The problem is, what is you're definition of often? And how effective is a (likely holstered handgun) in these situations?
Is an armed officer more or less likely to be shot at? Does the presence of an armed officer in all routine traffic stop increase safety overall (not just for the officer, but for non-violent civilians as well).
If you look at it from a cost benefit analysis, giving all traffic cops firearms may result in a small decrease in total deaths of officers during traffic stops [1], but a small increase in the total number of deaths of civilians (both innocent bystanders, and belligerent civilians who had no intention to seriously harm anyone).
That's probably not enough to justify arming all traffic cops. However if you add in the benefit of a firearm as a tool of compliance to aid in traffic cop's abilities to confiscate drugs and cash, you tip the balance towards arming them.
1. Even this is a big assumption, because we don't know if criminals are more or less likely to fire at armed vs. unarmed traffic cops.
46.3% of federal prisoners (nearly 85,000) are imprisoned for drug offenses -- not quite an absolute majority, but nearly triple the next highest category (Weapons, Explosives, and Arson.) [0]
Fairly recently, it was only 17% for state prisons [1], which is still quite significant -- about 1 in 6.
But both of those understate the degree to which the drug war drives imprisonment, because drug crimes (which tend to have a shorter sentence than violent crimes) account for nearly 1/3 of prison admissions, despite accounting for a lower share of total prison population. [1]
[0] https://www.bop.gov/about/statistics/statistics_inmate_offen...
[1] http://www.brookings.edu/blogs/social-mobility-memos/posts/2...
The contention upthread was that "Contrary to popular perception, the drug war is not a major factor in incarceration in the US."
Drug offenses alone (and the drug war produces, and includes imprisonment for, other, non-drug offenses) account for about 1/6 of the total prison population and 1/3 of total prison admissions (sources addressed in my previous post in this subthread.) Ending the drug war -- even if you consider only drug offenses, would eliminate a significant share of the total prison population, and a much larger share of total prison admissions.
The facts clearly show that, while drug offenses alone are not the majority of incarceration in the US (either by population or admissions), the Drug War is a very significant contributor to incarceration in the US.
Which is an interesting factoid, but irrelevant to the issue of whether or not the drug war is a significant factor in incarceration. That 70% of the increase from 1995 to 2013 is from that source (which, while it isn't specific to drug offenses, overlaps, rather than being disjoint from, the drug war) is almost entirely irrelevant to the issue of whether the drug war is a significant in the overall level (whether stock or flow) of incarceration, which it is and remains. (Basically, the main change is longer sentencing, which means simply increases the "stock" contribution of violent/property crimes to incarceration, without doing much to the "flow".)
> but we should not delude ourselves that the solution for police violence or violent crime will come from changing the rules about drugs.
Since no one has suggested that the solution to either of those would come from that source alone, that -- while true -- would seem to be a complete strawman.