Hacker News new | past | comments | ask | show | jobs | submit login
Why It Might Be Time to Rethink the Rules of Parole (fivethirtyeight.com)
49 points by ryan_j_naughton on Nov 13, 2014 | hide | past | web | favorite | 9 comments

Also I think we should stop making it a requirement of parole that the convict admit to the crime of which they have been convicted.

I certainly understand why parole boards have required these admissions. If someone has committed a crime but will take no responsibility for it, that certainly does not bode well for their future behavior.

The problem is that the justice system is not perfect, and occasionally convicts an innocent person. It's not right to demand that someone admit to a crime they actually didn't commit, and to deny them parole until they do.

Accordingly I submit that the only consideration for parole should be the convict's behavior while incarcerated.

> The problem is that the justice system is not perfect, and occasionally convicts an innocent person.

Occasionally? Given the 4% rate of false conviction for death penalty defendents (http://www.pnas.org/content/111/20/7230.abstract) and the fact that 97% percent of criminal cases dealt with in the U.S. don't even go to trial, it's far more than occasionally. Often. Quite often.

Without disagreeing, consider the following thought experiment:

The rate of conviction in Hypothetica is 100% for all crimes. If you're charged, you get convicted.

What percentage of the convicted were convicted falsely?

One of the central tenets of our criminal justice system is that it is better that two and a half guilty men be set free than for one innocent man to be falsely convicted.

On a related note, I was surprised to learn that there is NO parole for federal crimes. It was eliminated under Reagan in 1984 as part of "getting tough on crime".


Changing the parole system is going to be a tough sell to politicians ("it's always safe to be tougher on criminals"), law enforcement ("if you let them out, they'll do it again"), the right wing ("burning them alive is too good for them"), business ("we build prison five cough hundred-thousand dolla! incarcerate them long time") and victims (maybe the only people with the right to say yay or nay).

Which pretty much leaves only the left of center and if the last elections are any indicator, we'll be lucky not to get declared criminal by all of the above.

You also have over 1% of the adult population in prison (far over 1% in certain demographics). These people cannot vote (which is its own problem), however they have friends and relatives who can.

Bayes Impact (YC S14) is/was working on a project related to this, using risk models to help predict recidivism.

how about a risk model that given the evidence predicts/concludes guilty or not. I wonder whether any trial lawyers running something like this.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact