> after exploding in 2020-21, where there was the largest increase in the entire time series.
The IMGUR link includes the text:
Due to the full transition to NIRBS and the lack of data for agencies that are not fully transitioned, the 2021 data year cannot be added to the 5-, 10- or 20- year trend presentations that are based in traditional methodologies used
That's a giant flashing warning sign to data nerds everywhere that some change in the gathering | reporting of data took place exactly right when things "exploded".
The question remains, did crime (actual homicide numbers) really explode or did national reporting of homicide numbers suddenly improve | include backlogs | etc?
> We are basically back to 1995 homicide levels ..
These are absolute numbers .. and the US population has grown in 25 years, given the change in national reporting the per capita homicides rates may be more stable than the graph indicates.
The IMGUR link includes the text:
That's a giant flashing warning sign to data nerds everywhere that some change in the gathering | reporting of data took place exactly right when things "exploded".The question remains, did crime (actual homicide numbers) really explode or did national reporting of homicide numbers suddenly improve | include backlogs | etc?
> We are basically back to 1995 homicide levels ..
These are absolute numbers .. and the US population has grown in 25 years, given the change in national reporting the per capita homicides rates may be more stable than the graph indicates.
https://en.wikipedia.org/wiki/National_Incident-Based_Report...