Friday 26 December 2008

Murder, firearms and all that

Another bit of superficial reporting from the SMH, this time that murder by firearms is on the decrease.

If the reporters bothered to read the report, rather than just regurgitating the press release, they'd find this gem:

More than 93 percent of firearms used in homicides in 2006–07 were unlicensed and unregistered. Not surprisingly given the low level of legal ownership of firearms amongst those involved in homicide, 48 percent of firearm homicide offenders had a prior criminal history, compared with 37 percent of victims.
So the report is telling us what most of us know - crims kill other people with unlicensed weapons. This never stops the anti-gun crowd from making a song and dance everytime there is a shooting though.

And then this:

Handguns have made up the majority of firearm homicides in Australia in the past six years (see Figure 27). This has been increasing since the inception of the NHMP, and has remained steady in recent years. In 2006–07, 48 percent of firearm homicides were conducted with a handgun.
So why are we so concerned with getting rid of rifles, or making them hard to get hold of? I imagine that rifles outnumber handguns by a factor of 10 or more, yet handguns are used in the majority of murders. Given the circumstances outlined later, the reasoning for this becomes clear (ie, drunk or drugged people arguing over drugs or money).

We also learn that:

Males were most commonly killed for no apparent motive (28%); in alcohol-related arguments (21%); for money/drugs (17%); through domestic altercations (15%); and for revenge (12%). Arguments over money can occur in many situations.

The chances of being killed by a stranger are extremely small. If you have a violent, unemployed, broke, drug abusing boyfriend, then you might have a problem.

As to employment, or unemployment:

Unemployed persons are highly over-represented in homicide incidents, both as victims and as offenders. Thirty-five percent of victims and 41 percent of offenders were unemployed at the time of the homicide incident. Although the percentage of unemployment of homicide victims and offenders has fluctuated over the past 18 years, the rates have always been much higher than the national average.

During 2006–07, the unemployment rate in Australia ranged between 4.8 percent and 4.3 percent (ABS 2006 & ABS 2007), meaning that the unemployed are six or seven times over-represented amongst homicide victims, and about eight times over-represented amongst homicide offenders, in comparison with the general population.

Wastoids kill each other. No surprises there. The unemployment rate does not include those on sickness and disability benefits, who also make up a surprising number of killers. (Who would have thought that someone on a disability benefit would be fit enough to chase someone down the street and bludgeon them to death?)

To drugs and alcohol:

Many homicide cases involved substance use. Sixty-seven percent of victims had alcohol, illicit drugs, or both in their blood system when they died (refer to Appendix D). One-third (35%) had alcohol only; 16 percent, illicit drugs only; and 16 percent, both alcohol and illicit drugs. Male and female victims were under the influence of multiple substances and illicit drugs at similar rates (see Figure 29).

Male victims, however, were more likely (39%) than were female ones (24%) to be under the influence of alcohol only, and female victims were more likely (46%) than male ones (28%) to have no alcohol or illicit drugs in their blood system when killed, which may reflect their greater likelihood of having been killed in the context of a domestic dispute.

Perhaps the most critical comment in the entire report is buried in one line on page 33:

For as long as monitoring has occurred, Indigenous homicide-offender rates have been much higher than non-Indigenous ones. The offending rate per 100,000 Indigenous males in 2006–07 (14.0) was seven times that of non-Indigenous males (2.0) (see Figure 34). The 2006–07 homicide-offending rate of Indigenous females, 5.3, was nearly 14 times as high as the non-Indigenous female offending rate (0.4). The rate per 100,000 per year of Indigenous male homicide offenders has, however, fallen from its peak in 1990–1991 of 38.8 to 14.0 on the last measure, close to a third of what they have been.

This fall in Indigenous male offending has been a critical factor in lowering overall homicide rates in Australia.

The bit in bold was on the last page of the report before the data appendicies started - ie, where no on will ever find it.

If you ask me, the firearms amnesty has had little or nothing to do with a declining homicide rate.

The report tells us that:

The age of offenders (Figure 9) is a stronger predictor of homicide than the age of victims (Figure 4) in 2006–07. In males, there is a steep rise in homicide offending rates until they are in their mid twenties, followed by a gradual falling off through life.
If that is the case, why is no attempt made to correlate the age profile of the population with the homicide rate? Could it be that homicide is falling because there are fewer males in the crucial age group?

But the clanger of course is that last line about the fall in the male indigenous homicide rate being a critical factor in lowering the overall rate.

The report doesn't do it (possibly for fear of being branded racist), but I would be interested in seeing a graph showing the non-indigenous homicide rate. For all we know, it is stable and showing no signs of decline - which is a worry. I'll tell you why.

This report is being trumpeted as one showing a declining homicide rate, which must be due to more effective policing and firearms policies etc by state governments - that's how I would spin it. The intent is to make you and me feel safer and better protected by our competent, caring legislatures, Police forces and justice systems.

But what if that is not the case? Consider the two sets of numbers, which I have made up for illustrative purposes:


Year 1 - 100 + 20 = 120
Year 2 - 100+ 18 = 118
Year 3 - 102 + 15 = 117
Year 4 - 104 + 12 = 116
Year 5 - 105 + 10 = 115

The first number (the non-indigenous homicide rate) is going up, whilst the second number (indigenous murders) is going down, but more rapidly. Add the two together and you get a slowly declining rate.

But the use of a summary number, rather than the broken out statistics, gives an entirely false impression of how safe (or not) it is to walk to streets of suburban Sydney.

As we've been told time and time again, there's lies, lies and statistics.

No comments: