-health problems as pressing as the out-of-control firearm violence in the U.S. Because of this moratorium on funding, there aren't even official statistics on these events. Given the amount of media attention that is paid to these tragedies, it always surprises me that the complete lack of federal funding for research into the problem is rarely mentioned. So yes, I and many other researchers would like to devote more of our time to studying this problem, but there are only so many hours available of our time that we can afford to work for free.