> You're conflating inmate population with convictions, the latter being more salient in a discussion of plea deals. As another poster mentioned, these statisti

> You're conflating inmate population with convictions, the latter being more salient in a discussion of plea deals. As another poster mentioned, these statistics are methodologically biased towards more serious crimes with longer sentences. The original context is this comment, which discusses a percentage of US inmates (not a percentage of convictions): https://news.ycombinator.com/item?id=40906705 https://news.ycombinator.com/item?id=40906705 While it's true that people serving longer sentences represent a higher proportion of inmates than they do convictions, it also doesn't seem obvious to me which base denominator is more important. If anything, weighing convictions according to how serious they are seems more sensible than treating a small misdemeanor conviction equivalently with a murder conviction. Furthermore, plea bargaining cuts both ways: the net effect of a plea bargain is often a conviction on a lesser charge, with less or sometimes no prison time. > It's a question of what we value our prison system(s) for. I would hope for efficient deterrence and rehabilitation. That is, sentences likely to dissuade people from committing crimes, and best efforts to prevent offenders from offending again. You seem to think that the street-level homicide rate is too high, and I'd agree. I would say that, as such murders are generally crimes of passion or disordered thinking - instinct or illness (psychological, social) rather than reason - harsh punishments are not necessarily useful in deterring them. I don't necessarily think the threat of incarceration deters common criminals very much, but it does separate them from the population, which is preferable to simply leaving them on the street.