“We have a system currently that is almost entirely reactive, a system influenced by anecdote and emotion.... The beautiful thing about numbers is that they don’t lie.”
This was what California Attorney General Kamala Harris said about her OpenJustice data initiative as quoted by Washington Post reporters Aaron C. Davis and Wesley Lowrey in an article about the lack of data on police shootings. Kamala Harris was pointing out that most people get most of their information about police shootings from isolated incidents that make the national news and spark community outrage. The attorney general and candidate for U.S. Senate was touting her new statewide initiative to make public a statewide repository of criminal justice statistics with a fancy new website (openjustice.doj.ca.gov).
But this is 2016. Thanks to crowdsourced databases like Fatal Encounters and killedbypolice.net, and to the journalistic efforts from the Guardian and the Washington Post that use those sites to scrape data, we have a lot of transparency now, or at least more than we used to. Every officer involved shooting that makes the local news is now compiled into one of these databases, even though the majority of the incidents are non-controversial. Thanks in part to these efforts from the Guardian and the Washington Post, it is now known that over a thousand people get killed by police in the United States every year, that black people get killed at a rate 2 to 3 times that of the rate that white people get killed, and that a quarter of those killed are mentally ill, for instance.
The California OpenJustice data initiative does not contain the names of the deceased, the names of the officers who killed the deceased, or the circumstances of the killing. It does have identifying data like race, gender and age, and it does apply some statistical analysis to its limited dataset including arrests data, something hard to find from other sources. Its graphs are beautiful. But its depth is a little shallow. The OpenJustice data initiative can't tell us much about the precise moment when officers fired their weapons.
But none of these others datasets can either. The Guardian looked at weapons held by the decedents, but they didn't distinguish if the decedent had, for instance, fired a gun, point a gun, brandished a gun, or ran away from police with a gun. The Washington Post goes one step further and examines whether the officer was being attacked or not. But this doesn't quite get to the imminent threat that the officer may or may not have faced, and it is hard to second guess an officer if all you know is that the decedent had a gun.
This is why this database exists. I have researched every incident of police lethal force from January 1, 2014 through June 30, 2015 (for now). I have tried to include as much data from as many sources as possible for each incident. The data mostly comes from news reports and reports from district attorneys to determine exactly why the officer felt like he (usually a he) needed to use lethal force.
Datacards for each incident can be found here:
lethaldb.silk.co
It will be slow to load because there are so many incidents.
I wanted to focus on the split-second decision made by the officer and try to make sense of the reason that so many unarmed non-threatening people get killed by police. Because it is true what Kamala Harris says: the beautiful thing about numbers is that they don't lie.
