We're the Only Ones More Highly Rated Enough

So putting the police numbers at an annual rate gives you a rate of 0.01%. Both 0.01% or 0.0003% are both extremely low and the violations might not be comparable in that the private individuals might run into problems that a police officer (even one off duty might not run into), but the rate for police is still 23 times higher. [More]

I wonder if there's a statistically valid way to extrapolate incidents that are never prosecuted in the first place because of institutional interests in keeping a lid on things.