There are so many complicated variables in this, but one thing that is crucial to point out is that he keeps referring to state prisons, which is intentionally

There are so many complicated variables in this, but one thing that is crucial to point out is that he keeps referring to state prisons, which is intentionally misleading because many many drug offenders end up in federal prisons, where they make up 48.4% (https://www.bop.gov/about/statistics/statistics_inmate_offenses.jsp https://www.bop.gov/about/statistics/statistics_inmate_offen...) of the population. That alone makes a lot of this guy's commentary suspect, and since he didn't even mention race or private prisons as a possible factor it really tanks the smell test on this new theory of his.