Is The Increasing Number of Cops Killed by Perps Really a "Disturbing Trend"?

Ten times more civilians were killed by cops than cops were killed by civilians in 2008, but you won't find that information in Tuesday's New York Times story on the "disturbing trend" of officers killed by perps.

"As violent crime has decreased across the country, a disturbing trend has emerged," reads the lede of the Times story. "Rising numbers of police officers are being killed."

Newspapers live and die by trend stories. If three dogs are spotted in a park wearing S&M outfits, or five women in disparate parts of America are revealed to be potty training their cats, then a trend is happening, and The New York Times will tell you all about it.

In this case, the trend is "72 officers were killed by perpetrators in 2011, a 25 percent increase from the previous year and a 75 percent increase from 2008."

But we should ask ourselves: What makes a trend? If it's statistical significance, then 72 perp-caused deaths in 2011 versus 56 in 2010 is statistically less noteworthy than the increase in deaths from whooping cough (26 in 2010, versus 15 in 2009), Arthropod-borne viral encephalitis (9 in 2010, versus 2 in 2009), and malaria (9 in 2010, versus 3 in 2009); yet none of those increases made the front page of the Times.

The story also fails every other applicable trend test. 

Let's say, for instance, that there were other, less mathemetical considerations. Let's say the Times wanted to highlight an increase in perpetrators killing police officers as reflective of an increasingly anti-authoritarian or lawless republic—hence the use of the modifier "disturbing" before the word "trend."

That argument doesn't hold water, either. According to the FBI, "Most (57) of the alleged offenders had prior criminal arrests." As of 2009–the most recent year for which data is available–there were 1,021,456 law enforcement officers employed in the United States. Assuming that number didn't go down significantly in two years, is the killing of 72 LEOs versus 56 representative of a larger cultural shift? No. If that number went up, then the number of killings, as a percentage of total number of LEO, may be stagnant or even–gasp–a decrease. 

There's arguably an even bigger problem with the Times' story, and that's the absence of any data about how many civilians the cops have killed, even though that information is widely available, as demonstrated by the Advocacy Center for Equality and Democracy:

• From 2003 to 2009, 4,813 people died in relation to an arrest in "all manners of deaths." Each year ranged from 627 (2003) to 745 (2007). Source – Andrea M. Burch, U.S. Department of Justice, Bureau of Justice Statistics,Arrest-Related Deaths, 2003-2009 – Statistical Tables, November 2011.
• Of those, 2,913 (about 6 in 10) were reported as "homicide by law enforcement." Each year ranged from 375 (2004) to 497 (2009). See Burch.
• In the only year in which the NYT article and the Bureau of Justice Statistics report overlap, 2008, law enforcement killed roughly 10 times the number of people during arrests (404) than officers killed (41). See Burch.
• Since 2001, at least 500 people have been killed as a result of being tasered by officers in the United States alone.

The ACED then did a little spelunking in the Times archive, and found:

When the NYT reported the findings of a BJS study on arrest-related deaths (an earlier study covering only 2003-2005), it also included the number of deaths of police officers over that period, and even pointed out that there were "174,760 assaults on law enforcement officers during the three-year period." Again, today's article about the killing of officers made no mention of the number of civilians killed by law enforcement or police brutality.

And speaking of trends, here's one the Times didn't touch (but that Reason did):

In 2010, the Los Angeles County Sheriff's Department (LASD) had a 50 percent increase in the number of "perception" shootings—shootings in which "deputies perceive, accurately or not, that a suspect may be armed or going for a gun" and then fire their own weapons. According to a report compiled for the department by the Police Assessment Resource Center, a non-profit group that analyzes local and state police agencies, "A little more than half of those suspects were holding an object such as a cell phone or sunglasses that was believed by deputies to be a possible firearm."

LASD deputies fired on 260 people from 2005 through 2010. Sixty-one percent of those shootings involved suspects who were later determined to be unarmed. "Waistband shootings," in which suspects were fired upon after reaching toward their waists, increased from one-fifth of all incidents to one-third. In almost half of those incidents, the suspects were found to be unarmed. Ninety-six percent of the suspects fired upon by sheriff's deputies were black or Latino.

And that's just one county.