68

I have read comments and articles (an example) that claim that blacks are responsible for more than 50% of all homicides in the United States of America. How credible is this claim?

  • Comments are not for extended discussion; this conversation has been moved to chat. – Sklivvz Apr 30 '16 at 20:16
  • 4
    Also -- some of the comments were borderline racist. Posting racist flamebait will get your account suspended very quickly. – Sklivvz Apr 30 '16 at 20:17
  • 4
    The US is a strange place. Who is black? The US seems to call anybody "black" who has just one ancestor who was African. They may also have 1, 3, 7, or 15 ancestors who were light-skinned, but still they count as "black". The statistics should take into account the fraction of "blackness" to have any meaning. – RedSonja May 4 '16 at 7:49
  • 3
    @RedSonja That was true at one time. Now federal government considers white as "A person having origins in any of the original peoples of Europe, the Middle East, or North Africa", and permits people to state they are two or more races. See the last page of the following ATF document for example: atf.gov/resource-center/docs/non-immigrant-alienspdf/download – DavePhD May 5 '16 at 14:00
  • One must bear in mind that crimes are prejudicially "hung" on racial minorities. See Miami Herald article from July 12, 2018: “If they have burglaries that are open cases that are not solved yet, if you see anybody black walking through our streets and they have somewhat of a record, arrest them so we can pin them for all the burglaries,” one cop, Anthony De La Torre, said in an internal probe ordered in 2014. “They were basically doing this to have a 100% clearance rate for the city." – Daniel R Hicks Jul 12 '18 at 21:02
81

Yes, according to the US Department of Justice document Homicide Trends in the United States, 1980-2008:

Table 7, Homicide type, by race, 1980-2008:

Homicide offenders were:

45.3% White

52.5% Black

2.2% Other

Felony Murder Offenders were:

38.4% White

59.9% Black

1.7% Other

The report makes the following statement about potential bias due to unknown information:

While many agencies report supplemental data on homicides, much of the data concerning offenders may not be reported because no suspects were identified or the agency chose not to report the information. The most significant problem in using SHR data to analyze offender characteristics is the sizable and growing number of homicides in the data file for which no offender information is reported. Ignoring these homicides with no offender information would understate calculated rates of offending by particular subgroups of the population, distort trends over time among these same subgroups, and bias observed patterns of offending to the extent that the rate of missing offender data is associated with offender characteristics.

To adjust for homicides with no offender information, a method for offender imputation was devised that uses available information about murder victims for which corresponding offender information was provided as well as those with missing offender information. Through this imputation algorithm, the demographic characteristics of unidentified offenders were inferred on the basis of similar homicide cases—similar in terms of the victim’s demographic profile, circumstances of the homicide such as felony or argument, location of the homicide (region and urban), gun involvement, and year of the offense— for which offender data were provided. In other words, unknown offender profiles were estimated based on the offender profiles in offender-known cases, matched on victim age, sex, and race; circumstances of the homicide; location of the homicide; gun involvement; and year. Offender-based estimates in this report were imputed using this procedure. Other estimates in this report were based on homicides with known attributes, unless otherwise indicated. An estimated 30.8% of homicides involved an unknown number of offenders. For these homicides, the offender imputation method conservatively assumed the number of offenders to be one, likely resulting in an undercount of the number of homicides involving multiple offenders.

All rates were calculated using the estimated number of homicide victims or offenders as the numerator and dividing by the U.S. resident population estimates for the appropriate groups or subgroups. This report used bridged-race population estimates developed by the National Center for Health Statistics and the U.S. Census Bureau.

The report also finds that people are overwhelmingly being killed by members of their own race:

84% of white victims were killed by whites
93% of black victims were killed by blacks


For 2013, according to the FBI's Expanded Homicide Data Table 3-"Murder Offenders by Age, Sex, Race, and Ethnicity 2013 ":

out of 14,132 offenders:

4,396 White

5,375 Black or African American

249 Other

4,112 Unknown

This gives approximately 54% of "Black or African American" known offenders.

A different table from the same source, gives a subset, specifically one-victim one-offender murders, of the above set:

2755 White

2698 Black or African American

168 Other

102 Unknown

(here "Unknown" means age and/or sex but not race is known, there are really many more unknown offenders)

For this subset, the figure is 48%.

Unlike the Justice Department figures, there is no attempt to correct the FBI data for the bias of the unknown fraction of offenders, which is about 1/3 of offenders.

  • 6
    This assumes a perfect justice system. – Strawberry Nov 3 '17 at 15:36
  • 1
    Your math is skewed - 5,375 out of 14,132 is 38% - Omitting the unknowns would only be valid if you had strong evidence that the Unknowns would be distributed about the same... – Daniel Jul 12 '18 at 13:07
  • @Daniel Yes, I agree. Only the information above the horizontal line considers distribution of unknowns. – DavePhD Jul 12 '18 at 13:28
  • Through this imputation algorithm, the demographic characteristics of unidentified offenders were inferred on the basis of similar homicide cases Given that murder statistics in the US are quite vague, this statement does not give me any confidence in those numbers at all (even through they are probably the best you can get) – David M Jul 13 '18 at 4:33

You must log in to answer this question.

protected by Sklivvz Apr 29 '16 at 18:37

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

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