22 November 2013

Why crime statistics are crap

| johnboy
Join the conversation
3

The Guardian is having a look at squeals in the UK as police there admit you shouldn’t use crime statistics for anything.

Which makes one wonder about all of Simon Corbell’s grand claims based on our own police statistics.

Join the conversation

3
All Comments
  • All Comments
  • Website Comments
LatestOldest

Pork Hunt said :

If a crime occurs and is detected, the offender gets punished in some way (on the spot fine, suspension of license or even jail etc). Can these stats be used to measure levels of unlawfulness?

If that occurred, it would have to be recorded. There’d be reams of paperwork, even for infringement notices (on the spot fines), cautions and so on.

IP

If a crime occurs and is detected, the offender gets punished in some way (on the spot fine, suspension of license or even jail etc). Can these stats be used to measure levels of unlawfulness?

The surprise and disappointment is that anyone is surprised. As the writer points out, this has been known for decades. Real world example – when there was a recent spate of break-ins into unoccupied houses in my town, I expected a spike in recorded crime for the area (as the average number of break and enters is zero). I asked, and there wasn’t. Because even the people who told the police said “don’t bother recording it, as I am not making an insurance claim as too little damage was done”. And the police complied, so it is as if it never happened.

The writer notes that the gold standard for measuring crime in the UK is the British Crime Survey. This is worth noting, because there is NO equivalent in Australia. The ABS crime surveys http://www.abs.gov.au/ausstats/abs@.nsf/Lookup/4530.0Chapter1002011-12 cover a very limited range of offences, are not conducted as promptly as the BCS, and make little effort to actually count the total number of crimes experienced, which makes comparison with police statistics difficult.

There has been a huge amount of research in the UK and elsewhere on what happens to crime reports, but there has been little equivalent research in Oz, and when it has been conducted it seems to have had little influence on practice.

Even in the UK there is little information on the true level of fraud and white collar crime (particularly by employees) because often it doesn’t get detected, or if detected it often doesn’t get reported, and there are no (or few) surveys of businesses to establish their experiences.

The whole thing should be obvious if you refer to police crime statistics as “recorded crime” – because the act of “recording” is very discretionary, very subject to deliberate or accidental manipulation.

The other way to measure crime is to survey the public and ask them what crimes they have committed. It’s as flawed as crime surveys and police (recorded) crime statistics, but each is flawed in different ways, and would allow a “triangulation” to get a better ideal of the true level of crime, and more importantly changes in it. There is virtually no research in Australia on self-reported offending, although some of the longitudinal studies of child development will gradually cover it (but they have their own methodological problems, with the people dropping out of touch with the researchers over the years and decades being the higher risk groups for offending).

IP

Daily Digest

Want the best Canberra news delivered daily? Every day we package the most popular Riotact stories and send them straight to your inbox. Sign-up now for trusted local news that will never be behind a paywall.

By submitting your email address you are agreeing to Region Group's terms and conditions and privacy policy.