# Are one-third of Americans 'killed by strangers' killed by police?

According to Granta - Violence in Blue,

One-third of all Americans killed by strangers are killed by police.

Is this true?

• Two factors I see immediately that will make this difficult to really pin down: The relationship between murder victims and offenders was unknown in 49.5 percent of murder and nonnegligent manslaughter incidents, so "all Americans killed by strangers" is a base with a huge uncertainty. And then, "killed by police" would include justified incidents as well as unjustified ones... Commented Jun 3, 2020 at 10:02
• The title is misleading: a third of a quarter is one twelfth. The link explains how the statistic was derived, and why it is an educated guess, answering the question. To dispute it, someone would have to make another computation which would also have to be based on incomplete or overlapping information. Actually although the statistic looks bad – implying the police must be bad – I would prefer that more people are lawfully killed by police than by non-police. In fact, I don't want any people to be killed by non-police. So my controversial take is that the statistic is not high enough. Commented Jun 3, 2020 at 11:14
• @WeatherVane: Well, I don't want any people killed by police either... Commented Jun 3, 2020 at 11:40
• @DevSolar both would be ideal. I was just playing the devil's advocate. Statistics can be used to prove or stir anything, such as "just look at how bad the police are." Commented Jun 3, 2020 at 11:49
• @gnasher729 Yes it would.
– Simd
Commented Jun 4, 2020 at 13:04

To a reasonable approximation this appears to be true, provided we restrict ourselves to deliberate killings, although one quarter may be a more accurate figure.

The total number of murders in the US for 2018 is 15,498, according to the FBI, and this roughly agrees with data from other sources.

The US Department of Justice says that approximately 80% of murder victims know their killers, meaning that around 3100 people were murdered by strangers.

Recent statistics for police killings are harder to come by, but various sources indicate a rate of around 1100 for the years 2015 and 2016, and no indication that this would have changed significantly in 2018. Statista gives 967 people shot to death by police in 2018, but excludes those killed in other ways (George Floyd would not be included). It includes all killings, not just murders.

Taking 1100 police killings and 3100 people murdered by strangers, then we can conclude that the number of people killed by the police is about one third of the number of people murdered by strangers. We are neglecting here any people killed by strangers that are not murders, such as gun or vehicle accidents or self-defense killings. This is not exactly the claim being made, but it is probably where the figure comes from.

A better calculation would be to consider not just the number of stranger murders but also the number of justifiable homicides (e.g. self-defense killings). The police killings already include justifiable homicides, and there are around 300 justifiable homicides by private citizens a year giving us about 4600 deliberate killings by strangers, of which just under one quarter were killed by police.

Note that most police killings are considered justified, which means that far less than a quarter of people murdered by strangers were murdered by police.

For comparison the same calculation for the UK gives a figure of about 1 in 200.

• Comments are not for extended discussion; this conversation has been moved to chat.
– Jamiec
Commented Jun 5, 2020 at 10:06
• The Statista page does not distinguish between homicide and murder. It appears to treat the words as synonyms. I suspect that its headline is incorrect, and that it reports homicide rates rather than murder rates. It follows that adding justifiable homicides to the Statista numbers is incorrect. Even if the Statista figures truly are murder rates, this answer assumes that the 300 justifiable homicides are all "by strangers," for which there appears to be no justification. Commented Feb 3, 2021 at 14:29
• Doesn't change if I use FBI statistics. Commented Feb 3, 2021 at 14:43
• Why do you assume that the justifiable homicides were killing of strangers? For that matter, why do you assume that the police killings have no overlap with the reported murder victims. There are certainly cases where people who were police committed homicide. Commented Feb 3, 2021 at 17:57

this appears to be false as written. police only kill about 1k people per year, whereas car crashes kill somewhere around 30k people a year in the usa. 53% of car crashes involve more than 1 vehicle. if we assume that half of the car crashes victims in those are "strangers" to the driver who caused the crash. that would put the ratio around 8:1 without even counting other causes of death (e.g. crime). if we were to rephrase the claim as "one third of people who are shot to death by strangers are shot to death by the police", then that might be close to the truth.

• Comments are not for extended discussion; this conversation has been moved to chat.
– Jamiec
Commented Jun 8, 2020 at 9:07
• Death in an automobile accident is almost never homicide. Commented Feb 3, 2021 at 14:32
• @phoog as is death by police shooting. at least according to the court system. Commented Feb 5, 2021 at 4:19
• No, death by police shooting is always homicide. It's usually not a crime, however. Commented Feb 5, 2021 at 4:44
• I'm talking about the legal distinction between murder and homicide, which is what matters here. There are many forms of homicide that are not murder, some of which are lawful and others are not. Examples include justifiable homicide (e.g. self defense), judicial homicide (i.e., execution of a death sentence imposed by a court), negligent homicide (i.e., killing through carelessness). The last is unlawful, but not murder. The first two are lawful but still homicide. Commented Feb 25, 2021 at 9:45