Scott’s article Prison And Crime: Much More Than You Wanted To Know went hugely viral. Some thoughts:
1. The baseline value is too high for criminals on a utilitarian basis. Criminality is positivity correlated with lower IQ, unemployment, worse prospects, low income etc. Criminals when not doing crime can be assumed to have a lower ‘value’ compared to the general population on a per capita basis. The cost of incarceration is thus close to zero in terms of lost human capital.
2. I have seen some make the argument that it’s possible for punishment to not be bad, but at the same time the criminal is prevented from committing more crimes by limiting his freedom. I don’t see these as compatible.
The very act of physically preventing a criminal from committing crime must imply some sort of imposition being placed on the free will of the criminal. Even the nicest of penal colonies are inhumane in this sense.
3. To reduce costs, from an ROI perspective, a case can be made for punishing scalable crimes more harshly than crimes which do not scale as well, even if the latter are more heinous. So longer sentences for property crime, particularly theft, and shorter sentences for manslaughter.
Shoplifting, particularly in the context of a criminal shoplifting ring, is scalable across many stores, cities, and states and imposes considerable externalities on society. Due to thin operating margins, even ‘big box’ stores can be forced to close due to only a handful of shoplifters who are employed as part of a crime ring.
But the drunkard who accidentally knocks someone out in a fight, who falls backward and hits his head on a curb and dies, is not a scalable crime. This crime has much more limited externalities, those being the victim and his close family. It’s not like being a drunk scales that well compared to running a crime syndicate. By comparison, a store going out of business due to theft affects many employees and their families, and also the local economy.
In terms of sentencing, the U.S. criminal justice system is modeled after this approach. Despite the apparent gravity of manslaughter, the sentencing is shorter than the 20-years maximum for wire fraud or money laundering, which are often paired in the context white collar crime (wire fraud to procure the funds and laundering to hide or obscure it). Only first-degree murder carries a potential capital punishment, but it drops off quickly after that. The maximum for federal sentence for voluntary manslaughter is 15 years, and in California sentencing ranges from 3-10 years. Involuntary manslaughter is much less, typically 2-4 years. Also, being that violent crime is typically within the jurisdiction of the state instead of a federal crime, such sentencing may be eligible for parole.
4. Assigning value on a utilitarian basis to human life is dubious overall. It assumes no substitution effects
(interchangeability) and full employment. Let’s assume an Amazon warehouse worker dies. How much value is lost? It’s hard to say. Due to slack in the labor market, he would quickly be replaced by someone who is unemployed looking for that job. On the other hand, someone like Jeff Bezos is irreplaceable and has no obvious substitutes.
Overall, I think Scott’s argument puts too much weight on the value of human life, which is imprecise, and overlooks the economic cost of crime, which is more quantifiable and more applicable.