Thread Reader
Katrina S.

Katrina S.

Nov 20
18 tweets

This is a thread about the demographics of murderers and murder victims, using the recently released FBI NIBRS 2021 data Main things to keep in mind: - "White" to the FBI includes Hispanic - There's insufficient data for all other races

This includes murderers by race split into age-ranges Sadly the ranges NIBRS provides are quite wide
27.7% of White murder victims are killed by non-White offenders 9.7% of Black murder victims are killed by non-Black offenders
Nearly half of all murders happen at night, between 8pm and 4am
There are a few ways to estimate the demographics of murderers Beware: this likely underestimates Black murderers as unsolved cases are disproportionately of Black Victims e.g.: 673 White Victims have murderers of unknown race 1913 Black Victims have murderers of unknown race
By combining the Arrest data with US Census population data, we can get an estimate of the rate of committing murder across ages Ex: - [1 out of every 13900] [White Males over 15y/o] commit murder in 2021 - At these rates, we'd expect 1/220 WMs to commit murder in his lifetime
One interesting point, we see that the Male offender data is very consistent between the two main demographic sets in NIBRS (arrestee & victim) The Female data is less robust, likely due to the smaller numbers involved (NIBRS removes some entries below a threshold for privacy)
Alternatively, you can compare rates against some baseline This chart includes comparisons against: - "Average American Over 15yo" - "White Female of the same age-range group" Here, we see [Elderly White Females] commit murder at 0.03x the rate of the [Average American]
Inter-/intra-racial murders by victims sex Female portion of murder victims: 29.0% of White-on-White 25.3% of Black-on-White 17.0% of Black-on-Black 10.7% of White-on-Black In actual numbers: 25.2% of White Females killed are by Blacks 6.8% of Black Females killed are by Whites
Note: all data in this thread comes from the Oct 31 2022 NIBRS Estimation data Specifically from filename "Indicator_Tables_1.csv" which covers national-level data Available from FBI Crime Data Explorer
Looking at the ages of murderers by race, and the ages of the people they choose to murder (Note: this is categorized by the "Murderer's Race" in both cases: Top is "Murderer's Age" Bottom is "Victim's Age")
TIL: "Children under the age of 14" is the only category of Murder Victim where Male & Female Murderers are all within the same ballpark
Separating out Hispanic ethnicity from the White data is difficult, especially when you drill down At the top level, for Ages 15+: - Hispanics are ~21% of Whites - And commit ~31% of Whites murders
White & Hispanic Males over 15 y/o are together 31.7% of the US population and commit 30.1% of the murders in 2021 Black Males over 15 y/o are 4.9% of the US population and commit 56.7% of the murders in 2021
One useful way to think about these single year murder rates is to project them across a hypothetical "lifetime" At 2021 rates, you would expect about 2% of all Black Males (1 in 50) to have committed murder by the time they reach 25 years old
This is the per capita murder rates normalized per 100,000 population. Using this data we see: - 0.0810% of White Females are projected to commit murder in their lifetime - 0.0825% of Black Males (15+) commit murder last year
Due to missing data, there's a lot of uncertainty in the Hispanic numbers All previous tweets have used the most conservative possible assumptions for Hisp (2nd chart in this tweet) 1st chart here uses more realistic demographic assumptions, and is likely closer to reality
These are the 2021 murder rates by sex & race per 100,000 population Assuming these rates hold, we can project that: - 0.274% of White Males - 4.508% of Black Males - 0.174% of Hisp Females will commit murder in their lifetime (Includes both ests for Hisp, as prev described)
Katrina S.

Katrina S.

Just a Jewish yogi who loves politics and fun 🥰
Follow on Twitter
Missing some tweets in this thread? Or failed to load images or videos? You can try to .