> If a particular means consistently yields good results, is it bad means?
I think so. Here's an example: over 3/4ths of released prisoners are re-arrested within five years. Half are re-arrested within a year of release[1]. If my end is substantially reducing the crime rate, I could reasonably claim that mandatory lifelong sentences would achieve that end: recidivism would be eliminated, and there would be a strong disincentive for anybody to commit even minor crimes. But this is clearly a terrible means, both legally and morally: the punishment doesn't fit the crime, and we end up condemning 25% of the prison population to a life in jail because of a statistic.
I think so. Here's an example: over 3/4ths of released prisoners are re-arrested within five years. Half are re-arrested within a year of release[1]. If my end is substantially reducing the crime rate, I could reasonably claim that mandatory lifelong sentences would achieve that end: recidivism would be eliminated, and there would be a strong disincentive for anybody to commit even minor crimes. But this is clearly a terrible means, both legally and morally: the punishment doesn't fit the crime, and we end up condemning 25% of the prison population to a life in jail because of a statistic.
[1]: https://www.nij.gov/topics/corrections/recidivism/Pages/welc...