There's probably no realistic way to find out, but I'd be curious how the numbers would change if it were for how many people as distinct individuals have ever been arrested (whether just once or multiple times)?
Here is the link to the report the article is based on. I did not go through the table (
FBI — Persons Arrested)
It is worth noting that their "arrests" include being summonsed to court when one is not technically arrested.
The overview is:
Nationwide, law enforcement made an estimated 12,408,899 arrests in 2011. Of these arrests, 534,704 were for violent crimes, and 1,639,883 were for property crimes. (Note: the UCR Program does not collect data on citations for traffic violations.)
The highest number of arrests were for drug abuse violations (estimated at 1,531,251 arrests), larceny-theft (estimated at 1,264,986), and driving under the influence (estimated at 1,215,077).
The estimated arrest rate for the United States in 2011 was 3,991.1 arrests per 100,000 inhabitants. The arrest rate for violent crime (including murder and nonnegligent manslaughter, forcible rape, robbery, and aggravated assault) was 172.3 per 100,000 inhabitants, and the arrest rate for property crime (burglary, larceny-theft, motor vehicle theft, and arson) was 531.3 per 100,000 inhabitants.
Two-year arrest trends show violent crime arrests declined 4.9 percent in 2011 when compared with 2010 arrests, and property crime arrests decreased 0.1 percent when compared with the 2010 arrests.
Arrests of juveniles for all offenses decreased 11.1 percent in 2011 when compared with the 2010 number; arrests of adults declined 3.6 percent.
Over 74 percent (74.1) of the persons arrested in the nation during 2011 were males. They accounted for 80.4 percent of persons arrested for violent crime and 62.9 percent of persons arrested for property crime.
In 2011, 69.2 percent of all persons arrested were white, 28.4 percent were black, and the remaining 2.4 percent were of other races.