Statistics can be damning.
They can equally be misleading, more so when the parameters of gathering the statistics are not clear.
With the aim of clarifying numbers and further understanding results, I crunched some numbers.
Two separate news organizations (from two countries), the Washington Post and The Guardian, are running concurrent tallies on deaths caused by law enforcement in the United States. Their results are slightly different--the Post's current tally (as of August 18th, 2015) is 616, while The Guardian's tally as of the same date sits over 100 people higher, at 729. Some of these discrepancies may be due to methods of categorization and preliminary assertions. Additionally, although both list the deaths in subcategories of gender, race, age, and cause, The Guardian stats are able to be sorted according to subsets such as race and whether the victim was armed vs unharmed, while the Washington Post does not currently have that feature.
Using the data mined by these sites (who have gleaned their data via online resources, police records, and reporting), I have crunched some of the numbers further to come out with percentages that give a more succinct summary. Separately, I researched law enforcement deaths and have included them below to see at what rates police officers being killed and if the stats of police officer deaths warrant the reactions of officers in taking down armed and unarmed suspects during confrontation.
From a logical imperative it is important to note two things: simply because a victim is armed does not inherently mean he/she is a threat; additionally, to be unarmed does not indicate innocence. However, being armed does increase the odds of an escalating violence, and to be unarmed and facing an armed police officer is a tactical disadvantage that deserves a de-escalation of reaction on the part of the officer in question. This is where the statistics of armed vs unarmed deaths by police can become so persuasive: they speak to the reactions of police officers and the level to which they get it wrong.
To further understand these statistics, they would have to be taken as a percentage of a whole. These are strictly deaths as measured against all deaths. They are not deaths as measured against all police responses. IF, for fictitious example, 0.02% of all police responses to whites resulted in drawn weapons, whereas 0.05% of all police responses to blacks resulted in drawn weapons, we would have a more clear picture of reactions on the basis of the whole. As it stands, these statistics are isolated against themselves: they are deaths as a percentage of all deaths total, not to all responses total. This still leaves gaps in an understanding of bias, but we can still glean some understanding of the statistics as they stand.
Percentage of total killed by U.S. law enforcement, 2015 to date:
Via the Washington Post (616 total):
48% White; .00015% of population
26% Black; .00042% of population
16% Hispanic; .0002% of population
(Asian specifics not given)
Via The Guardian (729 total):
49% White; .00017% of population
26% Black; .0005% of population
15% Hispanic/Latino; .0002% of population
2% Asian; .00009% of population
(Population totals calculated against most recent census data)
Armed vs unarmed deaths by law enforcement, via The Guardian:
White:
283 -- Armed
63 -- Unarmed
Total: 346
18% of whites killed by police are unarmed
Black:
127 -- Armed
58 -- Unarmed
Total: 190
30% of blacks killed by police are unarmed
Hispanic/Latino:
84 -- Armed
20 -- Unarmed
Total: 107
18% of Hispanics killed by police are unarmed
Asian/Pacific Islander:
11 -- Armed
3 -- Unarmed
Total: 14
21% of Asians killed by police are unarmed