I work in law enforcement intelligence, and my team frequently creates policy papers, reports, and so forth that focus on crime trends. It is often tempting to use Unified Crime Reporting (UCR) statistics to make a point. Our policy is to only use that source if no other information is available, due to lack of reliability.
UCR data is collected from local, state, and federal law enforcement agency reporting. We find considerable variations in reporting standards, timing, and reliability. Different jurisdictions categorize crimes differently, and this variation causes serious doubts in the validity of UCR data.
Researchers and your readers should take studies using UCR data with a grain of salt. – ARM
o o o
The data from the UCR has a major skew resulting in the first of several downward movements resulting in changes to how crimes are defined and coded. In 1993, the DOJ changed the definitions of just about every crime, and the results begin showing in 1994. How can I know this? There are several ways; the easiest non-inside police circle secret squirrel way is to go to the UCR table tool, enter a state and all crimes, and look at the trend approaching 1994, then loom at the data for 1994. Do this for the same state for discrete crimes, such as murder, and you will see a huge decrease from 1993 to 1994.
Those stats are raw and not based on per capita. So with some analysis, the truth behind the data shows just about all crime rates are up for the last 30 years, not down. Run the per capita analysis and then glance at U.S. Census population figures for counties, and you see for yourself.
Here is an example. Texas is booming and has been for close to 20 years. Get the population for a county (Dallas County) and then take the same number of murders from the UCR for the same year, do some math to get a per capita (number of murders divided by 100,000, then multiplied up based on the population in increments of 10000). This takes 3-5 minutes per year per county. In every case you will see crime going up and up, and then as demonstrated in 1994 data, several hard corrections down, with no attributable reason.
I know the old saying “figures lie and liars figure”, but in this case some simple math will show anyone that crime rates are going up higher than per capita… except in gun rights friendly areas. – GJS