Does social media coverage of fatal police-citizen incidents act as a virtual contagion? If a member of the public is killed by police, does it lead to future violence against law enforcement? Conversely, if an officer is killed in the line of duty, does it lead to future violence against citizens?
These are some of the questions raised in a new study published today in the scientific journal PLOS ONE by Seattle University faculty William Parkin and Matthew Hickman in the Department of Criminal Justice, Vladimir Bejan in the Department of Economics, and Veronica Pozo, Department of Applied Economics faculty at Utah State University. The research explores relationships between law enforcement killed in the line of duty, fatal use of force incidents and social media activity related to the Black Lives Matter movement. The researchers wanted to know if, at the national level, there is any empirical evidence of retaliation on the part of police or citizens for fatalities occurring in either group. At the same time they wanted to understand the influence, if any, of social media attention to tensions between the two.
The study found evidence that when an officer was killed in the line of duty, there was retaliatory violence against citizens by police on the same day. However, the relationship was different when comparing minorities and non-minorities. Specifically, an unexpected increase in the number of officers killed was associated with an increase in the number of minorities killed, but a decrease in the number of non-minorities killed. While the study cannot speak to whether the shootings were justifiable, the findings could signal a racially biased reaction to public killings of police.
The study also found evidence that an unexpected rise in fatal shootings of non-minorities by police were associated with an increase in fatal shootings of police during the same day. But fatal shootings of minority citizens by police were associated with a decrease in fatal shootings of police. These effects were statistically significant, but not as strong as the impacts of increases in officer line-of-duty deaths described above.
Importantly, all of the relationships described held steady while controlling for the daily number of Black Lives Matter-related tweets. These tweets could support or oppose the movement and might have nothing to do with fatal police-citizen violence. The tweets were a proxy for how much attention the topic received on any given day.
"Black Lives Matter-related tweets were associated with increases in both the numbers of minorities and law enforcement officers killed," Pozo says. "This implies that social media can rapidly spread a negative message and act as a contagion."
An uptick in Black Lives Matter-related tweets was associated with increases in fatal shootings of minority citizens, but not non-minorities. The effect of Black Lives Matter-related tweets on officer fatalities was also statistically significant, but not as strong as the effect on minority fatalities.
"Our study provides strong initial evidence of an increase in deaths of minority citizens by law enforcement following an unexpected increase in officers killed in the line of duty," says Parkin. "It also provides initial evidence that unexpected increases in social media posts highlighting the conflict between law enforcement and minority groups are associated with an increased risk of fatal contacts between the two."
Explaining these relationships
One theory that can help explain the findings is called "terror management theory." This psychological theory suggests that when an individual's mortality becomes either consciously or subconsciously salient, they are more likely to defend the dominant worldview. They will react positively to persons associated with "in groups"—social groups with which an individual identifies—of that perspective. They will also react negatively to persons associated with "out groups"—social groups with which an individual does not identify.
If police are made aware of the line-of-duty deaths of fellow officers, which the researchers argue is a form of priming, it may alter how they interact with individuals associated with perceived "out groups," such as the Black Lives Matter movement. Law enforcement officers would tend to view non-minority citizens as part of the "in group." and minority citizens as part of the out group. Engaging differently with members of the "out group" during contacts, compared to the "in group," could explain the opposite relationship in the deaths of minority versus non-minority citizens after an officer is killed in the line of duty.
Parsing the data
Researchers looked at nearly two years of daily data, from January 2015 to September 2016. The data was drawn from the Officer Down Memorial Page, The Washington Post's police shootings database and Twitter. The data was disaggregated on two different categories to understand whether the frequency of fatal use of force incidents varied based on the race or ethnicity of citizens killed (white non-Hispanics and minorities).
The models looked at whether an unexpected increase in any of the variables was significantly associated with a future change in other variables. For example, an unexpected increase in "fatal shootings of police" is defined as more than the average daily murders (by one standard deviation). The models allowed researchers to predict the number of citizens killed by police, as well as the number of Black Lives Matter-related tweets up to seven days following this unexpected incident.
Explore further: Analysis uncovers racial bias in fatal shootings by police
Vladimir Bejan et al. Primed for death: Law enforcement-citizen homicides, social media, and retaliatory violence, PLOS ONE (2018). DOI: 10.1371/journal.pone.0190571