Let's pretend that researchers are investigating acts of violence between players during hockey games. And let's further pretend that they are interested in determining if violent behavior has a racial component.
Would you be surprised to find out that most acts of violence happen between white players? Well, of course not. The National Hockey League (NHL) consists almost exclusively of white people. (A figure from 2011 claims the league is 93% white.) So, if the researchers do not control for the fact that most hockey players are white, they could come to the erroneous conclusion that white hockey players are more likely to be violent.
For this reason, race, age, and other demographic factors are commonly controlled in epidemiology studies. It makes no sense to compare one group to another group if researchers do not bother to control for confounders -- that is, factors like race or age that can cause researchers to draw the wrong conclusion.
Yet, precisely that just occurred in a new paper that concludes that police violence disproportionately affects people of color.
A Race-Baiting Epidemiology Paper Smears Police Officers
NBC News ran the following click-bait headline:
The journalism, as usual, is a dumpster fire. But the paper upon which it is based is even worse.
The authors, who published in the Journal of Epidemiology and Community Health, examined how police shootings affected a measure known as "years of life lost" (YLL) among various racial and ethnic groups. This measure is used to determine the burden of disease. For instance, if a person is expected to die at age 80 but dies from cancer at age 72, then for that individual, YLL is eight.
They applied this metric to police shootings. Their conclusion was that people of color (namely, blacks and Hispanics) had more years of life lost than whites. But here's the rub: The authors didn't control for the types of crimes being committed. That's epidemiological malpractice. In fact, it's such a substantial oversight that it constitutes a fatal flaw to their research.
Why? Because the police respond differently to different crimes. They don't generally pull guns on jaywalkers. They do, however, pull guns on people thought to be involved in a violent crime. And there is an incredibly large difference between racial/ethnic groups in the U.S. when it comes to violent crime.
Consider homicide. CDC data show that, compared to whites, the homicide rate among blacks is 8 times higher and among Hispanics two times higher. Obviously, in such a situation, a police officer is more likely to use lethal force. Thus, the relevant question is: If the underlying crime is the same (e.g., homicide), is a police officer more likely to use lethal force against a black or Hispanic person than a white person?
Research suggests the answer is no. Harvard economist Dr. Roland Fryer, who is African-American, published a paper that examined this issue. Despite his finding that police were more likely to use some type of force against blacks and Hispanics, for officer-involved shootings, he concluded that there were "no racial differences in either the raw data or when contextual factors are taken into account."
Don't Believe the Media on Issues Involving Race
Racial issues are so difficult to discuss in this country partially because the media lies about them non-stop. Any story that fans the flames of racial tension is immediately broadcast all over the world. The media -- and apparently, some researchers -- are willing to undermine social cohesion and smear police officers by spreading falsehoods for the purpose of getting more eyeballs. That may not quite reach the definition of "evil," but it certainly gets close.