I feel like there's almost nothing said about how between 1980 and 1984, homicide fell by over 20%, robbery by almost 20%, agg assault by 7%,* and rape by 3%.
We have a good theory abt a major reversal of the declines, but the decline itself goes unremarked upon.
Like, if crack hadn't arrived, would that decline have continued? Given that the Great Crime Decline began just five years later, find it hard to think it's some sort of cohort-age artifact.
Did crack interrupt something already started? Making 1991 a dubious start date?
I mean, property crime declined over almost that entire period. Declines were happening.
But if the Great Crime Decline is really something that started in 1980 with a crack interruption in 1984, that throws off, like, a TON of stuff, right? (I see you, lead and abortion and...)
Has there been anything written on this? Feel like every account starts the decline in 1991, which is at one level is literally true. But ... 1980-84 is intriguing.
(* Agg assault bottoms out in 1983, not 1984. That's the footnote above.)
So one more detail. Homicides rise until 1980... as do the number of agencies reporting the FBI. When that levels out... violence falls (until 1985).
It's hard to look at the "more agencies reporting" graphs and the "FBI crime data moves up while victimization data moves down" graphs and not think that that has to be at least PART of the story.
Even though I feel like it is almost never part of the story told.
That doesn't mean that the "rising crime sparked mass incarceration" story is wrong, just that it is missing a word:
"rising REPORTED crime sparked...." The reporting, and perhaps how we thought about it, is real. Just... what it was reporting is... more confusing.
Every day I increasingly doubt that I know anything at all about anything. Today has been a day that really helped solidify that feeling.
From 1960-80, annual murders rose ~14K, from ~9K to ~23K, and we invested SO MUCH in prisons and punishment in response.
From 1999-2019, drug OD deaths rose by ~54K, from ~17K to ~71K. Prelim CDC data says ODs will jump ~20K JUST in 2020.
Nowhere near the political response.
Both target the young--from 1999-2019, those under 40 account for ~80% of the years of life lost to homicide, and ~60% of years of life lost to drug overdoses.
But drug ODs cost those under 40 1.34 million years of life lost, vs. 650K for homicide. More than twice as much.
To be clear, homicide has collateral costs that drug ODs do not--like the recent paper that simply hearing a gunshot within a few blocks does real harms to children.
But both are sudden and traumatic and tear apart families' lives.
For guilt/innocence, state gotta prove everything beyond a reasonable doubt. But for sentencing, the (constitutional) rules mostly go away (30,000-tweet thread on Blakely omitted here).
Totally fine to, say, sentence based on something found by a preponderance.
So: “acquitted.”
“Acquitted” doesn’t = “innocent.” It means “not proven beyond a reasonable doubt.”
Over course of trial, judge may come to think ~60% likely def is guilty. That’s not BRD, but it is preponderance.
I've seen a bunch of tweets today abt how we spend $x per person on Rikers.
That's not how jail/prison finance math works. At Rikers, think ~90% of spending is wages and benefits. Which means total spending is fairly insensitive to population, and it isn't "going to" detainees.
It also means as jail pops fall, spending-per-person-in-jail will almost axiomatically rise, because unless we lay off staff, we'll be spending the same amount, just... per fewer ppl held there.
It's important to stop talking abt jail/prison spending in "per detainee" terms.
Talking about "spending per person in jail/prison" misleads ppl abt where the money is going (wages, not programming), and also means we grossly overstate the savings we hope for when cutting back prison/jail pops (since total spending won't move absent LABOR, not pop, cuts).
In 2016 Survey of Prison Inmates, of those serving time for a VIOLENT crime that is NOT homicide or rape:
• 62% claim NO ONE was physically hurt
• 21% admit serious injury (death/rape/stab/shot/broken bones)
• 17% admit minor injury
According to the BJS prisoner count, there are ~350,000 ppl in for violence other than rape or homicide (SPI-16 est is 375K, so... close).
That would be over 200,000 ppl in prison for violence who claim to have hurt no one.
Further complicates "violence" as a category.
Of course, these are self-reports, which are always noisy and messy.
Still, even if we assume some desire to downplay harms, still suggests many in prison for violence didn't cause physical harm (betting a lot are robberies).