[oskapt]

That’s not recidivism, which is the return to committing crimes after release. This is a punishment for bad behavior while inside the jail. It’s a method of behavioral control in real time. I doubt that anyone who gets out of jail is using the argument of, “maybe I’ll have to eat Nutraloaf if I go back to jail” as a factor in choosing to commit or not commit further crimes.

[fsckboy]

>That’s not recidivism, which is the return to committing crimes after release.

you raise an interesting point, and I did a bit of a dive into various dictionaries, new and old: chances of re-incarceration after release is an important aspect of criminology and public policy, but it's not built into the word recidivism.

recidivism is "backsliding", so it's more a case of "draw a line or a gradient, then measure recidivism across that line or gradient"

in the context of prisoners, it's probably a good idea to be clear about the sense of the definition chosen.

[jerkstate]

it's a small point and maybe a different word would have been more clear (because I agree, recidivism commonly refers to committing crimes again after being released from jail), but none of the definitions I found for recidivism were specific to release from jail - the definition seems to be committing offenses again after some type of intervention. in the case of the nutraloaf, you're already in jail and you commit some additional offense (say, starting a food fight). the article seemed to indicate that in most cases, the offender avoided doing the sorts of things that could get them back on nutraloaf - in my opinion, that meets the definition of recidivism.