Oh, no, you're atually a little off base on that...
The number of LEOs "on the beat" actually DO have a measurable correlation to violent crime. The more cops, the more crime...
There are nearly 5 times as many sworn LEOs in the USA today than there were in 1950. The population has doubled in that same time. The violent crime rate has nearly tripled. These are national averages...
Regionally, however, the majority of the nation has seen significant decreases in violent crime since 2007. In the regions with major decreases in violent crime, there have also been tremendous INCREASES in the issuance of CC permits.
However, cities and states that are not "shall issue" or otherwise restrict or prohibit carry have seen increases in violent crime that are practically a mirror image increase to the decreases in carry-friendly areas.
The increases in violent crime in "victim disarmament zones" has, in fact, more than offset the decreases in violent crime in carry-friendly jurisdictions. These "victim disarmament zones" have also had higher-than-average increases in LEO numbers.
You do the math...
More guns in the hands of LACs = less crime....
But more cops + fewer guns in the hands of LACs = WAY more crime...
And of course, in areas that have seen above-average in LE personnel, there has also been an above-average increase in police brutality, civil rights violation cases, and excessive force lawsuits. Unfortunately, these cases are usually not included in the Federal UCR as "violent crime" because most of these police-committed instances of violence do not result in criminal charges being files against the offending officers--most end in civil suits or are completely thrown out and the officers "get a pass". So actually, the violent crime statistics for carry-unfriendly jurisdictions are lower than the reality of the situation, because they don't count the thousands of annual incidents of bad shootings, beat-downs, excessive force, false arrest (kidnapping), and general abuse of power by LEOs...
Of course, we must remember that correlation DOES NOT equal causation. Statistics are like poodles. If you treat them nice enough, you can make them jump through flaming hoops...