Did '96 make a difference?

Simon Chapman, Philip Alpers, Kingsley Agho and Mike Jones have a new paper out today, looking at the effect of the 1996 gun law reforms on mass shootings. Like the Baker and McPhedran paper (whose homicide and suicide findings I blogged on), Chapman et al use time series data on homicide and suicide from 1979-2003, and compare trends in the pre-reform and post-reform periods.

I’m still trying to figure out some of the details of the paper (eg. I’m puzzled as to why the focus is on the impact of the buyback on the trend in the rate rather than the level of the rate). But meanwhile, I’ve posted below the key graph from the paper. If it’s not large enough, the full paper is online here.

This entry was posted in Economics Generally. Bookmark the permalink.

46 Responses to Did '96 make a difference?

  1. ChrisPer says:

    The initial measures all claimed good results because the absolute rate was lower after than before. Then someone pointed out that the absolute rates had been falling since a decade or so earlier, and there was no evidence of an increas in the slope of the trend.

    Myself, I would ask what trends should we compare with. Mass shootings are black swans, and these ‘good persons’ claiming their credit because a man hangs himself or gases his children instead of shooting himself – they make my skin crawl.

  2. Christine says:

    Well: focus on trend because they’re not using ARIMA as did Baker/McPhedran, so they have to deal with the Don Weatherburn and others comment that there was already a downward trend in firearm suicides and homicides pre-1996, and since there was no acceleration of that downward trend post-1997 the laws didn’t really work.

    I’m not particularly sure this paper proves anything either, mostly due to questions of need for a control group, but I also haven’t looked in detail, and I’m off home for the night. I fully expect to see 20 comments by tomorrow!

    PS – did you note that your posts on this have been Wikipedia’d, and by a couple of the commenters on the threads at that (g’day guys, if you’re reading)? And that the Coalition of Law Abiding Sporting Shooters said nice things about you, even while disagreeing? “Other criticisms of the paper were made by economist Andrew Leigh, who pointed out that if the trend broke through the lower 95% confidence limit it would have to be negative numbers of murders, obviously an impossibility as a measure of success. The most important point to make is if it takes someone of the quality of Andrew Leigh to produce a valid criticism of the work, we have to wonder what a good benefit-cost evaluation would be like.” Indeed! Fun and games.

  3. ChrisPer says:

    The mass shootings were actually a different kind of event than the usual run of murders – see http://www.class.org.au/ideas_kill.htm . The reasons they stopped, and the reasons they have re-started overseas, are related to the media framing of the story which they helped create.

  4. ChrisPer says:

    Hi Christine! Thanks for noticing.

  5. Bindy says:

    I was just looking at Table 3 in the latest paper, and comparing the statistical outcomes with the earlier research by Baker and McPhedran. Different statistical methods, same outcomes – no significant change in the decline in firearm homicide, a significant change in the rate of decline of firearm suicides, a significant change in non-firearm suicides. So basically the two papers seem to have independently (or so I assume) come up with pretty much the same general findings. Which is interesting in itself. I have to say that claiming a decline in overall gun deaths is a bit questionable, since looking at the data most of those deaths are suicides. I don’t see any evidence that suicides and homicides were weighted differently to take this into account, so I’m inclined to think that any overall changes are due purely to the results for suicides.

  6. Simon Chapman says:

    On a dermatological scale of skin creepiness, how do others rate ChrisPer’s blithe dismissal of the deaths of the 112 people in mass (4 or more) shootings from 1981-1996 as exotic “black swans”, a mere inconvenience to the verboten possibility that the gun laws might have been associated with any positive outcome? Our paper looked for any evidence of method substitution for both homicide and suicide — none found, so what’s your point, exactly, about people gassing their children?

    The CLASS website is such a hoot: all two of them are mutually priapic about the idea that Port Arthur was caused by people in the media and in gun control publicising the powder-keg potential of the old laws down there to allow someone like Bryant access to the guns he so easily got hold of. The total flakiness of this is mind-boggling, not to mention the interesting way that it lets them worm off the hook for their actions in opposing the reforms that we now know, 10 years on, saw such events disappear.

    Imagine if we were talking instead about people concerned about the easy availability of bomb-making chemicals and who were advocating for these ingredients to be heavily restricted. As governments dragged their feet, someone makes such a bomb. So it now the fault of the people who were raising the alarm about what could so easily happen who are responsible. Not those who stood in the way of reform. Dear me.

    These people are wedded to this self-serving, wise-after-the-event idea that it’s only mad or bad “others” who shouldn’t be trusted with guns. All their hunting and gun club mates are by definition responsible. Thomas Hamilton (Dunblane) was a gun club member, as was the 19 year old from Efurt, Germany who shot 16 at his school in 2002.

  7. Christine says:

    OK, so I’m not home yet, but already this is developing nicely.

    ChrisPer: a pleasure, though I sort of agree with Simon on your skin crawlingness business. No need to attribute bad intentions to anyone, is there?

    Simon: yes, it’s a bit silly to try to say no-one is hurt by law abiding gun owners, who by definition can’t include anyone who uses a gun to kill someone. A little bit circular. But they’ve got some point, still. BTW: add the recent Montreal shooter (Kimveer Gill?) as a gun club member, I think (though everyone at the club thought he was weird? Can dig out the stories, if you like). There was much talk, btw, of how the polices’ past experiences of such school shootings in Montreal meant they were able to reduce the number of lives lost in that incident: just one possible reason for a correlation in deaths across time?

    Bindy: It’s basically the same dataset (not exactly, but close), which is why they’re similar in the way the data looks. The only way to get different results is to use different statistical models. Simon et al use logs, while Baker/McPhedran use levels. Point to Simon et al. B/McP use a time series econometric model with an implicit time trend plus serial correlation, Simon et al assume there’s no correlation across time other than a time trend. Possible points either way, depending on the data. The big difference is really in the conclusions drawn. One says nothing changed after 1996, one says a lot changed.

  8. Bindy says:

    Christine, that is my point – same outcomes. I must have a play with the dataset(s) some time and see what happens with a few other models.

  9. Christine says:

    Bindy: fortunately, both papers print all their data in a table, so you can play to your heart’s content.

  10. ChrisPer says:

    Thanks Simon for your re-introduction of sexual hyperbole into the evidence-based gun debate. Someone said to me once, astonished, that the gun debate ‘wasn’t personal’; but you and Phillip Adams with your ‘priapic’ contempt for your neighbours, and John Howard wearing a bullet-proof vest to meet with our friends chose to make it so.

    In Ideas Kill, I state the NCGC directly contributed not only to the deaths of Port Arthur victims but demonstrably guided the suicide of Allen Burrows. The Coroner found it to be so, even if you dropped it in the memory hole in the frenzy of 1996.

    I do not attribute bad intentions to them and you; I attribute self-serving bias, one of the most basic human cognitive traps and natural for a professor of influence tactics and activism who has spent his career being stroked by peers and media.

    Of course the NCGC contribution to the deaths was not intended – but it was real. The culpable part was not theirs, but the journalists selling advertising time with blood and titillation. They and their industry knew better but chose to act that way.

    I heard you on Margaret Throsby’s program and you sounded such a lovely fellow. That Vladimir Ashkenazy work by Chopin is the one my wife walked down the aisle to when we were married 20 years ago, too. I resolved when I heard you to believe you are a good fellow at heart.

    http://www.class.org.au/ideas_kill.htm In case any readers missed it.

  11. ChrisPer says:

    As to the ‘creepiness’ Simon, I am looking at causes in human nature; if all those deaths could be prevented by understanding the cause and changing people’s behaviour it will be worth the embarrassment.

  12. Peter W. says:

    Most shooters heard of many cases of suicide, following the gun crushing program in 97, of blokes who were shattered after they had to hand in treaured shotguns or rifles, because they were told they could no longer be trusted to own them. In my own case, one of my wife’s cousins hanged himself, after having to seeing several of his late Father’s guns crushed, along with being ridiculed by Police as “another of you crazy gun owners”.
    Many other stories abound of people who had to give up their shooting sport, because of the more arduous requirements…stressful indeed!
    Looking at the Alpers/Chapman charts, (table 2) that has now hit home, with the “suicide by non-firearm” rate having increased from an average 10.00/100,000 up until 1996, to 12.91 (in 97) and 13.09 (98) then 11.75 (99).
    Adding up all those extra suicides, during and just following the gun crushing, gives a total of more than 1000 people driven to suicide by the trauma of being vilified by media and gun control fanatics.
    We are now working on a proposal for a National Monument to be erected in Canberra, to these “Victims of Gun Control”.

  13. Frank says:

    When I walk through an airport metal detector I am assumed to be a potential terrorist. When I get pulled over by the police for RBT, I am assumed to be a potential drunk driver. When the checkout girl at Target asks to look in my backpack, I am assumed to be a thief. I don’t feel affronted by these procedures. I fully support them. Some shooters though, find such assumptions personally challenging. When 90%+ of the community wanted murderous weapons like Bryant used banned, those who can’t bear to be parted from them now want our sympathy and for us all to take their word that they would never do such a thing themselves.

    Peter W asks us to believe that “more than 1000 shooters” topped themselves in the three years (97,98,99) after from the stress and humiliation of being vilified in the media as being potential Martin Bryants. Shooters must be such sensitive people! Perhaps we should worry about about such fragile personalities being allowed to own guns. I mean, what’s the distance from being unable to deal with such “trauma” and killing yourself, to not being able to deal with it and lashing out at the world?

    I looked at the Injury Prevention paper being discussed. Table 2 shows that gun suicides FELL from 382 (’96), 329 (’97), to 235 (’98). Peter, why would all your 1000 suiciding gun owners chose to suicide by non-firearm means do you think? Aren’t you talking unadulterated nonsense here?

    Blogs like this allow some truly astonishing stuff to be paraded. It must be genuinely embarassing for the 99% of shooters who are normal people.

  14. Peter W. says:

    Frank, the airport detector and RBT analogies cannot be compared to the gun bans. You are not stopped from flying to your destination, nor ridiculed by media commentators, or University Professors, should you choose airplane as method of travel! Howard did not say, after the 911 plane attacks, “We must ban aeroplanes, because I can see no good reason why someone would need to travel way up in the sky, when they can sail around the world in a safe boat.” As for RB testing, that does not stop you driving, nor remove your freedom to choose what type, make or model of car to drive! After the 1996 gun laws, many had to give up their sport completely. Those who did choose to fight their way through the multitude of laws and regulations, as I have done, are severely limited in the type, make and model of “sporting equipment” they are allowed to use.
    There are hundreds of examples where the 1996 longarm (and 2002 handgun) laws can be held up to ridicule as having nothing to do with “public health and safety” but to sheer bloody-mindedness of Howard. Why am I able to go deer hunting in Canada or NZ and use a semi-automatic, centre-fire rifle, but not allowed to in Australia? Why can I own and use a semi-automatic 12 gauge shotgun on my farm, but not use the same gun to teach a young girl to shoot at a clay target club? (it is lighter and has less recoil, so would be ideal…)
    As for the “non-firearm suicides” if you are unable to feel compassion for those victimised and vilified, then that’s your own problem. If you cared about restrictions on personal freedoms and resented Government interfence in your personal affairs you would join the good fight against oppression.
    The ones I heard of (first and second hand) suicided by “non-firearm” no doubt because they were decent, law abiding citizens, who had handed their guns in to be crushed.
    Note also, firearm suicide had been reducing steadily since 1979, (as had firearm homicide) so your comment is meaningless.

  15. Frank says:

    But hang on Peter, the buyback only required semi-automatic and pump action shotguns to be handed in — not all guns. Many shooters would have owned guns of both the about-to-be-prohibited classes and and those which were not affected by the law. They could have easily done the job with a non-prohibited gun. Those who were so depressed that they wanted to kill themselves (what, because they could only get one shot away at a deer before it took off instead of spraying dozens at the poor creature, in true sportsman-like fashion?) could have still used a gun in probably nearly every case. But according to you, they all for some reason used another method, given that gun suicide went way down.

    You are painting a picture of people who are so obsessed about not being able to own particular classes of guns that life is not worth living. Do you feel victimised because you can’t own an tank? A ground-to-air missile launcher? Bomb-making equipment? You can still hunt. You can still fire bullets at targets till your heart’s content. You have to admit, someone who has that refined a sense of feeling “victimised” has probably got pretty profound mental health problems .. and as I said, not an ideal person to have any gun.

    Yes, suicides had been falling, but the point that the data in the paper shows is that the rate of decline prior to the laws more than doubled after the laws, with no substitution — how would you explain that?

    With your thousand names for the memorial — can you list just 10 of them?

  16. Terry says:

    Frank, I don’t know if PeterW can or cannot list 10 people who have suicided because of the impact of Howard’s 1996 gun laws, but I know of one. However I am hardly likely to name that person here

  17. ChrisPer says:

    I know one too. I have known a few suicides, and survivors of suicides, but only one personal contact that fits this description.

    Frank, community standards were such that we had a ‘three years hate’, where expressing contempt for shooters personally, as human beings, was quite normal. A lady target shooter I know was told at work she was unfit to breed, an exceptional viciousness for reasons I won’t go into. Thats what the suicides were about, not ‘a choice of toys’.

  18. Bindy says:

    Interesting – I’m getting autocorrelation in both the firearm homicide and firearm suicide data. That would indicate against using negative binomial.

  19. Peter W. says:

    Really Frank, are we getting into a discussion over what is “better” or more socially acceptable; suicide by firearm, or suicide by “other methods”? If you are prepared to accept the conclusions of the Alpers/Chapman report, that the 1996 gun bans resulted in drops in firearms deaths and firearm suicides, then part of that act of faith, is to also accept that in parallel, it “caused” an increase in non-gun suicides.
    BTW, I have several names, apart from wife’s cousin, but not for this discussion.
    From my recall, the ABS a couple of years ago, made special reference to the “drop in gun deaths being matched exactly by the increase in suicide by hanging”. Can’t find any reference in A/C report to that.
    Back to the report and table 2…There has been no account taken of “Manslaughter” in any of the data. A.I.C report 2005 noted that manslaughter accounted for on average 45 cases per annum, peaking in 2000 at 48. That seems to be statistically significant! (Help here please, Bindi?).
    Perhaps the “drop in murders” (gun and non-gun) was just through legal argument and definition?
    While I have your attention, Bindi, there had been one single event of mass murder (n=35) in Tasmania, during the past 200+ years of white settlement. (ignoring the early Government-sanctioned killing of natives). How do we do the statistics on the probability of that happening again, in Tasmania, if such firearms had NEVER been removed?
    How about Queensland, (n=6) where gun ownership is higher and many kept their semi-automatic guns?

  20. Kevin Loveridge says:

    I am licenced firearms owner in NZ. NZ would have been a good control group for the Alpers /Chapman report. Since NZ did not have any bans/buy backs yet homicide and gun related homicide is very low. Australia should look at the NZ model and then they wouldn’t have wasted all their money on the buy back.

  21. ChrisPer says:

    A good suggestion, Kevin.

    We had the case in an early AIC report that referenced a Canadian study. They found a trend in falling violence after a law was passed. What they did not report was that adjoining US states had the same falling trend with no change in laws. The Canadian Auditor General Department issued a report bollocking the original study, but it was already received wisdom in Australia that the Canafdian laws helped.

    Typically these studies provide no clear model of cause (law in actual action) and effect (individual behaviour) which would produce the population-level change. Thats the gap that the Contagion Effect fills.

  22. Christine says:

    Bindy: definite autocorrelation, although be careful with checking for that without using a time trend of some sort as well. What’s the indication against a negative binomial? I see some papers suggesting can’t violate the independence assumption in that model, others suggesting it’s OK, but haven’t encountered it before so am being very impressionistic.

    Peter W.:Re manslaughter: the charts refer to homicide, but as far as I can tell the official description is “Deaths due to external causes: assault”. Other external causes are “intentional self-harm” (ie suicide), “accidental” and “undetermined intent”. These do not correspond to legal categories. From the descriptions in the accidental category, I believe this is something that would not include manslaughter. By default, manslaughter should therefore go in ‘assault’. But someone with more knowledge of the data should correct me if I’m wrong. ICD-10 classifications of death are here (not fun reading, btw):

    Re statistical significance: there’s almost no way that an increase in three deaths per year from 45 to 48 counts as statistically significant – these data series are way too volatile.

    The question of mass murders is a tough one, because they’re low probability events. Generally, what you do is to calculate a probability of such an event occurring in any given year, using some baseline (pre-law change) data set, and then calculate the probability that another such event doesn’t occur after the intervention. Eg if there had been 13 cases in the 18 years to 1997, then using those figures the probability is 0.72 for a mass shooting in any given year. The probability of no recurrence in the next 8 years (98-05) is 7.5%, assuming independence of the events (which as Bindi points out is unlikely). So it’s low probability, but not zero. After 15 years, the probability is under 1%. If there are no more mass shootings in the next decade, or even if there is only one, I’d probably be happy to attribute this to the gun laws.

    Two big problems with this are: (1) post-law change period may differ from pre-law change period for a variety of reasons other than the law change, and this might affect the probability of mass murder; and (2) if the events are very low probability, then it’s hard to get a good estimate of the probability in the first place, and you may have to wait a very long time before you can tell if the intervention worked or not. There’s no way anyone in their right mind would look at mass gun murder in Tasmania if there’s only been one incident pre-1997 and that was in 1996, and then try to estimate the probability of recurrence.

    For what it’s worth, Peter W. and ChrisPer, I am somewhat in agreement with Frank re the argument that the gun laws had negative effects because some former gun owners suicided after being forced to hand in their guns. No matter how tragic the individual cases (and I have no doubt they were tragic, and my sympathies to all affected), the idea that a large proportion of gunowners felt so distraught at being forced to give up their guns that they killed themselves doesn’t seem to me to be something that backs up the argument that most gun owners are just like most non-gun owners, and are therefore not likely to cause any harm with their guns. If you think that non-gun owners don’t understand gun owners, and you are trying to bridge that understanding gap, I don’t think it is in your interests to make this point, and especially not to suggest that the suicides number in the thousands.

  23. ChrisPer says:

    You make some very good points. I don’t know why Peter puts the number of buyback-related suicides at order of a thousand, but knowing someone personally may tend to an overestimate. My own feeling is that based on the tendency of this kind of story to spread, the number is likely to be order of tens at worst. That is roughly the equivalent of the Port Arthur Massacre over three years, as the most extreme response to the three-year nationwide – erm – expression of decency – against a minority of roughly one million people.

    It is most interesting to me though how you read a different message to the one the writer thought he was sending. We are making a point that for three years, the over-the-top contempt Simon shows in his note above was normal from much of the community and particularly the modern values enforcers – the media, their partner activists, the public sector and politicians.

    People grew up, and quietly revised their values, around the time of the Sydney Olympics with their inclusive pride in Australian identity. But let me ask you:

    There remain 850,000 licenced Australian shooters. At least 120,000 of them care enough to be members of the SSAA, and many write letters to the editor and politicians and all that ordinary activist stuff. Feelings are very high, and remain so.

    In the last ten years, how many Australian shooters have assassinated a politician or activist for their political actions? How many did before the 1997 laws?

    An Australian leftist hit a pensioner in the head with a brick for going to a political meeting they didn’t like. How many shooters have done that? Did our demos trash the place like some demonstrations, or have we threatened people with murder, beatings and rape like certain union officials have?

    How many have set off truck bombs in a crowd of civilians? How many ever have? How many have gang-raped innocent women, or beheaded hostages, or shot up cars outside Sydney churches and rampaged through suburbs smashing windscreens, hurling bricks and bottles at cops?

    Say it out loud for us. Say the number. Like this: “____ ordinary shooters have assassinated, murdered, firebombed, raped, robbed and run violent riot through Australia in the last ten years.”

    Shooters are your neighbours, your business contacts, your students, we are your family. We do not deserve the prejudice that the gun control discourse imposes.

  24. ChrisPer says:

    Sorry about my emotion. For you guys its an econometric problem, I guess.

  25. Patrick says:

    Never underestimate the power of undisclosed preferences to distort/influence econometric or other ‘technical’ analysis ChrisPer!

  26. Christine says:

    ChrisPer: I do understand (though I’m sure not fully) your emotional reaction to these issues. One reason I’m treating it as a purely econometric problem is that that is where my expertise is on this topic, and because the discussion started off because of a published paper claiming to some statistical results, and arguing those results lead to strong policy conclusions. I’ve been trying to drag the conversation back to evidence because I think we need to have good analyses in order to make good policies. This is something I think all the commenters have agreed on, and I think it helps to have a bit of common ground in what is obviously a hot button area.

    On my previous comment showing that I read your point about suicides of gun owners differently from the way it was meant: that was the point I was making.

    Some questions for you: is the primary concern of gun owners with post-1997 that they were vilified or that they were forced to hand in guns? Could we have had one without the other? Are advocates of gun control necessarily imposing prejudice on shooters?

  27. ChrisPer says:

    Christine, thanks for your questions – and your dispassionate approach!.

    1) For me, the vilification mattered far more. I was not actually a gun owner at the time; I had got rid of them all 8 years before on a spiritual challenge, though I still had an interest. The 1996 thing was so offensive that I decided to get back in, to own weapons suitable for ‘social work’ as reactance against the prejudice, and to teach my kids to shoot as they grew up. As it turns out I got far more interested in activist psychology – why the hate thinking is what it is – and the evidence trail on benefits and costs. And reading, tennis and windsurfing are what I do for fun in real life.

    For 650,000 others, the loss of individual firearms was an extremely significant issue. The dicks=guns thing is rubbish, but their symbolism is important in ways that are very heavily discounted in ‘cosmopolitan’ values. I would compare it in enthusiasts with the significance of a family farm, or a lifetime home.

    2) To the question ‘could we have had one without the other’, I believe not. Vilification is born out of the root motivations of gun control politics, and our media do not have a ‘mute’ button.

    3) Are advocates of gun control necessarily imposing prejudice?

    Well, many who support the control measures view them as ‘just common sense’ and ‘not personal’. But the first flash of emotion when you meet authoritative names is surprising – extreme distaste. I am not talking activists, but mainstream criminology professors and the like. The behaviour of journalists and radio presenters often seems to be based on moral status display, not examination of facts.

    This issue is seen in moral terms by gun control proponents because it is a marker issue for the ‘cosmopolitan class’: that is, self-identified ‘educated’ or ‘enlightened’ people see it as necessary to agree with them to view you as ‘good’ people. It comes as a set with anti-racism, pro-choice, pro-feminism, pro-gay, pro-environment values. The best of people on both sides can address differences with courtesy and depth of knowledge, but journalism and ordinary peoples’ need for for the shortcut produces extremely polarised us-them thinking.

    Interestingly, the thinking of shooters is not ‘enemies of the above’. Most of them see themselves as community backbones, aligned with respect for those values , but with a ‘traditional’ slant toward self-reliance and positive masculine social roles. The criticisms of them as a group are based on projection, plus a very small rate of offenses that feed confirmatory bias.

  28. Frank says:

    Christine, several of the gun lobby correspondents on this page are doing their best to project an image of shooters as thoughtful, decent people, open to evidence. Many … most … obviously are. But if you want a glimpse of the more frothing end of the shooting movement — and as you read their remarks, remember, these are people licensed to own guns — take a read of the stuff here http://www.synect.com/forum/index.html Very scary!!!

  29. Bindy says:

    Well said Christine, let’s get this back to evidence! Is it just me or do the conclusions tend to overlook what emerged from the analysis of relative trends? The conclusions appear to be based on relative rates, not relative trends. I guess I’m just finding it a little odd that the results from Table 3 really aren’t discussed at any great length, so in light of that I think I’ll reserve judgement about whether the conclusions this paper reaches about the buyback can be sustained. I’m also not sure why Table 3 contains redundant analyses – if you have a significant decline in firearm suicide and non-firearm suicide, why analyse whether you have a significant decline overall? It’s a foregone conclusion. I’d have thought some weighting would be needed for the analysis of total firearm deaths, also, to accommodate the number of suicides in that total, but there isn’t any mention of differential weighting or some other control.

  30. ChrisPer says:

    re http://www.synect.com/forum/index.html

    Very scary? Man you are drawing a long bow. There are a couple of loose speakers, but we freely correct each others excess there. Basically it is between shooters, we are ‘at home’. Violent talk, offensive treatment of women or other races is not tolerated; but politically incorrect free speech is. I have been much corrected by people there, and particularly recall someone setting us straight after someone dissed gays, I personally discredited conspiracy theories of Port Arthur, and with many members I slammed a post which used a metaphor of physical violence for verbal activist conflict, and it was deleted toot suite by the moderator.

    That forum has had at least 3 incarnations. A previous one was slammed on Crikey, by a writer who put in his own text to make our writing seem racist and violent.

    Different audience here, we are very much your guests.

  31. Compass says:


    is your post some type of red herring to prevent close exmination of Christine’s valid comments or because you just don’t like gun owners?

    I prefer the evidence based discussion.

    Two papers have now been published with the same statistical results (firearm homicide NS, firearm suicide S, but the trend for homicide and suicide without firearm also S) and I was particularly interested in Christine’s:

    “Two big problems with this are: (1) post-law change period may differ from pre-law change period for a variety of reasons other than the law change, and this might affect the probability of mass murder; and (2) if the events are very low probability, then it’s hard to get a good estimate of the probability in the first place, and you may have to wait a very long time before you can tell if the intervention worked or not.”

    From a socio-economic point of view this suggests that the Chapman et al. paper have overlooked some serious issues which would cut straight across the claims that destroying some 600,000 plus guns (at a cost of over 500 million) was in any way useful. Some of the possible other issues were highlighted in the earlier Baker and McPhedran paper and it remains to be seen if an economist would be brave enough to compare (or even try and seperate out) the effects of buy back with the initiatives of mental health programs and several states introducing ‘tough on law and order’ campaigns, which included truth in sentencing post-1996.

  32. ChrisPer says:

    Compass, interstign comment thanks.

    I think that it is extremely doubtful that destroying all those guns did anything constructive. But I think the absence of mass shootings since the buyback is critical to interpreting what really happened.

    The buyback per se is unlikely to have done anything. The media publicity and advertising and vilification of shooters however probably contributed by over-writing in people’s minds the earlier media message that massacres were easy and rewarded the perpetrator with perverse glory.

    Laws harmonising license categories and requiring a shooters licence after a background check, and safe storage, could have probably made a big difference to public safety without spending the half billion and engendering all that hate.

  33. Peter W. says:

    It is impossible to undertake a true “scientific” study of the effects of the 1996 gun bans, since there was never any inquiry into the events which lead up to and “caused” Martin Bryant to murder 35 people. After the tragedy, many of the survivors and relatives of the victims, asked for a judicial inquiry or government investigation. Many pointed to a failure of local police to act when M.B. had been causing disturbances firing his guns late at night. Other experts highlighted the influence of M.B.’s collection of violent and sadistic videos and called for severe censorship of such films. Others referenced the Ch 9 A Current Affair “How to get a gun and shoot people” film which aired only in Tasmania. (Chris already mentioned contagion effect). Yet still more experts called for a review of mental health services, and questioned the Richmond Report of a few years earlier.
    However, all those possible/probable causes were ignored.
    If a full and detailed investigation had been carried out in 1996, it may (or may not) have led to a decision to restrict gun ownership, but there would have also been other recommendations, especially regarding the identification, diagnosis, treatment and monitoring of those with mental health problems. (perhaps also charges against Ch 9 producers over the ACA program).
    Since there was never any full and thorough investigation, any conclusions 10 years later cannot have any credibility. There are far too many variables; the A.I.C. noted that advances in medical assistance have saved many people who in previous years would have died. Most centres where crowds gather are now monitored by CCTV and/or armed security guards. Meanwhile, the mentally ill still roam the streets killing and maiming themselves and others!
    ABC TV 7.30 report of 17th may 2005 emphasised that 36 murders in NSW alone, during the years 1999-2003, were committed by mental health patients! Hey! That’s more than died at Port Arthur, murdered in a short period of time, in just one State of Australia.
    What Howard’s Gun Bans and fans of Gun control have done is divert attention and resources from the real problems in our Country.

  34. ChrisPer says:

    I really disagree that it is impossible to undertake a good study for actual evidence of the effects you speak of, Mr Whelan.

    We could estimate the effect of combined better medical treatment, mobile phones and other advances that reduce deaths, by checking ratios of injuries to deaths in auto accidents and other injuries.

    We could disaggregate stats of woundings, suicides and murders by ‘narrative elements’ to correlate and test contagion effects and times to treatment, if only we could get the full files from the police and ambulance services. Research permission for that might be very hard to get and the time involved quite expensive.

    Nevertheless, the way the AIC paper on measuring the effects put it, they were as helpless as the note above suggests; yet they are the ‘official umpire’.

    Today our newspaper has the faces of three girls, two of whom killed the third. This kind of crime is so rare, yet in the catchment of this newspaper the same story was published a few months ago – two girls killed a third. A number of elments are common. The effect was immediate, too – a retarded boy aped and strangled a young girl in a shopping centre disabled toilet just one week after all the publicity.

    Contagion effects are strong, effective and totally denied by our yellow press. The NCGC contribution to causing the Port Arthur Massacre is real, as partners in the sensationalism of the press.

  35. Frank says:

    Peter – -can you please let us have the details of all the gun massacres that occurred in the last 20 years where many people were seriously injured, taken to hospital with gunshot wounds and survived thanks to these amazing advances in hospital procedures you refer to. Spare us some secondary reference alluding to general improvements in medical procedures. I work in the public hospital system and believe me, you’re in fairyland here pal. Gunshot wounds, as I’m sure you know, can be very, very ugly. Anything but the gun laws, eh?

  36. ChrisPer says:

    Frank, the research is there to support it about trauma in general and murder in particular.

    In the case of Port Arthur, for instance, 35 dead and 37 seriously wounded I think. Thats up to 72 dead, if we did not have quality health care and antibiotics. You do good work in the hospital system, and it is getting better.

  37. Frank says:

    Chris — I think Peter is saying that better medical care is a variable responsible for the gun death rate declining (“advances in medical assistance have saved many people who in previous years would have died”). If there research is there, where is it? I’m asking him to point to examples of mass shootings since Port Arthur where people were injured but saved by this better medical care. There have been no major advances in anti-biotics since 1996 that I’m aware of … which advances are you referring to?

  38. ChrisPer says:

    Interesting how you are determined to confound different categories. Death rate from shooting wounds may have declined, but you would exclude it because you only asked about massacres?

    Good quote here:
    From Classical Values blog:
    “I soon noticed that there’s a downside to debunking fraudulent people or claims. The people who make them up — and most of those who agree with them — simply don’t care. Because the characters and claims are invented to support what they already believe fervently, debunking them does not “count.”

    Lies presented in furtherance of a greater “truth” are not really considered to be lies, at least not in the moral sense. The idea is to persuade people, and if fictional people or incidents have to be used, that’s OK, as long as it’s in the interest of the greater truth.

    The problem I have with this approach is that I don’t like being lied to. Even when I agree with the cause the lie is intended to support. I don’t find lies emotionally fulfilling because they pollute the process of thought. When lies are presented as “news reports,” it’s even worse, because it makes me distrustful every time I pick up the paper or turn on the television.

    It is time we addressed the behaviour of media and how they contribute to violence by showing the way for weak minds.

    Today’s paper has the face of a girl killed by two other girls; a rare class of murder indeed but the indications are that they were imitating a murder in our state a few months ago. Just a week after the first one a little girl was also murdered by a retarded boy, very likely influenced by the media coverage.

    Ideas can kill, and the National Coalition for Gun Control seems to have helped the irresponsible media give Martin Bryant his idea.

  39. ChrisPer says:

    Whelan referred to a report issued by the Australian Institute of Criminology, Homicide in Australia
    2001–2002 National Homicide Monitoring Program (NHMP) Annual Report

    A.I.C report no. 46, http://www.aic.gov.au/publications/rpp/46/ “advances in emergency medical care (these may have saved many serious assault victims who might otherwise have died; see Harris et al. 2002);”

    Harris, A.R., Thomas, S.H., Fisher, G.A. & Hirsch, D.J. 2002, “Murder and medicine: The lethality of criminal assault 1960–1999”, Homicide Studies, vol. 6, no. 2, pp. 128–66.

    From Britain:
    http://news.bbc.co.uk/2/hi/health/2419705.stm The rising level of gun crime on Britain’s streets is putting increased pressure on hospital staff who have to deal with the victims.
    But the medical experience from regular shootings and new training means victims are more likely to survive their injuries than a few years ago. …

    Professor Ryan said the fall in mortality rates was due to improvements in the chain of care.

    He said: “The fact that we have improved pre-hospital care, care in the emergency room and now improving care in the operating theatre, I’m not surprised morbidity and mortality is going down.”

  40. Frank says:

    Chris, you say that I am trying to confound matters by asking for clarification about how medical advances have reduced deaths from massacres. Recall that this entire discussion was precipitated by Andrew posting on the Chapman et al paper which made the blind-freddy observation that since the law reforms there have been zero massacres, something which Christine points out is of exceedingly low probability given their occurenec in the previous decade (ergo something is likely to have caused that change)..

    Peter on Dec 22 wrote above about the Bryant massacre & in that context, advanced improved medical intervention as a confounder. I asked him to then let us know how many mass shootings had occurred since where the wounded survived rather than died because of improved medical intervention. Of course, there have been none.

    I have had a look at the Harris paper you helpfully cited. It of course deals with all forms of deadly assault/attempted murder (stabbings, bashings as well as shootings). It looks at data from 1960-1999, a period in which I agree there have been immense advances in emergency medicine. But Peter advanced this notion as a confounder in the context of an explanation (I think… he is very hard to follow) of the post gun law changes in gun homicides (ie that part of the reduction .. which he also argues has not taken place!) .. is due to improvements in medical care. There simply have not been significant increases in emergency care since 1997.

    As we know, Peter is fast and loose with his facts — to put it mildly. He’s the one who told gave us the priceless line earlier in this blog that he had calculated that 1000 shooters had suicided in a couple of years after Port Arthur and was about to erect a memorial.

    I’m signing off & going on holidays. Happy & peaceful Christmas to all.

  41. Bindy says:

    Frank, I think you may have misunderstood Christine’s entirely valid point about mass shootings.

    It was not that the low probability indicates something has caused a change, but that it’s hard to know if there has been a change because we’re dealing with a rare event with low probability.

    She’s quite right in that regard. I’d like another five or ten years of data before accepting the claim that the gun laws have prevented mass shootings.

  42. ChrisPer says:

    I guess the difference is Frank, we knew the ones who died. I don’t believe the 1000 figure, but I knew one personally. So did Peter.

    You are confounding obviously different things for rhetorical effect.

    The Baker & McPhedran and Chapman and Alpers studies both looked at overall stats of deaths, as deaths are relatively easy to get data on and have high political value. These figures are affected by anything that converts possible deaths to survival of individuals. That is the evidence-based present discussion.

    The fact of no gun massacres is to me more than interesting. It is a confirmation that our assumption of the causes of massacres at the time was wrong. We thought it was the evil of humanity plus the availability of guns.

    Well, only 40-60% of the banned guns were thought to have been handed in so they are still plenty available, and in the hands of LESS law-abiding people. Notice how bikies seem to still have the good gear when its siezed, and the serious drug dealers still have semi-auto rifles if we believe the seizures reported in the press. So semi-auto guns are still in the hands of violent people. Ordinary people still have all the other kinds, and they can be used in good or bad ways according to their choices.

    People still have all the evil they used to, that hasn’t changed.

    So why no massacres? John Howard and Adam Graycar said they ‘didn’t pretend for a minute that these laws would prevent future massacres’.

    What did?

    The assumptions about gun availability were wrong. The assumptions about ‘evil in the nature of man’ was wrong. The contagion effect theory gives us the reason. The press turned around and gave the deviants a new story. No longer was it ‘guns are easy to get, and here’s how to get millions worth of fame and free publicity! You can try this at home!’.

    The new message – the publicity of the buyback, and the public idea that we somehow ‘fixed’ the gun laws – sent a new message. ‘Massacres don’t happen here now.’

    It took the combination of mental illness, sloppy police checks of a foreign student’s bona fides, and weeks of publicity for a couple of no-hopers in Washington killing innocents, to create the new conditions for another massacre – but it was stopped as it started by bystanders intervening.

  43. Peter W. says:

    Frank: Well, I’m still waiting for your explanation as to why in 1997 did the “non-gun suicides” rate increase so dramatically and increased again in 1998 and was still high in 1999?
    As for the analysis of “mass murders”, the paper under discussion is a good example of “post hoc, ergo propter hoc”, or the comon fault in logic which assumes that because one event followed a certain event, that the first event was the CAUSE of the second event. e.g. When you wash your car on Sunday and it rains on Monday, you would claim that your washing the car, CAUSED it to rain!
    What I am claiming is that if there had been no buy back, gun massacres may have still have not occurred during the past 10 years.
    I used “advanced medical procedures” as an example of just one of the variables. Other advancements which may have influenced the outcome include: most people now have mobile phones, so can call for asistance: most popular centers are now patrolled by security guards, CCTV monitors main areas, etc.
    A Merry Christmas and Happy Hanukkah to all.

  44. Peter W. says:

    Strange how sometimes an outsider with a fresh approach can put an entirely diferent slant on this topic! At a post Christmas party I got into discussion with a friend who is a non-shooter, not involved with shooting, nor the various complex pro-or anti-gun arguments.
    He knew my passion for freedom, gun rights etc, so the topic of “no more gun mass murders” came up. I started on the detailed statistical analysis involved, method substitution etc and how the Gun Control groups had claimed the gun laws had worked and claimed a victory.
    He stopped me short by saying, “But hang on! All you gun owners have to do is hope that there is another Martin Bryant out there, who didn’t hand in his guns, so he can go on a rampage again. Then the gun lobby can claim that the gun laws did NOT work, we can get rid of the gun laws and you can get all your guns back and get on with your lives.” I was stunned at this concept, that gun owners would be hoping for another mass murder by gun, until my friend went on, “But no! If there was another mass gun murder, the gun control people wouldn’t see that as the gun laws having failed, but as proof that we needed even tougher laws!”
    We laughed and had another beer.

  45. ChrisPer says:

    A reply to Chapman and Alpers work has been published.


  46. ChrisPer says:

    Baker and McPhedran have written an analysis of Chapman, Alpers at al.

    Click to access Baker-McPhedran_critique_of_chapman_et_al.pdf

Comments are closed.