Tag Archives: protest movements

A Theory in Search of Evidence

On Sunday, the Pittsburgh Post-Gazette published an editorial headlined “Murder on the Rise.” It states that the homicide rate is up this year “in more than 30 major American cities”—but not including Pittsburgh. It repeats the “theory” that this change is due to the “Ferguson effect,” which argues that police, facing criticism from African-American activists, have been “less aggressive in patrolling problem neighborhoods.” It concludes that a return to the “bad old days” of high homicide rates is unacceptable and that law enforcement and the Justice Department “must bring their resources to bear to figure this out.”

The statistics they cite come from an August 31 New York Times article whose authors, Monica Davey and Mitch Smith, surveyed an unspecified number of cities and reported that at least 35 of them have seen increases in “murder, violent crimes, or both.” The article is accompanied by a chart showing increases in the homicide rate of between 4% and 76% in ten cities. But not all cities have seen more killing. They mention three cities where murders have not increased.

There are several problems with the Times article. First of all, their statistic has neither a numerator nor a denominator. Since they lump murder together with other violent crimes, the authors don’t specify exactly how many cities reported increases in murders. More importantly, they fail to report how many cities they surveyed—a critical point, since if the actual homicide rate is unchanged, half of cities can be expected to show increases just by chance. Finally, they give no summary statistic indicating whether the overall homicide rate in the cities surveyed is up or down, by what percentage, and whether the change is statistically significant. This is important since homicide rates in many cities fluctuate quite a bit from year to year, and the number of violent crimes was unusually low in 2014.

The authors had to conduct their own survey because there are no up-to-date, authoritative data on homicides in the nation’s cities. The FBI’s Uniform Crime Reports, which provide city-level crime data, do not come out until the following year. This lack of hard data allows people to claim that the crime rate is either increasing or decreasing, whichever their ideology leads them to prefer, based on incomplete samples.

Even if the homicide rate has increased significantly, there is nothing to connect it to the protests following the death of Michael Brown in Ferguson, MO, one year ago. To their credit, the Times mentions the research of criminologist Richard Rosenfeld, who found that homicides in the St. Louis area peaked before Michael Brown was shot, and who states that there is no evidence of a “Ferguson effect.” Unfortunately, this finding is buried deep in the article and is surrounded by stories about specific murders and theories which lack empirical support.

A man holds his grandson at a rally protesting the death of Walter Scott in Charleston, NC. (Photo: AP/David Goldman)
A man holds his grandson at a rally protesting the death of Walter Scott in Charleston, NC. (Photo: AP/David Goldman)

Three days later, the Washington Post got into the act in an article about the shooting death of Illinois police officer Charles Gliniewicz. Although there is no evidence that race or anti-police sentiment played a role in his death, Jim Pasco, executive director of the national Fraternal Order of Police and several other law enforcement sources hold the Black Lives Matter movement responsible for what they imply is an increase in the killing of police officers. One of them blames the “dangerous national rhetoric that is out there today.” One Black activist who disputes these claims is also quoted.

Buried in the middle of this article is a critical fact. The National Law Enforcment Officers Memorial Fund reports that 24 police officers have been killed by suspects so far this year, “the second lowest number in the past five years.” Yet according to a September 1 Rasmussen poll, 58% of likely voters believe “there is a war on police in America today,” while just 27% disagree.

Ta-Nehisi Coates criticized the Times article as an example of “false equivalence,” since the authors don’t make a clear distinction between opinions and facts. Unless they read the story carefully, readers could easily conclude that “there is as much proof for the idea that protests against police brutality caused crime to rise, as there is against it.” (The author of the Post-Gazette editorial seems to have fallen into this trap.) This same argument applies to the Post article.

Social scientists refer to this journalistic practice as false balancing. It’s found, for example, in articles about climate change which imply that scientists are evenly divided as to whether the climate is changing, and which fail to evaluate the quality of the evidence each side presents. Cautious journalists have been transformed into stenographers, faithfully reporting what everyone says but never examining whether what they say makes sense. Paul Krugman once suggested that if candidates of one party said the Earth is flat and and the other party said it’s a sphere, the newspaper headline would read “Views Differ on Shape of Planet.”

As of this writing, at least 820 people have been killed by the police so far this year. African-Americans, with 13.2% of the population, account for one-third of these deaths. For those deaths in which the victim is unarmed, the Black percentage is even higher. While it is clear that major changes in policing are needed, law enforcement is digging in its heels and fighting back with theories such as the “Ferguson effect.” By failing to make it clear that claims of a “war on police” are without empirical support, the corporate media play into the hands of those who are trying to convince the public to sacrifice the civil liberties of African-Americans in exchange for an illusory increase in public safety.

Update (9/11/15):

police1

The American Enterprise Institute published this chart showing the number of gun-related police deaths per capita from 1870 to the present. As they point out, exaggerating the danger to police has been used as a justification for the increasing militarization of U. S. law enforcement.

You may also be interested in reading:

White People Don’t Riot:  A Manual of Style For Ambitious Young Journalists