According to https://skeptics.stackexchange.com/a/47814/25402 1 in 2000 men's cause of death is police violence while the corresponding number for women is 1 in 33 000.

Risk is highest for black men, who (at current levels of risk) face about a 1 in 1000 chance of being killed by police over the life course. The average lifetime odds of being killed by police are about 1 in 2000 for men and about 1 in 33000 for women.

Is that difference proportional? Naïvely, do men commit 16.5 times as many crimes as women? Less naïvely, do men commit 16.5 times as many crimes of types that motivate the police to use lethal violence?

With the latter I mean that one could motivate that it is reasonable that e.g., armed robbery or school shootings is met with more violence than e.g. shoplifting or speeding on a freeway.

  • 10
    When the police interacts with people, those people are usually only suspected of committing crimes or misdemeanors. A person isn't considered a criminal until a court determined that the person is guilty , and from that point on that person will usually interact with the corrections system, not the police system.
    – Philipp
    Jun 15, 2020 at 8:57
  • 4
    The type of suspected crime is not what determines (or at least it is not what should determine) how the arrest happens. No matter how horrible the charges are, the police is only to put you into custody, not to decide or administer punishment, and proportionate force is allowed as an answer to resistence/violence to arrest (in particular, lethal force is usually considered valid only for self-defense or defense of other people).
    – SJuan76
    Jun 15, 2020 at 11:42
  • 3
    Note that, depending on the type of crime, the probability of the suspected crime resulting in a police interaction may vary from close to 0 to close to 1. If you're interested in all crime, you'd have to come up with other methods (such as surveys) to estimate total crime vs. reported crime vs. crime with a police interaction.
    – gerrit
    Jun 15, 2020 at 12:12
  • 5
    This is a highly misleading question. First, you have to separate out the numbers killed by police because they were actually committing crimes from the number that were just being harassed, or simply in the way, e.g. en.wikipedia.org/wiki/Shooting_of_Justine_Damond Second, police (being mostly male) are likely to harass women differently from men. You might compare the male/female ratio of people coerced into sex by police.
    – jamesqf
    Jun 15, 2020 at 16:22
  • 5
    @jamesqf The question is perfectly reasonable if we're gathering information on that harassment. Jun 15, 2020 at 18:41

2 Answers 2


The FBI Uniform Crime Reporting (UCR) Program, the latest year for which statistics are available at time of writing being 2018, gives a breakdown by sex for arrests made during the year.

In particular, for arrests in general, 5,684,385 males were arrested, compared to 2,126,700 females. This represents a ratio of about 2.67 to 1.

With respect to 'crimes of types that motivate the police to use lethal violence', this is a little harder to broadly define, as clearly every arrest is different, but if we use the FBI's definition of violent crime - "offenses of murder and nonnegligent manslaughter, rape, robbery, and aggravated assault" - the figures are 313,411 males arrested compared to 82,854 females, a ratio of about 3.78 to 1.

The only crimes which had a ratio of male arrests to female arrests greater than 5 to 1 were Robbery (5.62), Murder and nonnegligent manslaughter (7.19), Weapons; carrying, possessing, etc. (9.71), Sex offenses (except rape and prostitution) (13.41), and Rape (30.51).

The only crime for which more women were arrested than men was Prostitution and commercialized vice, with 80% more women (15,243) arrested than men (8,456).

Note that arrests are obviously not equivalent to actually committing or being found guilty of the offense, but are probably a better metric of interaction with police.

  • 3
    This answer has the major unstated assumption that arrests are proportional to crimes committed, which is certainly not the case. Not only are many arrested people innocent, there are also large numbers of crimes for which nobody ever gets arrested.
    – gerrit
    Jun 15, 2020 at 12:07
  • 11
    @gerrit see the last sentence
    – CDJB
    Jun 15, 2020 at 12:07
  • 9
    @gerrit arrests would seem to me to be the best indicator of interaction with police, however. Perhaps my 'and', should be an 'or' though.
    – CDJB
    Jun 15, 2020 at 12:09
  • 3
    True, arrests would seem a reasonable indicator of interaction with the police, in particular such interactions which may turn violent. It's rather the question that is somewhat poorly phrased here.
    – gerrit
    Jun 15, 2020 at 12:10
  • 3
    @DarrelHoffman According to the FBI, the statistics are collated from "law enforcement agencies submitting 12 months of arrest data for 2018". Presumably, this then includes occasions where suspects were arrested and then died in custody, but not those who were killed in an arrest attempt. The annual number of people killed by US police seems to waver around 1,000, however, so although a huge number in real terms, doesn't make much of a dent in the statistics above.
    – CDJB
    Jun 15, 2020 at 15:29

A 2018 report from the Department of Justice outlines "Contacts Between Police and the Public, 2015" (published 2018). The report notes

Overall, a higher percentage of males (22%) than females (20%) experienced police contact (table 1). This was driven by a higher percentage of police-initiated contacts among males (12%) than females (9%)

Table 1:

Demographic       U.S. population     Any contact        Police-initiated    Resident-initiated
characteristic     age 16 or older                           contact             contact
Male*                  122,968,400 | 27,038,300  22.0%  | 15,339,100  12.5%  | 12,537,300  10.2%
Female                 130,619,000 | 26,431,000  20.2 † | 12,076,800   9.2 † | 14,523,000  11.1 †

*Comparison group.
†Signifcant diference from comparison group at the 95% confdence interval. 

Arrest counts are not proportional to police interaction. Using the arrest counts for the same year (2015) by gender (male, female, both summary ) shows 6,067,584 male arrests and 2,238,335 female arrests.

To answer the question

Is the death rate due to police violence between men and women proportional?

Then, the incidence of police interaction by gender is not proportional to

The average lifetime odds of being killed by police are about 1 in 2000 for men and about 1 in 33000 for women.

Additional considerations:

  • Police bias based on gender is not considered
  • The above table shows annual police contact, it is not a direct comparison to an individual's "lifetime odds" as police contact may vary with age
  • 1
    Off topic but related study: prisonpolicy.org/blog/2019/05/14/policingwomen , notes "Women make up an increasingly large share of arrests" 27% up from 16% in 1980. And "Use of force: Nearly doubled for men since 1999, but more than quadrupled among women". And "Street stops: Significant racial disparities among men, but not women" but notes "Arrests during stops: Significant racial disparities among women, but not men". And finally "Use of force during stops: Rates of Black women similar to white men; Black and Latino men most likely to experience force"
    – BurnsBA
    Jun 17, 2020 at 21:27

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .