Hacker News Reader

Comment by Retric | Hacker News Reader

What about prison? It's ~1% of the adult populaion.

(People in the military, prison and institutions are excluded from these figures.)

Why isn't the military counted as part of the work force? its not like people are drafted into it in the US.

Perhaps the more interesting question since the prison population is excluded: what about ex-cons? Ex-cons have a much more difficult finding work for a variety of reasons, and I believe that ex-cons as a percentage of the work force is significantly more than it used to be. So how much of the rise of non-working males is due to the rise in the number of ex-cons?

Ummm... didn't you read the article?

> (People in the military, prison and institutions are excluded from these figures.)

Or 5% of the black adult male population in the United States. http://en.wikipedia.org/wiki/Statistics_of_incarcerated_Afri... and a 30% chance of being in prison over a lifetime http://www.theguardian.com/world/2003/aug/19/usa.garyyounge

If you count imprisoned men, do you count those who work while being jailed? Seems like it's too convoluted to be included in this metric, though certainly worth noting.

~1% of the adult population? Whoa.

As others noted, they're excluded from the study.

But if you're shocked at 1%, 2.9% are in jail or on probation or parole. They are mostly non-violent offenders. Laws are unequally applied toward African Americans at a ratio of about 10 to 1 (when accounting for incidences of the illegal behavior in each population) with African Americans receiving 60% longer sentences than whites for the same crime.

Prisoners are then put to work in industrial manufacturing for as little as 18c/hr and as much (ha) as $1.50/hr. The prison manufactures are guaranteed to win any government bid, regardless of whether their bid is higher than competitors.

Sorry for the tangent, pet peeve.

Curious question - if hauling someone through the justice system is how we determine whether illegal behavior actually happened or not, then how does one separate the metrics "laws being unequally applied" vs "incidences of illegal behavior"?

Incidence of illegal behavior is generally determined through polling. Basically, you ask people anonymously and they're more honest. For example "Have you smoked marijuana in the last year?"

There are various biases that impact the validity and reliability and there are methods for accounting for those biases. Rates of criminal behavior in the population is a very active and mature area of research.

Another method is extrapolation from statistical trends in crime. For example, most violent crime occurs within race. So you can take victims of violent crime that were not solved and extrapolate fairly good estimates of the perpetrator's traits from the victim. Of course, you still have to account repeat offenders and other confounding factors.

Short answer is, there's not just one way.

That is very interesting, but you would need to know the level of trust the groups have compared to how strong their norms for not breaking the law would be in order to normalize the results (e.g are people of one minority group less likely to trust strangers compared to the majority?) which are factors that you can't know.

Yes, but by the same token, the fairness of he judicial system is also subject to factors you can't know. It is unreasonable to assume hat the crim justice system Is an island of unbiased rationality, considering the steady stream of corruption revelations.

Reply on news.ycombinator.com