Lies, damn lies and Tim Blair’s stats

Tim Blair blogs an item about Australian gun laws and crime rates:

Despite Australia having “the most up-to-date” gun laws, gun crimes still happen somehow:

From 1999 to 2002 the number of robberies involving firearms in Sydney’s most populated areas rose by 34 per cent, while handgun homicide has grown from 13 to 50 per cent since Martin Bryant killed 35 people at the Port Arthur tourist site in Tasmania in April 1996.

Those stats are from a UN conference, by the way. The proposed solution? More laws!

Being a curious little armadillo, I went to the Australian Bureau of Statistics website and checked the latest figures.

I found that in fact the homicide rate involving firearms has remained static between 1996 and 2001 (the latest figures). The number of murders using firearms fell from 0.5 per 100,000 population to 0.3, while the number of attempted murders rose slightly from 0.6 to 0.7. As for robberies, the number of robberies using firearms was 8.7 per 100,000 people in 1996 and exactly the same number in 2001 (although it rose and then fell in between those years).

Why the discrepancy between the official statistics and those quoted by Tim? Well, I think it has something to do with the qualifying words to both quoted figures: robberies … in Sydney’s most populated areas and handgun homicide. The statistics simply show that while overall firearm robbery rates didn’t change at all, there are some hotspot suburbs where they rose significantly. Hardly a surprise, and it says nothing at all about the efficacy or otherwise of gun laws (one way or the other).

However, the fact that homicides by handgun rose by 13 to 50 percent (presumably as a proportion of the total number of firearm homicides which, as we know, didn’t rise at all) appears to be rather more significant. It must mean that homicides using rifles and shotguns (i.e. firearms which are not “handguns”) actually fell by the same amount. Since rifles and shotguns (at least the semi-automatic variety) have been subject to the post-Port Arthur gun laws promoted by PM John Howard, while handguns haven’t, the figures strongly suggest that the Howard gun laws have been effective (contrary to the fond wishes of the gun lobby). No doubt that’s why, as Tim observed, the conference paper presenter was suggesting that similar restrictive laws concerning handguns might well have a similarly beneficial effect on crime rates. I’m indebted to Tim Blair for drawing my attention to these very interesting figures.

What appears to be occurring is a substitution effect. Because it’s more difficult for potential offenders to get their hands on a semi-automatic rifle or shotgun, they’re using handguns or knives or syringes instead. Now, on one view that doesn’t provide much cause for rejoicing. The victim ends up equally dead whatever weapon is used. However, it’s much less likely that a situation involving a junkie holding up a chemist shop, or an irate spouse or neighbour in a domestic situation, will spiral out of control and result in homicide where a gun is not involved. I strongly suspect the vastly greater gun proliferation in the US is the main reason for their much higher homicide rate. We’d be much better off if all junkies robbing chemist shops were armed with knives or syringes rather than guns (of whatever sort). There’d be many fewer deaths, although the same number of robberies.

As I wouldn’t want to be equally guilty of making selective or misleading use of statistics, I should also mention that, although firearm crime rates for homicide and robbery haven’t risen since 1996, rates for use of firearms during kidnappings have risen significantly, while firearm useage rates during other categories of violent crime (assault; sexual assault) have fluctuated but shown no clear pattern either upwards or downwards.

Also, although the use of firearms during violent crimes generally hasn’t risen at all, the total number of those violent crimes rose quite substantially right through the 1990s (that is, there is significantly more crime, but no increase in the use of guns to commit them). That in itself tends to suggest that gun laws have been effective. However, it isn’t possible to draw any rational connection between the increase in overall crime and changes in gun laws (not least because the increases began long before the Port Arthur laws were introduced). By the same token, and despite the discredited attempts of John Lott Jnr, it also isn’t possible to draw any rational connection between American gun laws and the fairly dramatic fall in US crime rates during the 1990s. Instead, Peter Saunders and Nicole Billante make a quite persuasive case that both the US and Australian trends are explained, at least in part, by changing law enforcement and sentencing patterns. US policing has become tougher and more effective, as has criminal sentencing, and that has been reflected in a significant fall in crime rates. The opposite pattern has been evident in Australia (and the UK).

The bottom line? The lefties are correct that tougher gun laws are effective in reducing the commission of crimes using guns. The righties are correct that tougher policing and sentencing are effective in reducing crime rates overall. As the title to Saunders and Billante’s article succinctly puts it: “Find ’em, catch em’, put ’em in prison: how to reduce the crime rate.” It isn’t really rocket science, it’s just that prevailing left ideology prohibits any admission that “law and order” policies actually seem to work, while prevailing right ideology doesn’t allow the possibility that restrictive gun laws work too.

About Ken Parish

Ken Parish is a legal academic, with research areas in public law (constitutional and administrative law), civil procedure and teaching & learning theory and practice. He has been a legal academic for almost 20 years. Before that he ran a legal practice in Darwin for 15 years and was a Member of the NT Legislative Assembly for almost 4 years in the early 1990s.
This entry was posted in Law. Bookmark the permalink.
17 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Alan
Alan
2022 years ago

I’m too slack to hunt out the figures on recidivism among offenders sentenced to prison as opposed to offenders sentenced to probation, community service and the like. While I agree with find ’em, catch ’em’ what are your thoughts on bleeding heart sentencing options?

Gummo Trotsky
2022 years ago

Ken,

It took me a little while to get the point of your argument re the increase in handgun use – which, I now see, is that handgun usage probably increased because handguns were being substituted for the harder to get long-arms.

Re “find ’em, catch ’em, put ’em in prison” the problem I have with Laura Norder politics is the last stage: the community and politicians lose interest once we’ve turned ’em into prisoners. A lot of Laura Norder advocates seem unaware of the three R’s of punishment: retribution, restitution and rehabilitation. The first R is generally considered sufficient for political purposes, along with the flawed assumption that heavy retribution (up to and including the death penalty) acts as a deterrent to crime. if I may be allowed a sweeping generalisation, there’s too much historical and criminological evidence to show that this belief is plain wrong.

Ken Parish
Ken Parish
2022 years ago

Alan,

“Bleeding heart” solutions have an important place for juvenile offenders and first offenders. The great majority of first offenders never commit another crime, so it doesn’t make sense to lock them up. That is especially so since current research suggests short sentences (up to 1 year) tend to be counterproductive. Any deterrent effect is outweighed by the “school for crime” effect, so we end up making crime rates worse not better.

Research suggests that community “conferencing” options, where the offender is confronted by the victim, other members of the local community and his own family, can have positive measurable effects for juvenile and first offenders. The offender is confronted with the effects of his behaviour, suffers shame, and negotiates a punishment with the victim and local community, which is then enforced by authorities. Another “diversionary” sentencing option that has shown some promise is one in North Queensland involving young Aboriginal offenders. They do a “roughing it” program where they learn to harvest useful bush medicine and other plants and seeds, learn the skills to turn them into saleable commodities, and then get jobs selling the produce they’ve created (at Kuranda markets from memory). Peter Botsman wrote an op-ed piece on the program recently, and results appeared quite impressive.

However, for adult and second and subsequent offenders, the tougher US approach appears to be the way to go.

Gummo,

Your position isn’t borne out by the facts. Research shows that increasing the probability of being caught (effective policing) and imprisoned DOES have a very significant deterrent effect. On the other hand, the evidence that longer sentences increase the amount of deterrence is much more equivocal. The dramatic US reduction in crime rates seems to flow more from an “incapacitating” effect of VERY long sentences than from deterrence as such. Many US states have habitual offender statutes and/or “3 strikes and you’re out” mandatory sentencing laws, which result in EXTREMELY long sentences (often life) for habitual offenders and those sentenced for the third time for a serious indictable crime. Because a very high proportion of the total crimes is committed by this small group of core serious offenders, locking them up for a very long time has a very significant downwards effect on overall crime rates. However, it also has very serious social effects on the families and communites containing most of those core offenders. In the US that means black and hispanic communities; in Australia it would mean Aboriginal communities. These are hard decisions. Adopting US policies almost certainly WOULD result in major crime reductions, but at significant social cost.

Gummo Trotsky
2022 years ago

Ken,

I stand by my position – the historical example is Robert Peel’s creation of the Police Force, which was much more effective in deterring crime (because offenders could now reasonably expect to be caught) than previous “string ’em up in public to warn the others approaches”. My objection is to the reduction of law and order politics to the equation “tough on crime = more people in prisons for longer”, in preference to “tough on crime = effective policing + effective punishment”.

Still, I have to confess that most of the knowledge that I have on the subject of criminology is second hand – mostly consisting of a few memorable bits and pieces of trivia that rubbed off over a few years of living with a criminologist ;)

Ken Parish
Ken Parish
2022 years ago

Gummo,

From what I can see, you’re not actually disagreeing with me. I conceded that the research seems to indicate that the deterrent effect flows from the level of certainty of detection and imprisonment, not the length of the sentence. However, very long sentences can have significant crime-reducing effects for reasons having nothing to do with deterrence. The reason, as I said, is incapacitation: frequent serious offenders don’t commit crimes while they’re in prison (except against each other), so if you imprison enough of them for long enough then crime rates are affected.

Although it’s probably impossible to prove that proposition irrefutably, except by dodgy data-mining techniques like John Lott uses, I nevertheless think the US bundle of “tough on crime” initiatives is the most likely explanation for the fairly dramatic reduction in crime rates they’ve achieved through the 1990s.

Gummo Trotsky
2022 years ago

Ken,

Glad we got that sorted out. BTW on the nastiness of punishment as an inherent deterrent to crime, I think the ACOSS report on crime rates & imprisonment rates in Victoria might be relevant – both are the lowest for any state in Oz *gloats*

Anthony
Anthony
2022 years ago

And yet we Victorians live in a Police State, where we are regularly – and maliciously – fined if we exercise our God-given right of driving too fast.

Ken Miles
Ken Miles
2022 years ago

I was under the impression (and I’ve read virtually zero research into the topic – so take with a big grain of salt) that the probability of being caught was far more important than the punishment.

cs
cs
2022 years ago

The first bit is nicely unpicked Ken, but I’m having trouble following this:

… both the US and Australian trends are explained, at least in part, by changing law enforcement and sentencing patterns. US policing has become tougher and more effective, as has criminal sentencing, and that has been reflected in a significant fall in crime rates. The opposite pattern has been evident in Australia (and the UK).

I thought we have gone nuts on hiring more cops and locking people up, to no apparent improvement (or, in fact, on Don Weatherburn’s stats, which I’ve got here somewhere, everything has just got worse … as you suggest … but despite the tougher policing and sentencing). Moreover, could it not be suggested that the divergent patterns between the countries may resemble the divergent unemployment levels? Or haven’t I followed you properly?

Ken Parish
Ken Parish
2022 years ago

Chris,

It might be worthwhile reading the Saunders and Billante article (although you may have philosophical objections to doing so). There are particular attributes of US policing strategies (e.g. the “broken windows”/zero tolerance stuff) and sentencing laws (especially “3 strikes” mandatory sentencing, habitual offender laws, sentencing commissions establishing guideline sentences etc), which I suspect in aggregate have been at least partly responsible for the fairly large fall in US crime rates over the last decade and a bit. Most of these initiatives are not duplicated in any part of Australia, although WA had a “3 strikes and you’re in” law at one stage (although I think it only related to car theft). I wouldn’t even attempt to argue that there aren’t also other factors involved in the US crime rate drop, although precisely what they are isn’t all that clear to me.

I’m not sure the analogy with unemployment levels really helps our understanding all that much either. It wasn’t all that difficult to see why the US was achieving significantly lower unemployment levels than Australia through the second half of the 1990s: they had a larger and more prolonged boom than us, and a more deregulated employment market which therefore was able to absorb more low-paid “working poor”. Moreover, since the boom ended, the differential has evaporated: I see the latest US unemployment numbers are 6.4% whereas Australia’s rate is about 6.0%. It’s nowhere near as easy to isolate a clear and obvious social, demographic or economic reason for US crime rates to be falling significantly throughout the 1990s while Australia’s and Britain’s were rising. That’s one of the major things that leads me to suggest that policing and sentencing regimes probably provide at least part of the answer. I’m certainly open to be persuaded otherwise, but these are quite marked phenomena that warrant more analysis than they’ve received.

cs
cs
2022 years ago

Far enough Ken, but it could be interesting to watch the rates as the unemployment levels rise, as poverty will presumably challenge the policing practices at the margin.

Scott Wickstein
2022 years ago

I doubt unemployment alone makes a scrap of difference to the crime rate.

Australia had an experience in the 1930’s of 20% and more unemployment, yet no history of the period (that I have read, anyway) makes reference to a crime boom.

cs
cs
2022 years ago

Surely crime is an option that, even if one only vaguely subscribes to the ‘hedonist’ calculus, becomes more attractive at the margin as other options for advancement close down. Surely it also becomes a more attractive option as other constraints on crime break down (such as traditional communitarian sanctions). I’m not sure about the 1930s’ ‘crime rate’, and nor am I sure of the validity of this being a stand-alone index of anything between such very different historical periods, but the 1930s was marked by a sudden spate of hangings: 8 in NSW and Victoria between 1938-40, more than twice the number hung in the prior 15 years. Do you have any more basis for your argument Scott?

24601
2022 years ago

Looks like you’ve lost a lot of right-wing and libertarian readers – coz nobody is defending gun-rights here.

As far as I knew (both before and after reading this blog) crimes involving guns haven’t changed much since the new laws. Ergo, the laws are not effective. They are, however, expensive and a large restriction on a generally peaceful and law-abiding section of our community. It’s easy to destroy other people’s hobbies/sports on a whim – but if the whim turns out wrong, then at least do the right thing and let them have their sport back.

Lott’s finding weren’t discredited – Lott the human was. Just because rumour has it that Keynes had sex with aminals, doesn’t mean fiscal policy is ineffective.

The gun debate has long ago left the rhelm of reasonable debate. The fact that identical guns, except for their action (pump v bolt) can be treated differently is absurd. And having a month waiting period for a second gun that has a lower calibre than the first is also against common sense. Also, it should not be a criminal offense if you fail to update your living address to the police within 6 days of moving. But hey – who cares. Shooters aren’t real people are they? Nothing like kicking a minority when they’re down.

Ken Parish
Ken Parish
2022 years ago

24,

If the fatuous standard of the “debate” on Tim Blair’s blog on this issue is typical, then I wouldn’t be losing a lot of sleep if it turns out you’re correct that people of that ilk have stopped reading this blog because they prefer reading material that reinforces their own existing prejudices rather than having them challenged.

You say “crimes involving guns haven’t changed much since the new laws. Ergo, the laws are not effective. They are, however, expensive and a large restriction on a generally peaceful and law-abiding section of our community.” I make two points about that argument:
(1) The figures to which Tim Blair drew our attention appear to indicate that crimes involving guns HAVE changed significantly since the new laws were introduced in 1996. Robberies involving the types of weapons affected by the laws have decreased by “from 13 to 50 per cent” in absolute terms, even though the overall number of robberies has been increasing. If a similar effect flows from restricting handguns, we could expect a real impact on homicide rates (because the only option for most junkie robbers would be substituing weapons that are much less potentially lethal, like knives and syringes).
(2) I make the more general point that the huge ongoing disparity between homicide rates in the US versus Australia, Britain etc is most likely explained in large part by the much greater uncontrolled proliferation of guns in American society. That’s a very high price to pay for leaving the harmless hobby of the responsible majority of gun owners largely unregulated. Note that I would never support banning either handguns or rifles and shotguns (semi-automatic or otherwise) completely. I simply support a regulatory regime that, as far as reasonably possible, keeps weapons out of the home and reserved for use at shooting ranges and when hunting. That seems to me to strike an appropriate balance between the freedoms of shooters to enjoy their hobby and the rights of the rest of the community to protection from violent crime, gun accidents, and domestic violence events that spiral into tragedy because of the presence of a gun.

You also made a substantive point in relation to Lott, and it merits a substantive response. Lott’s apparent fabrication of defensive use surveys (and the Mary Rosh nonsense) isn’t his only problem. John Quiggin and others have also highlighted what appear to be highly dubious “data mining” techniques in his more important primary research purporting to show positive statistical correlations between restrictive gun laws and high crime rates. There’s been much research and analysis in this area, and it’s impossible to do justice to it in a comment box. The bottom line is that the proposition that gun ownership helps to reduce crime (by deterring criminals who will fear homeowners might be armed and dangerous) isn’t credible. As my primary post suggests, it’s far more likely that the reduction in US crime rates through the 1990s is explained predominantly by changes in policing, sentencing and imprisonment practices (and maybe at the margins by unemployment rates as well).

24601
2022 years ago

I’ll be honest. I didn’t read your latest comment. Not for any mean reasons – I just have to get back to work. I just forgot to mention one thing last time and that is: I don’t read Tim Blair’s blog. I don’t read it because it is all about spin, with little substance. Sure – he spins it well, but I prefer ‘thinking’ blogs.

Patrick
2022 years ago

One point about substitution. Handguns are NOT more dangerous than knives. The survival rate for people attacked with knives or handguns are about 85%.

Long arms (rifles, shotguns) are much more deadly. A survival rate of about 25% if I recall correctly.

So substituting handguns for longarms SHOULD reduce the death rate. That it hasn’t means something else is counteracting this trend.

I personally have no problem with middle class, decent people such as myself having firearms. Maybe if there was a income test? :)