Gary Becker's biggest mistake

[littlewing]

One of the more successful, though still brutal, experiments in handling crime, in my opinion, was the penal colony in New South Wales, Australia:

https://en.wikipedia.org/wiki/History_of_New_South_Wales

From 1788-1792, it started out tough with thousands of professional criminals ill-fitted for the skills required. 1/4 of the 2nd fleet lost their lives. The rest survived near starvation. However, between 1810-1821, it transitioned from a penal colony to a budding free society.

I can't think of many successes like that in our history where so many career criminals were reformed so quickly, the sacrifices of lives not withstanding.

[idlewords]

A lot of those criminals were people who were transported for trifling offenses like theft. The real driver behind Australian settlement was overpopulation in Britain, not the need to get rid of some hardened criminal class. The Fatal Shore is a good, readable history of this social experiment.

Note also that this "budding free society" was responsible for horrible atrocities against the native population until fairly recent times.

[Retra]

I don't know if you can count that as a success. And almost all of those people "lost their lives" in the sense that they had much that was a part of their prior life stripped away.

You could send criminals to the moon and find that those who survive don't commit crimes, but that is one step below ending their lives, and not a crime prevention program significantly more successful than just executing them.

[josu]

tl;dr: His mistake: longer sentences didn’t reduce crime as much as expected because criminals aren’t good at thinking about the future; criminal types have problems forecasting and they have difficulty regulating their emotions and controlling their impulses.

[jbapple]

I think the notion that "more police on the street [makes] punishment more quick, clear, and consistent" needs supporting evidence.

[erikpukinskis]

This is a harder sell in the connected age, because every time someone on early release commits a serious crime it is widely reported and proof in the public's eye that longer sentences make sense.

But when someone is released early and quietly finishes out their life that doesn't make the news.

[gcb0]

> he argues that an optimal punishment system would combine a low probability of being punished with a high level of punishment if caught

the Trailer Park Boys series has an episode were small criminals decide to lower their crimes even more so police would ignore them.

[skywhopper]

> "models are to be used but not to be believed."

Data scientists, please take note.