Another issue is that a minority of violent crimes are even reported to the authorities. According to the National Crime Victimization Survey, between 40% and 50% are reported to the authorities.
Even taking that into account, when I took a look at the FBI UCR statistics for a recent article (https://borncurious.blog/i/152677807/violent-crime-rates-are-at-or-near-all-time-highs-in-most-major-american-cities-false), I found they were less than the 40 to 50% of violent crimes reported via the National Crime Victimization Survey. Thus, it appears that the number of crimes processed by police agencies and reported via the UCR are even less than the number of crimes reported to police. (It's also suspicious that there is a surge in crimes processed through the UCR every December, which makes me think police agencies are catching up on their reporting at the end of the year.)
Thus, crime statistics record a fraction (crimes processed) of a fraction (crimes reported to authorities) of actual crime. Because of this, there is a big gap between perception of crime and crime statistics.
I'm not familiar with the Real-Time Crime Index. Perhaps it's better than the FBI's UCR?
Good point. Even if the data is accurate, there are still issues, as I noted, but as you point out, crime is underreported, which creates more issues. Although homicides tend to be accurate for what it is worth.
The Real-Time Crime Index has the same underporting problem as they are still getting data from police reporting. Their value is that by sampling they are able to have data without the year-plus lag that FBI UCR has.
Another issue is that a minority of violent crimes are even reported to the authorities. According to the National Crime Victimization Survey, between 40% and 50% are reported to the authorities.
Even taking that into account, when I took a look at the FBI UCR statistics for a recent article (https://borncurious.blog/i/152677807/violent-crime-rates-are-at-or-near-all-time-highs-in-most-major-american-cities-false), I found they were less than the 40 to 50% of violent crimes reported via the National Crime Victimization Survey. Thus, it appears that the number of crimes processed by police agencies and reported via the UCR are even less than the number of crimes reported to police. (It's also suspicious that there is a surge in crimes processed through the UCR every December, which makes me think police agencies are catching up on their reporting at the end of the year.)
Thus, crime statistics record a fraction (crimes processed) of a fraction (crimes reported to authorities) of actual crime. Because of this, there is a big gap between perception of crime and crime statistics.
I'm not familiar with the Real-Time Crime Index. Perhaps it's better than the FBI's UCR?
Good point. Even if the data is accurate, there are still issues, as I noted, but as you point out, crime is underreported, which creates more issues. Although homicides tend to be accurate for what it is worth.
The Real-Time Crime Index has the same underporting problem as they are still getting data from police reporting. Their value is that by sampling they are able to have data without the year-plus lag that FBI UCR has.