Tuesday, October 05, 2010

Prison populations and sleight of hand statistics

A commentor from yesterday points out that most prisoners serve only short sentences hence they do not qualify for rehab programmes. That is why only slightly more than half of the prison population is working or training apparently.

The commentor links to this;

Clearly the majority of offenders ending prison sentences spent relatively short periods of time (less than six months) in prison as sentenced prisoners. Less than five percent of those released each year have spent more than three years in prison.

Now that paints a very different picture from the last Prison Census (discontinued after 2003) which looks at a point in time population as opposed to all prisoner movements. This table shows that only 4.9% of prisoners were serving a sentence of 6 months or less. Granted that was 2003 but the statistical phenomenon will persist. The same happens when considering how long people spend on benefits. If it is worked out over all people who ever used a benefit, the time is dramatically shorter than if it is calculated at a point in time.

Leaving aside the issue of remand prisoners, it looks like over 90 percent of prisoners are serving sentences over 6 months but a large chunk of them are not involved in work or training.

And returning to the remand or short sentence inmates, given the illiteracy and life skills deficit problems surely there should be mandatory programmes operating for these inmates as well.

2 comments:

Anonymous said...

Although you've accused me of statistical sleight of hand, I was just using the best and most recent data available on how long people are spending as sentenced prisoners (i.e not on remand and therefore eligible for programmes).

As well as being almost a decade old now, the evidence you are pointing to does not, in fact, show how long people spend as sentenced prisoners but rather the sentence imposed on them. These are actually different things. The time someone spends as a remand prisoner is taken off the sentence imposed once it begins. For example, if you were given a one year sentence by the judge, but spend six months on remand (and some people spend more than two years as a remand prisoner) then you would only be a sentenced prisoner for the remaining six months.

There are also other reasons why prisoners may not be eligible for education or work programmes including:
- having a high security classification
- being a segregated prisoner
- having broken the rules of the prison (e.g. drug use)

You will note that none of these excluding factors are related to the 'perverse and unintended effects' of welfare nor the unwillingness of prisoners to participate. So accuse me of statistical sleight of hand all you want but the real sleight is the flawed logic of your original argument.

Lindsay Mitchell said...

Anon, It was not my intention to "accuse" you. Probably my choice of words for the blog post title was poor. What I meant to highlight was the statistical phenomenon of variation between point- in- time data versus other. My apologies if you thought I intended otherwise. I accept the reasons you have put forward. But they don't negate the reason I suggested. No doubt there are multiple factors at play.