Thursday 24 September 2009

Juvenile detention and how to stick your head up your arse

I really, really love it when a government agency releases a grab bag of statistics and some "thoughtful commentary" to go with them. Here is the latest lot:

JUVENILE offenders who are sent to jail are just as likely to re-offend as similar law breakers who receive community orders, fines or other non-custodial sentences, a study shows.

The findings, released by the Australian Institute of Criminology, cast doubt on the deterrent effect of jail for young people and question the cost-effectiveness of detention.

'The idea that you can deter juvenile offenders by locking them up is wrong,'' said the lead author of the study, Don Weatherburn, who is director of the NSW Bureau of Crime Statistics and Research.
The story goes on and on, but it is written by Adele Horin, who is a notoriously wet socialist hag.

If you've been reading this blog for long, you'll know that I like to go to the source material instead of relying on the SMH to regurgitate and spin it. And that's what I did.

The first place I went was to the NSW bureau of Crime Stats and Research, where the author of the above report actually works. I found an interesting report on juvenile burglary which stated:

Even if the risk of apprehension for a burglary is as low as five per cent, the average offender will be caught after 20 burglaries. Some burglars commit this many burglaries in less than two months (Stevenson & Forsythe 1998).
Some burglars commit as many as 20 burglaries in less than 2 months. Remember that.

Stick with me:

As with previous studies, we assume that the rate at which a burglar or motor vehicle thief appears in court convicted of a burglary or motor vehicle theft is determined by the frequency with which they commit these offences. To be more precise, we assume that the more frequently people commit burglary and motor vehicle theft, the more frequently they are arrested and convicted of these offences. This assumption is supported by empirical evidence of a close relationship between self-reported offending and the frequency of arrest and conviction (Hindeland, Hirschi & Weiss 1979; Maxfield, Weiler & Widom 200, Farrall 2005).
Got the gist of that? The more crime a criminal commits, the more frequently they are arrested and convicted.

It gets better:

there are two groups of offenders involved in these offences; one of which offends and is convicted at a high rate, the other of which offends and is convicted at a much lower rate. High rate burglars are convicted at a rate that is more than 6.6 times higher than low rate offenders, while high rate motor vehicle thieves are convicted at a rate that is more than 8.5 times higher than low rate motor vehicle thieves. Thus, although high rate burglars make up just 6.4 per cent of the burglar population, they account for about 39 per cent of all convictions for burglary. Similarly, although high rate motor vehicle thieves make up just 4.6 per cent of the motor vehicle thief population, they account for 34 per cent of all convictions for motor vehicle theft. In short, the vast majority of high rate offenders were convicted at least once and many were convicted several times. The vast majority
of low rate offenders, on the other hand, were not convicted at all during the observation period.
The guts of that is that there are two types of crims - the persistent buggers and the minimalist offenders. High rate offenders were convicted at least once and many were convicted several times.

Early evidence provided strong support for the assumption that there are two distinct populations of offenders with different offending rates. Farrington, Blumstein and Moitra (1986) analysed the correlates of recidivism amongst offenders in the Cambridge Study of Delinquent Development and found evidence that there are two distinct groups, which they labelled, respectively, desisters’ and ‘persisters. Patterson,Debaryshe and Ramsey (1990) found evidence for what they call ‘early’ and ‘late’ starters and hypothesised that the genesis of offending in the former group lies in coercive parenting whilst for the latter group it lies in delinquent peer influence. Moffitt (1993) research also suggests a distinction between what she calls ‘adolescent-limited’ and ‘life course persistent’ offenders. She contends that the delinquency observed in ‘life course persistent’ offenders is due to a combination of genetic and environmental factors, whereas the delinquency observed in ‘adolescent-limited’ offenders is due to the onset of puberty and exposure to antisocial peers.

The way I read that is that some people are born losers. They start committing crime young, and they go on to commit it for life. Others fool around when the hormones kick in, and they grow out of it and become respectable citizens. Not all crims are the same.

Their disproportionately large contribution to the total volume of offending means that small reductions in rates of re-offending can be expected to have a significant effect on crime.
What that means to me is that we need to separate young crims into two groups. Those that are busted doing something stupid, but who are really just addled teenagers, should be let off with a warning - but they should still be stamped on hard enough to jolt them into behaving better. Those that are losers should simply be put away for as long as possible, simply to prevent them from doing what they do - which is commit crime. Alternatively, put an explosive collar with a GPS on them and let them go. If a burglary is reported, the police simply detonate any collars within say 1km of that address.

There is something else to note as to why youngsters are locked away. From this report at the AIC:

Crime types:

Violent - 43.3%
Property 34.4%

43.3% of juveniles were convicted for a crime of violence. Now if a teenager attacks another kid with a brick and takes a chunk out of his skull, I'd say the best thing to do is lock the little bastard up. Juveniles are not just in prison because they shoplifted a bag of lollies - many are in there because they are already nasty, drunken, drug-fucked, violent little psychopathic scumbags. Excuse me for suggesting that we keep them off our streets and out of our houses and cars and businesses.

In short, this story in the SMH is a crock of shit. Here's what we now know:

  • 43% of juveniles are convicted of crimes of violence
  • a prolific burglar may rob 20 houses in less than two months
  • the more crimes a crim commits, the more times they are arrested and convicted
  • high rate offenders commit 6 times more crime than low rate offenders, and they are likely to have been convicted multiple times
  • some juvenile crims are persistent losers who will go on to live a life of crime
Now, if the world was perfect, our juvenile prisons would be full of the highly prolific criminals, and first time offenders would only be locked up for something like stabbing someone else.

If that is the case, then consider perhaps two burglars who each robs 20 houses in two months, and is then caught and sent to prison. Before prison, they might rob 60 houses in 6 months - one every 3 days. After prison, one of them goes straight, and the other goes back to robbing - a 50% success (or failure) rate - which is what the study showed. Now, if they were in prison for 6 months, that means 120 houses not burgled. In the six months after they are released, it means a further 60 houses not burgled.

What would be the result if neither was locked up?

We know that 50% go straight, and 50% go back to crime. What it means is that in the entire 12 month period (6 months where they could have been in prison plus 6 months afterwards), 120 houses will be robbed. One robber will go straight, and the other will carry on robbing houses every 3 days. The difference between Plan A and Plan B is that when they are both in prison, no houses are robbed, which effectively halves the crime rate.

One would also hope that we are locking up the worst crims - those committing a crime every 3 days or so. If after a stint in prison, 50% go from committing a crime every 3 days to committing none, then I think that is a big success. What the author of this report does not allow for is that those that are not locked up may be committing occasional petty crimes, whilst those that are locked up are committing regular "felony" type crimes.

I could go on and on, but that's enough for one night.


Anonymous said...

Explosive collars Eh!, one could have fun with that idea. Work out where the likely lads (or ladettes) live, and go and rob the house next door to them. Sit back and watch the show.


Anonymous said...

A great writeup Boab. It's always nice to see the real facts being presented rather than the "good facts" that too many of our lefty media come out with.

You might be interested in, which goes into all this from a slightly different direction. (I think I saw this link on Romeo Mikes blog some time ago.)