Gun laws save lives

It’s not a surprising conclusion, but given the controversy on this topic, it’s important to get the stats right. Andrew Leigh and Christine Neill have done a study concluding that, while the data set is too short for a conclusive resolution, the best estimate is the gun buyback undertaken by the Howard government after the Port Arthur massacre has saved between 1000 and 2500 lives. The work of Leigh and Neill is a response to a very dubious study claiming no effect that came out last year.

13 thoughts on “Gun laws save lives

  1. Take 2

    Intuitively I feel this estimate is an order of magnitude too high. Excluding suicides I would want to divide the data into homicides attributable to registered vs unregistered guns if that info was available. Similarly I would like to know just what numbers of working guns in either category are in circulation. Then I’d like to run the data though more models (not just ARIMA) than most of us have with Excel. Finally I’d like to compare Australian data with that of flakey countries like the US and non-flakey countries like Canada.

    Where I’m coming from is that your Ivan Milat types didn’t hand in their guns during the buyback. Therefore I believe the effect was relatively minor and may have prevented just a few ‘heat of the moment’ events.

  2. “Intuitively I feel this estimate is an order of magnitude too high”
    You don’t need to use intuition. You can look at the paper and see if there is something wrong with the statistics/assumptions. If that happens to be the case, then you will know why your intuitions deviate from the analysis.

  3. There can be no doubt that gun laws save lives, if only from accidental shootings. The same logic should be applied to internet pornography which, if site blocking software was applied at service providers, would help to limit the flood of material available to our young adolescents. Guns have a very visible immediate effect on a community. Pornography flood is a hidden addictive destroyer the full consequencess of which are yet to become visible. Some pornographic material is unavoidable, if not necessary, but the saturation that can be achieved through the internet is not at all healthy.
    What is the connection between guns and pornography? Proponents of both use the same argument, freedom of individual choice, to support the perpetuation of access. Both are income vehicles of undesireable elements in our society. Both play to a sense of disrespect, if not dominance, of one person towards another. Both ooze out of the United States.
    Australia, at least, has a handle on one of these social distrotions.

  4. With total homicides in Australia around the 300 mark per annum and those using guns considerably less (53 victims in 2003-04, 40 victims in 2004-05) , the reduced death rate from guns must presumably be mostly reduced suicides and accidents. Whilst I support gun laws, I think, with the current debate’s focus is on homicide, the numbers for suicides and accidents ought to have been prominent, as the impression is that we’ve seen a very large drop in gun homicides.

  5. Hermit says “Where I’m coming from is that your Ivan Milat types didn’t hand in their guns during the buyback.” The guns buy back was not aimed at serial killers (see the restrictions on class C and Class D weapons) but on running amok massacre( Port Arthur).
    The study of Jeanine Baker and Samara McPhedran has not been well discussed. I would suspect that there is insufficient power to be able to detect any difference( Ken’s point) Other methods ( is exact ARIMA in StatXact) should have been used, The outcome variable is not normally distributed

  6. BilB – try as I might, I’m unable to locate a single deceased person whose cause of death includes ‘access to p*rnography’. As a ‘hidden addictive destroyer’, it must rank lower than tim-tams or train-watching.

  7. Hal, like all “hidden addictive destroyers” the consequences are masked. There have, however, been in recent years a number of highly publicised suicides directly related to pornography addiction. Look again. My comment was that the internet flood of pornography is just a few years old and the damage to individuals, relationships and families is yet to become fully apparent. Look in todays SMH for an article about a chief prosecutor.
    Getting back to guns, the real effect of restriction to weapons is that it prevents many people from having and wanting them in the first place. Ivan Milat did not need guns to be a monster. With one female victim he severed her spine with a knife from behind disabling her to magnify his sense of control as he continued the torture. Any thing is a weapon to such individuals.
    People having guns in an ever crowded world guarantees fatalities of both humans and animals. If every American with a gun took up their constitutionally protected “right to have a gun” and hunt, then they could kill every significant mammal in America in less than a week. The population of the world when the American constitution was drafted was less than 1 billion people. It is now 7 billion people. There are just too many people now. What was right 200 years ago is not appropriate now.

  8. The problem with Sinclair Davidson’s formulation is that it is so general that it is useless and not really worth debating.

    Of course, employers have a right to employ and, of course, once a worker is employed, he or she has a duty to live up to the terms of his or her employment contract. But the real issue is what should be the government’s role in setting the legal framework or rules of the game for such contracts. Sinclair’s formulation is consistent with a free for all by employers but it also consistent with a set of worker protection rules which ensure fair and equal bargaining, sets a reasonable minimum wage, protects working conditions and in which governments invest heavily in human capital to improve the employability, skill endowment and earning power of disadvantaged workers.

  9. This could equally well read “John Howard saves lives.” But dont hold your breath waiting for the Lancet and so on to blazon that headline accross their front page.

  10. The article by Leigh and Neill is full of rather tortured over reach as it trys to articulate conclusions that the data and their own analysis does not support. They even go to great pains to disown their own conclusions before making them. Strange indeed.

    Baker and McPhedran’s response here:
    Nice charts here:

    Focusing on suicide for a moment.

    1. We do not now have less guns in Australia than we did in 1996. The $500 million dollars expended on the buyback reduced the number of semi-automatic rifes in Australian society (reputedly) but the total number of firearms in Australian society has not declined. Many and probably most people who sold their semi-automatic weapons to the federal government under duress used the compensation funds to buy an alternate firearm.

    2. When using a gun to kill yourself the fact that the firearm is capable of discharge a second bullet by a simple second pull of the trigger is not overly meaningful. After the first bullet you are generally very dead or at least incapable of pulling the trigger a second time.

    3. It is extremely hard to see in pure mechanical terms how a law reform that has shifted lawful owership from semi-automatic firearms to conventional bolt action firearms would have any causal impact on suicide by firearm, regardless of what people might gleem from statistics on suicide trends.

    Focusing on mass killings.

    1. The data analysis in the report by Leigh and Neill looks at homicide in general as well as at suicide. It does not review data on mass killings in any depth and certainly not across the long timeframes they advodate.

    2. In discussing the substitution effect Leigh and Neill don’t discuss examples of non-firearm mass killings.

    3. Clearly the analysis by Leigh and Neill says little about the impact of the 1996 gun controls on mass killings. As such it is not a useful reference for a discussion about any supposed causal linkage between the frequency and impact of mass killings and specific firearm controls.

    Focusing on Homicide.

    1. The most relevant statement by Leigh and Neill (ignored by the media reports) is the following one:-

    As we point out, the high degree of variability in the underlying data and the fragility of
    the estimated results with respect to different specifications and even statistical
    packages used suggest that time series analysis alone cannot conclusively answer the
    question of whether the NFA cut gun deaths.

    2. If you can’t conclusively demonstrate a benefit from $500 million dollars in expenditure then the only sane conclusion is that you should not have been so quick to expend the money. 1996 was knee jerk politics (I was wrong to support it at the time).

  11. Terje, since the post notes the uncertainties, I don’t really know what you are complaining about here. As regards conclusive demonstration of net benefits, this is not a sound basis for public policy. We have to decide the best policy on the balance of probabilities, and that balance favors stricter gun laws.

  12. John – I was making observations not complaints. Do you have a problem with my observations?

    I agree that we need to use a balance of probabilities basis in formulating policy. However that still leaves a lot of scope to discuss what outcomes are “best”. And I don’t agree that on the balance of probabilities the 1996 gun controls have any demonstrated net benefit whilst they did have a clearly demonstrated financial cost and a clear reduction in individual freedom. This is not to say that all gun controls are on balance negative.

    The idea that suicide with one bullet is better than suicide with two bullets does not stack up.

    Out of interest during the discussion on this issue it has been indicated that a figure of $2.5 million is the common policy value placed on a human life. Do you know what figure is generally used to account for a reduction in personal freedom and autonomy of decision? Or is this presumed to be zero?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s