Also, you have to consider what is the definition of an offense is to be considered as reoffending. Normally, that data includes parole violations as arrests. You can violate your parole engaging in what most people would consider benign behaviors. It is a confounding factor that many people don't even think about when looking at that sort of data.

Additionally, the "some college" stat is cherry picking. Many state prisons offer "some college" to inmates in the form of courses. It is likely inmates in a state prison system for any significant period of time have have taken "some college". That survey states that a good number of inmates took "some college" while in prison (page 7).

Look at college graduates. That document itself states that it is incredibly unlikely for someone with a bachelors degree to even go to prison. 2% of white state prison inmates aged 20-39 have any college degree (almost 3 decades ago in 1997, and crime rates have fallen since then).

Finally, the reimprisonment rate is much lower in the US in general, some 37% over three years, even as low as 19% in Oregon [1]. And that is for all persons, not bucketed by educational attainment.

Some programs claim around 2% - 4% general recidivism rate for inmates they fostered to obtain college degrees [2]. Again, "recidivism" is a large bucket that includes supervised release violations, rearrests for anything (including violations), re-convictions, and re-imprisonments.

I would wager that, if you could find data on it, that the chance of re-conviction for a new offense for someone who has completed parole, a bachelor's degree, and found gainful employment in said career is near-zero.

[1] https://pmc.ncbi.nlm.nih.gov/articles/PMC6743246/

[2] https://www.vera.org/news/back-to-school-a-common-sense-stra...

[3] https://www.ussc.gov/research/research-reports/effects-aging...

[4] https://www.uschamber.com/assets/documents/uscc_business_cas...