4 ms·
Perhaps silly, but let me explain a little more. Here's the scenario I envisioned after reading your original comment: Say you have 10 people who have committe
by elmuchoprez 14y ago
Perhaps silly, but let me explain a little more. Here's the scenario I envisioned after reading your original comment:
Say you have 10 people who have committed a crime and under the current system 9 get convicted and go to jail for 10 years and and 1 gets off on a technicality. So, there's collectively 90 years of prison time being served.
Now imagine we introduce plea bargaining and offer 5 years to all those accused and all 10 take it. The conviction rate goes up, but the total amount of jail time being served is drastically down.
Now I don't know what kind of "discounts" plea deals usually offer, or what it increase the conviction rate to, but I doubt it's an "obvious" win if you're counting by the number of years convicted offenders are removed from society.
Reading your silly answer, I think you're arguing something even less "obvious", which is that willingness to commit a crime is a direct function of age. And if that's the case, and if we're measuring benefit by a decrease in one's willingness to commit a crime, well then your system makes sense because any jail time is essentially better than none because the majority of the decrease in willingness to commit a crime happens at the front end of the sentence anyways. But I think that requires a broad assumption about the correlation of age and crime.
- jpwagner 14y agoThat's a fair point. I would argue that we're not trying to optimize time in prison but benefit to society. Your argument relies on a linear(ish) relationship between the two.