The five means by which the War on Drugs can drive up incarceration
rates (or punishment more generally) considered in Part II are (1) the direct
incarceration of drug offenders, (2) the re-incarceration of all types of offenders
due to drug-related parole violations, (3) the impact of drug incarcerations
on prison admissions instead of prison populations, (4) the extent to
which prior drug offenses trigger repeat-offender enhancement, even for
non-drug crimes, and (5) the effects of large-scale drug arrests and incarcerations
on neighborhood social cohesion, and the connections between social
stability and incarceration.
Not only does it not ignore that question, it addresses it empirically. You should read past the 14th footnote, where you say you stopped; this is addressed towards the end of section 2, about halfway through.
No, all of those things are true, they just don't have overwhelmingly large effect sizes. They are, undoubtedly, still driving up the incarceration rate.