American police kill civilians at a shocking rate compared to other developed countries

It's official: 25-year-old Freddie Gray's death was a homicide, according to Baltimore State Attorney Marilyn Mosby. Criminal charges will now be brought against the six police officers involved in Gray's arrest.

Gray was arrested on April 12 and lost consciousness in the back of a police van. When he got to the hospital, he was in a coma, his spinal cord was severed at the neck, and his voice box was crushed. He died a week later. Video of the arrest showed police putting Gray into the van. He was conscious and handcuffed. He also seemed to be in pain. Police didn't immediately explain how Gray suffered the spinal cord injury that killed him.

More from GlobalPost: North Korea, Venezuela and Russia are having a field day with Baltimore

A furious Baltimore community had had enough. They came together to demand justice for Freddie Gray and to continue the now months-long national protests against police killings of people of color, including Michael Brown, Eric Garner, and Walter Scott. They protested. Some rioted.

US police brutality against civilians is in the spotlight once more, and the world is watching.

There's so much we don't know about why and how often people die during police encounters, and yet, we know enough to know that American police are killing far, far too many people compared to police in other developed countries.

While no US government agency is keeping reliable records of how many people die each year during encounters with police, the best estimates suggest the number is no lower than 400 per year, and most likely around 1,000. Police in peer nations like Germany, Denmark, the UK and other liberal democracies — meanwhile — rarely kill civilians. Even accounting for population size, the frequency with which American police kill civilians is shocking. Not twice as often, or three times as often. We're talking factors of 20 to 70.

How many killed?

If the deaths of Michael Brown, Eric Garner, and Freddie Gray prompted you at any point to Google something like, “How many people do American police kill every year,” then you probably know that we don’t know.

More from GlobalPost: What people around the world are saying about Ferguson

You’d think state and federal officials would keep better track of something as serious as a law enforcement agent killing a civilian. You’d think agencies like the Department of Justice would publish annual reports, chart trends over time, map where police are killing civilians with higher frequency, and identify who’s dying and under what circumstances. This kind of information would be essential for training the nation's police more effectively, reforming departments that kill people in disproportionate numbers, and reassuring an increasingly suspicious and even hostile public that police are being held accountable. Well they don't keep those records, and in failing to keep them, they produce the exact opposite effects.

Trying to get facts about police killings means venturing into an internet wormhole. So let’s start with what data the government does provide.

The FBI provides information about what it calls “justifiable” homicides committed by police officers. Their data runs from 2009 to 2013. According to the data, police commit a little over 400 justifiable homicides per year. In 2013, they killed 461 people. So you’ll see the 400 number a lot when you try to figure out how many people police are killing.

But there are reasons to be skeptical of that number. Most obviously, it doesn’t include information on police homicides that were found to be unjustifiable. So the unjustifiable police shooting of Walter Scott on April 4 in South Carolina wouldn’t count. There are also questions about transparency, record keeping, and self-reporting by local police departments.

Another concern: The FBI’s justifiable homicides are almost all shootings. In 2013, 458 people were killed by police firearms. Three people were killed by “other dangerous weapons.”

It’s not clear whether the FBI is tracking deaths that result from police use of physical force during arrest, restraint, and processing at jail. That means the numbers likely wouldn’t include Freddie Gray’s or Eric Garner’s deaths.

You can get a different set of answers from the Department of Justice’s Bureau of Justice Statistics (BJS) data on “arrest-related” deaths. Unfortunately, it’s a set of answers BJS admits is so unreliable that it stopped collecting the data in March 2014.

Take it for what it is, but here’s what they tell us: Between 2003 and 2009, 4,813 people died during or shortly after arrest. About 60 percent (2,931) were deemed homicides.

Again, there are problems with these numbers.

First of all, after a self-audit, BJS decided that its collection method, which, like the FBI data, relied on self-reporting by local police departments, wasn’t producing reliable enough data.

Second, it’s not fully clear when you stop calling a person’s death “arrested-related” and start calling it a death “in custody.” BJS keeps those records, too. And once again, the data makes it impossible to know how many people police are killing every year.

BJS doesn’t explain, for example, whether it’s a custodial death or arrest-related death if a police officer asphyxiates a person while restraining them in short-term lock-up. And it doesn’t distinguish between homicides committed by police or correctional staff and homicides committed by fellow inmates. It lists just a single number for “homicides” in all state prisons and local jails.

That means there is no way to know whether a person was killed by police shortly after arrest while awaiting arraignment. Or whether a person was killed by another person in detention. Or whether a person convicted and serving their sentence was killed by a corrections officer.

We need to know these things. That we don’t is sketchy enough for BJS to have a FAQ about it on the website:

Are homicide counts limited to homicides committed by other inmates or facility staff?

No. The total homicide count includes all types of intentional homicide and involuntary manslaughter as ruled by a medical examiner or other official medical investigation. The total includes homicides committed by inmates. It also includes homicides as a result of staff use of force, such as positional asphyxia, or suffocation caused by the position of the inmate's body, while the inmate is being removed from a cell. It includes legal-intervention homicides (e.g., an inmate is shot in the process of escape).

For now, the closest we can get to a real answer about police killings comes from watchdogs and data junkies. Mapping Police Violence, an independent monitoring group, mined news reports and found that police killed at least 1,149 people in 2014. Nate Silver’s Five Thirty Eight also estimates that police killings could be somewhere around 1,000 per year.

What about other countries?

To make sense of these numbers, you need to know just how rarely police in other wealthy liberal democracies kill civilians.

Even when you account for population size, gun ownership, and violent crime rates, American police out-kill civilians by staggering factors relative to peer nations.

Take a look at England and Wales, where the Independent Police Complaints Commission (IPCC) keeps yearly records of police homicides. (Scotland and Northern Ireland have different police monitoring agencies.)

According to the IPCC, there hasn’t been a fatal police shooting in more than two years and over the same period only 26 people have died in police custody during arrest or in post-arrest detention. In the period from 2010-2014, a total of 62 people died in police custody.

Compare the police record from 2010-2014 to Baltimore’s police record, which we know thanks to an investigation by the ACLU. In Baltimore alone, 31 people died in police custody, exactly half the number who died in all of England and Wales.

England and Wales have a combined population of 56 million. Baltimore’s population is 622,000. Do the math and you realize that people in Baltimore were about 50 times more likely to be killed by police. And the disparity is even larger when you factor in shooting deaths. The Economist estimates that British citizens are 100 times less likely to be shot by police than Americans.

The story is the much the same in Germany, a country of 80 million people, where about 100 people have been killed by police since 1998.

In Iceland, police have only killed one person since the republic was founded in 1944. Granted, the population of Iceland is just 323,000. But it’s still impressive.

And here are some other comparisons published in a study by the Danish police on firearm use between 1996 and 2006.

They identified the number of people killed by police in several European countries and accounted for population by giving the number of people killed per one million residents. They used population figures from the year 2000.

Here’s what they found:

• Denmark: number of people killed by police between 1996 and 2006: 11 people — number of people killed per one million residents: .187 people  
• Sweden: 13 people — .133 people
• Norway: 3 people — .060 people
• Finland: 2 people — .034 people
• Germany: 81 people — .089 people
• The Netherlands: 24 people — .137 people
• England/Wales: 25 people — .042 people

If we take Five Thirty Eight’s estimate that 1,000 people are killed by police in the United States every year and divide it by the 2000 population of 282 million, the American situation for just one year would look like this:

US: 10,000 people killed — 35.5 killed for every one million residents

Are you comfortable with that?

Defenders of American police will often explain the disparity on the prevalence of guns in the United States. More people with guns and more gun crime means American police might be quicker to use deadly force. But if you compare per capita gun to per capita police homicides, that theory doesn't hold up.

The US has two to three times as many guns per capita as places like Denmark, Germany, Finland, Switzerland, and. But the number of police homicides per capita is staggeringly high. Twice as many guns means 70 times as many people killed by police?  

We clearly need to figure out why this is happening and what will stop it.

A start would be keeping records that it's happening at all.

Correction: Some of the numbers used to describe the Danish police study in an earlier version of this story were incorrect. The figures have been updated.