What Happened, How, and What Have Been the Results

updated 1 June 09


Long Term Homicide History pre '87

In 1987 the rate of homicide in Australia had leveled off (even higher than at the turn of the century) for about five years, after rising for about 35 years from a low during world war II.  Reported rape rates had increased about threefold since the mid 70s, robbery rates had more than doubled, and rates of serious assault had increased exponentially by a factor of about 4.5.1,2  Inflation had been very high throughout the 80s.3  There were a high number of labor disputes, although only about half as many per year as had occurred throughout the '70s.  A mid-1986 poll had found that violent crime was the number one issue for Australians.

In 1987, Australia experienced at least 32 homicides in a total of at least six multiple-killing incidents—including the Hoddle Street and Queen Street (Melbourne, Victoria) "massacres."  In a December "gun (control) summit," immediately after the Queen Street incident, Australian Prime Minister Hawke, the state premiers, and the Chief Minister of the Northern Territory agreed to establish and fund a "national committee on violence" (NCV), which was charged with investigating Australian crime, reporting, and making recommendations by the end of '89.  (The fact that the leaders had the "summit" showed that they already knew they wanted more gun control.)  Two of the states refused to go along with the proposed national approach (probably needing a "study" to justify it to their citizens).  But, Victoria and the Northern Territory immediately changed their laws to tighten control over semiautomatic firearms starting in '88.

A couple of months later (Feb 88), the Australian Institute of Criminology (AIC)— part of the federal government— released No. 10 of a "trends and issues" publication "Firearms and Violence in Australia" authored by a Duncan Chappell (director of the AIC) and three of the AIC staff.  In it the authors examined claims released in '83 by the Canadian government for three years of results following the 1978 start of new gun restrictions in Canada (that the authors erroneously identified as bill "C-15").  They falsely represented the Canadian law as including firearm registration.  They said gun suicides and gun homicides dropped after 1978 and that the proportion of violent crimes committed with guns declined.  And they said that the reduction of gun homicide was almost matched by the reduction in overall homicide, so people had apparently not substituted other methods to kill in the absense of guns.

The proportion of crime committed with firearm is not a valid measure of policy effectiveness because such measures ignore the very real possibility that criminals will simply use other tools and that crimes can be prevented by the fact that people other than the criminal minority have firearms.  And the use of suicide with firearm as a measure of effectiveness is not valid primarily because those who want to commit suicide will just use other means if their preferred means is not available.

The valid measures of policy effectiveness are crime rates and suicide rates, not gun crime rates and gun suicide rates.  Canadian suicide rates have not varied significantly from 1977 up through at least 2000 in spite of the C-51 law, the more restrictive C-17 law of '91, the even more extreme C-68 law of '95, and anything else Canada has tried.

Regarding their claim about homicide rates:

1. There was no absense of guns.  Canadians didn't turn in their guns or destroy them.  The new law restricted only purchase of guns, not their possession.

2. Examining their figure showing the history of homicide, gun homicide and non-gun homicide, it is apparent that the reductions (and the level state of the non-gun homicide) the authors mentioned had begun three years before the law went into effect.

The paper then asserted that, "a reduction in the number of firearms in Australia, and further restriction on their availability, would undoubtedly prevent considerable death and injury."  And it recommended a number of controls which would have "two main thrusts," one being "reducing the number of weapons in Australian society."

On July 7, 1988 professor Chappell signed a memorandum of understanding (MoU) on behalf of the AIC and, hence, the Australian federal (commonwealth) government.  The other party to (signer of) the agreement was a Margaret Anstee, head of the Crime Prevention and Criminal Justice Branch Centre for Social Development and Humanitarian Affairs, a subsidiary of the U.N.

Rather than addressing a need to scientifically investigate the relationships between different kinds of violence and private possession of firearms, a matter that had hardly been studied at the time, the agreement between the branch center and the AIC declared a need to severely restrict the private possession of firearms in Australia, and described a dozen or so measures to be undertaken—such as registering members of all gun clubs and banning all self-loading centerfire rifles and shotguns.

The National Committee on Violence (NCV) met first on 19 and 20 October, 1988.  The chairman of the committee was the same Professor Duncan Chappell, director of the AIC.  He and the AIC no doubt provided much useful or necessary "information" to the committee, and connected the committee with "good sources" of information during the following year as the committee solicited input and had five further meetings and eight community forums around the country.

Immediately upon formation, the NCV set about to produce some publications in a "Violence Today" series.  Number 1 of these, "Violence, Crime and Australian Society," authored by professor Chappell, was disseminated by Feb '89 after being given as a presentation to the South Australian Police Academy on 13 July 88.  In this paper he stated that gun control would be more effective than tougher sentencing and additional police resources, and that a powerful gun lobby had blocked some state attempts at gun control.

There were invited speakers at the NCV meetings and community forums throughout '89.  Gun Control Australia and the Shooting Sports Council of Victoria were represented at the fifth committee meeting on 22 and 23 July 1989.  And the Tasmanian Coalition for Gun Control spoke at the community forum in Tasmania.4

The committee received considerable participation from some churchs, police commissioners, domestic abuse organizations, media, health and welfare organizations, education institutions, health practitioners, anti-violence organizations, child welfare advocates, numerous municipalities, two gun control organizations, and two sport shooting organizations.5  Gun owners would say later that their inputs to the process were ignored.

On 10-13 Oct 1989 the committee hosted a "National Conference on Violence" in Canberra.6  It had one workshop/panel on "Firearms and Violence."

Enter the Australasian Police Ministers Council (APMC).  The council was a committee of top law enforcement representatives (ministers) from the commonwealth government, each of the states and territories, and New Zealand.  The name of the council was (reportedly) not just "Australian" because the council included New Zealand representation.  The council had entertained firearm "reform" in 1987 (and again in 1990) according to the now defunct Australia "buyback" web site "history" page, which also indicated that commonwealth governments (leadership) prior to the one that came to power in 1996 had been unsuccessful in their efforts to bring the Australian states and territories together to implement "effective" nationwide firearms controls.  Note that this history implies that it was a given that such controls were needed and would be effective.

In 1989, the state of New South Wales (NSW) passed its Firearms Act of 1989, so NSW had firearm registration, owner licensing and virtual gun bans starting in 1990.  The next year the state of Queensland passed its Firearms Act of 1990, so some kind of owner licensing began there the next year ('91).  Handguns and fully automatic firearms had been severely restricted almost throughout Australia for several years already.  In August, 1990 five people were shot to death with a shotgun in Surrey Hills, Sydney, NSW.

In 1990 the AIC published the NCV report on NCV findings and recommendations.  The report, Violence:  Directions for Australia included many recommendations for violence reduction, including 13 for gun control.  These gun control recommendations were similar to what had been described in the AIC-UN MoU, but not the same.  In the section, before the recommendations, the report said as part of the justification for gun control:  "research has demonstrated the death rate for victims assaulted by [sic] guns is several times that of those assaulted with lethal intent by [sic] knives or other weapons."  Here was yet another instance of unwarranted reliance on information from the medical community, which fails to take into account that murders using knives do not involve victims going to a hospital (and dying there or on the way) to as great an extent as murders using guns.  Very often shooting a gun at someone results in people in the area becoming aware immediately that something has happened, because they hear the gunfire.  This results in aid being given to victims shortly after being shot.  A person killed with a knife can easily lie dead for hours before being discovered.  So, for '01 through '07 in Australia, attempts to murder using a gun were successful in 35.6% of cases while attempts were successful in 43% when using knives.

The report introduction erroneously stated that "firearms are used in approximately 40 per cent of homicides."  At the time gun usage in homicides was only about 27 percent, down from a high of about 36 percent around '65.

Near the end of the info on Police statistics, under Sources of information on violence in Australia the report said that the then current condition of police statistics on the use of firearms in crime "prevent governments from monitoring systematically the effects of current firearms policies."  So information was not available from which one could determine if a law change would have an effect.

A topic on weapon use in the report, near the end of a section on Risk of violence in Australia again restated the inflated portion of homicides committed with firearms and said the committee viewed the use of firearms in family violence and mass killings "with the gravest concern and considers that weapon use in violent crime is a major issue,"—[but apparently an issue that had already been settled].  In the conclusion to Part 1, the report acknowledged the importance of peoples' fear of violence irrespective of any actual risk of violence—and said, "Frequently part of the explanation for this [fear] lies in near-hysterical headlines in Australian newspapers."

As the proposals looked when funneled through the gun control media to the public, they were generally such as would seem quite reasonable to people who themselves had no interest in firearm possession or knew little about the connections between firearms and violence—or were naïvely trustful of government or accustomed to control by governors.  We often see that "the devil is in the details" and the details are never in the generalizations publicized by our gun-controlling politicians.

On August 17, 1991, a 20-year-old man shot and killed six people, stabbed to death another, then killed himself with his semiauto rifle in a shopping plaza in the Sydney, NSW suburb of Strathfield.  Note that carnage was being produced with firearms more lethal than handguns since having a handgun was extremely difficult because of past preoccupation with restricting handguns.  Note too that these two high-profile "mass" killings occurred in NSW after NSW instituted much tighter restriction of long gun possession.

After the mid-August Strathfield incident, the NSW parliament quickly established a "joint select committee upon gun law reform"  The term "select" apparently did not refer to qualifications of the legislators (ministers) who participated, as at least one of them was qualified only by her father and brothers having once been hunters.  The committee quickly put together a set of recommendations which were accepted by the cabinet before mid-November.7  The committee must have received some help from someone to get so much done in so little time, especially with unqualified participants.

Before '91 was done, NSW amended its firearms act 1989 to make changes recommended by the committee—including prohibition of firearms possession by a person under a domestic violence restraining order (called an "apprehended violence order," or AVO, in NSW) which might be issued without existence of any violence, and would result in prohibition of gun possession for ten years after the end of the order AND with no possibility of judicial discretion.

In a 1991 police ministers meeting, according to parliamentary library Current Issues Brief 16, 1995-96, the ministers considered the need for uniform national firearms laws, one of the NCV recommendations, and reportedly reached a "concensus" based on the recommendations of the NSW "select" committee.7  Later cooperation of the AIC with the council indicates that the AIC was probably involved with the council at this time as well.  Recommendations for "reform" were drawn up as a result of the 1991 meeting.  According to a '99 AIC document, sale of "military style" rifles was banned, starting about '92, in all states but Tasmania and Queensland.8

The next year, on October 27, 1992 a man shot to death six people in three towns north of Sydney on the central coast of (again) NSW.  On August 26, 1993, three people were shot dead by a man in another suburb of Sydney, NSW (again).  In the '92 annual meeting, the Public Health Assn of Australia declared their support for policies such as other gun controllers had been advocating.

In 1994 Duncan Chappell left the AIC and went to work directly for the UN's UNICRI (home of Martin Killias) for a year before returning to Australia (ending up at the Australian Research Council "Centre of Excellence" in Policing and Security, Griffith U.  Also in 1994 the Australian Institute of Health and Welfare said in "Australia's Health 1994" (p109), regarding suicide:

"Currently, the question of the potential benefits of more stringent firearm controls is being debated.  Although stingent gun control has plausibility as a possible intervention, available data and research cannot provide a clear picture of intended and unintended consequences of such a measure."
At its meeting in May 1995, the police ministers council noted the "varying registration and licensing systems applicable to firearms and firearm ownership within the States and Territories to control the use of firearms."  It established a team with members from each jurisdiction, and coordinated by the Victoria police minister, to "identify mechanisms which would achieve uniformity" (i.e., "like the most restrictive states") in a number of areas, including:  a system to control mail order sales; uniform standards for training prerequisite to licensing; uniform standards for security/storage; and a uniform approach to pistol registration.  The Victoria minister reported on subcommittee activity in the 10 Nov 95 council meeting and further consideration was set for their February 1996 meeting.9  It has also been reported that a commonwealth representative presented a proposal at this meeting.

At this time gun laws or their enforcement were apparently most restrictive in NSW and Western Australia and least restrictive in the states of Tasmania and Queensland, although it appears that Queensland had quite restrictive law and it is hard to tell for sure based on the conflicting and incomplete information from newspapers and gun controllers on the one hand and gun rights activists on the other.

Tasmania had passed their Guns Act 1991, and its restrictions apparently began January 1, 1993.  These restrictions included requirements that a person have a license for possession of a firearm, and that handguns be registered.  After the restrictions began, use of firearms in assaults increased somewhat steadily for four years (through '96).  Firearm suicide rates dropped from the peak they had reached in '92, back down to pre-92 levels, and so did overall suicide rates.  The drop, then, was caused only by the '92 peak rather than the guns act.  No trend can be discerned regarding homicide in the time period because the numbers were too small and variable for any reasonable level of statistical significance.

In early '96, Australia had been having some bad social and economic problems for several years.  Inflation had been very high throughout the 1980s.3  Major banks had failed in 1991.10  There was an almost nationwide recession in the early '90s.  There had been a large influx of immigrants from Asia, and unemployment had been very high for immigrants from non-English speaking countries.  Both Asians and Australian aborigines thought themselves poorly treated and had demonstrated against discrimination toward them.10

Unemployment had been very high for several years.3,7  Long term unemployment was high and many males had just stopped trying to find work.  Relative poverty, an indication of inequity within the society, was very high, exceeding that of Canada, England and New Zealand and exceeded notably by the considerably higher relative poverty in the U.S.A.11  Alcoholism, unemployment, poverty, crime, etc. were very high among aborigines.  Youth unemployment was very high.  The divorce rate and portion of households with only one parent had been increasing fast for several years.12  Productive (manufacturing and natural resource) industry had been in decline for years.  [Note that several of the problems just described are generally accepted as being strongly contributing causes for most types of violence and suicide.]

The public was naturally unhappy with the state of affairs, which they naturally blamed on their government.  A federal (commonwealth) election was called for, so the February 1996 police council meeting was postponed.9  A new federal government was formed, headed by John Howard.  An issue to distract the public from real issues, and serve as a unifying force, would certainly be helpful at this point.

In January, 1996 seven people were shot dead in a Brisbane, Queensland suburb.  Then, on 28 April 1996, 32 people were shot to death, 19 were seriously injured, and three others were otherwise killed near Port Arthur, Tasmania.  The public was immediately whipped into a gun control frenzy by the press, the medical community, the Australian branches of the international gun control cartel, the commonwealth government (including their "institute of criminology") and others.  Although polls done prior to the massacre indicated that the public was satisfied with the amount of "gun control" they already had, a major newspaper did a poll just a few days after the massacre (while all minds were "clear") and, not surprisingly, found high levels of support for extreme gun control measures.  This poll would be used thereafter by the commonwealth government and other gun controllers to claim that Australians supported the new gun laws to come.

Within ten days after the massacre, the research staff of the commonwealth's parliamentary library finished a report about the "gun problem" for the enlightenment and guidance of the gun-ignorant representatives (ministers) to parliament.  Or, was the report just waiting for addition of a release date when the massacre occurred?  This report9 had some valid and useful information about history and options for forcing uniform gun control.  It also had much misinformation and blattantly biased statements about "military-style" firearms, their lethality and evidence suggesting that increased firearms possession might cause increased violence—specifically, a study by Martin Killias for the U.N. Interregional Crime (and Justice) Research Institute (UNICRI).

About a week after the massacre (May 5) the police ministers had a meeting in which they agreed that firearm ownership was not a right of Australians, and that self defense was not a valid reason to possess firearms.  At another emergency "special" meeting on 10 May 1996 they approved a comprehensive gun program.   The program was virtually ready and waiting at the time of the meeting although it may have been adjusted some.  Additional resolutions were added in meetings on July 17 and November 15.  The state parliaments would have to pass some laws in order to fully implement all the agreed changes, but some changes were made immediately.  For example, the Tasmanian police minister thought that the existing Tasmanian Guns Act 1991 gave him authority to require registration of certain semiautomatic centerfire firearms, so he issued an order to do so on May 7, 1996 (even before the 10 May APMC meeting).

The new laws would virtually ban private possession of pump-action shotguns and all semiautomatic firearms (even of .22 rimfire caliber) and provide for the government to pay people over the course of a year (up through September, 1997) to turn in such firearms for destruction.  The money for this was to come from a 1-year increase in the "medicare" tax (after all, the public was told guns were a "public health" issue).  The laws would provide for a person to possess a non-military semiautomatic long gun, or pump action shotgun, for professional use only if the authorities could be convinced that the person had a "genuine need."  A "genuine reason" (to the authorities' mind) was required for a person to have a lesser firearm.  And protection of life was not to be a valid reason for having a firearm.

The official Australian buyback web site history page (now defunct) said that the agreement was the result of "detailed preparation and extensive consultation (i.e., conspiracy) over many years."  A representative of the Australian Press Council was a participant in a February, 1997 symposium of gun control advocates, convened by the Dept. of Public Health and Family Services, to plan "public health approaches to firearm violence."  Does anyone doubt that the Australian press was already prepared to keep the public convinced of the need for gun control when the Port Arthur massacre occurred ten months before?

There was objection by some people (basically gun owners) who were concerned about losing their right to have firearms, but resistance was divided.  Many gun owners and their representatives made statements accepting some of the restrictions and the basic concept that their governments could or should restrict based upon someone's idea of what need people might have for one kind of firearm or another.  A huge demonstration (150,000) occurred in Melbourne.  At least one other large (8000+) demonstration occurred in Brisbane, Queensland on June 29, 1996, but the media basically ignored these other demonstrations.

Most politicians caved in quickly to the media onslaught.  The AIC and the media worked on maintaining high support for the new gun laws.  On November 4, 1996 the AIC released "Violent Deaths and Firearms in Australia:  Data & Trends."  About 70 percent of this 96-page report (RPP04) is about firearm deaths.  The rest is about Australian homicide over the years.  On the same day the AIC also issued a media release entitled "Lax firearm laws mean more deaths."

The actual report is a generally professional statistical analysis with some scattered bias, like constantly referring to "firearm caused" deaths.  But the media release was blatantly biased and drew unwarranted conclusions about firearms prevalence and death rates.

The media release opened by declaring that the AIC report "shows that Australian states in which guns have been more easily available have significantly higher death rates than the national average"—and listing 1994 firearm (not total) death rates for the states and territories.  The media release title and opening statement tried to make the public think something about deaths in general although the material was only about those deaths in which firearms were the instruments used to cause such deaths.

Then, among findings mentioned, the media release stated that the firearms death rate had fallen over the past decade "as legislation of the late 1980s in some states has made firearms somewhat more difficult to obtain," that "those states in which firearms regulation and licensing have been less stringent have had significantly higher [gun} death rates than the national average" and that the usual small group of countries have less gun homicides than the USA because the USA has more guns.

All these statements were attempts to make the public think that the legislation in some states and countries had reduced gun death rates and that states/countries with less stringent control had higher death rates as a result, ignoring the possibility that the higher and lower gun death rates might be caused by something other than gun laws or gun prevalence.  The report actually had no facts upon which to base such conclusions.  Note that the AIC switched to addressing homicide rather than total deaths when discussing international comparisons since some of the countries have higher suicide rates than the USA.

The media release used the usual gun controller tricks of addressing "firearm" deaths, and also including accidents and suicides.  They focus on "firearm" deaths/injuries because they can always show reductions in these if firearm availability is reduced, even though total deaths and injuries are not affected and such availability reductions might actually cause more deaths/injuries than they prevent.  They include accidents and suicides in their discussions because Australia has so few firearm murders and other homicides that the numbers and rates would not tend to upset people much, especially in comparison to other types that are much more numerous.  Also, because the numbers are so small, they vary a great deal randomly from one year to the next so that a plot will not clearly show a definite trend over time.

Gun controllers are able to upset the public much more by including suicides in their death figures.  What the gun controllers don't want people to know is that it has been proven (and confirmed) that reducing the prevalence of firearms does not reduce (total) suicide rates even though it does in fact reduce firearm suicide rates.  People wanting to commit suicide simple use another tool/method when a firearm is not available.  [Doctors typically don't believe this because they know that a firearm is more certain to kill than some other methods that are frequently used, and they've told each other that some large portion of suicides are impulsive, spur of the moment events.  What they fail to account for is that there is no shortage of equally deadly methods, and a person who fails in a suicide attempt is very likely just to try again and again (often without others even knowing) until he or she succeeds.]

The truth is that the claim about availability was unfounded from a scientific research standpoint.  The report had not even addressed "availability" in the various states, even by reference to some existing analysis, much less quantified it.  (They couldn't reference an analysis because there had not been one.)

What the data in the report actually showed (in combination with what has been able to learn about gun laws in the late '80s and early '90s) was that:

  1. The "gun death" rate for NSW rose progressively for three years after effectivity ('90) of the NSW Firearms Act 1989, before finally falling for just the last two years of the period (meaning the gun death rate reduction was not caused by the law);
  2. The final drop in the NSW gun death rate did occur upon effectivity of the major '91 amendments to the law, but this was only for two years (not enough to qualify as a statistically significant trend) and the pattern actually resulted from a rather random combination of ups and downs of gun homicide and gun accidents added to a minor drop in gun suicides.
  3. The gun death rate for Tasmania showed no statistically significant change in the two years of effectivity ('93 & '94) of the Tasmania Guns Act 1991, and the small reduction was entirely the result of the rate being artificially high in '92 because of an abnormally high suicide rate for that year (i.e., there was no reason to think that the law had any impact on gun death rate); and,
  4. Only Queensland had a sustained decline that started with the effectivity of its new gun law, the Firearms Act 1990, and extending through '94.

Click here for details on what the report actually showed about "gun" deaths from 1983 through 1994 in the different states.

SUMMARY ON THE AIC MEDIA RELEASE:  The international comparisons are completely bogus because they selectively address only a few countries, because they ignore the many significant differences between the countries, and because they ignore the fact that some of the countries had still lower death rates in the past when they had more firearms.  Firearm injury death rates in the different states, and percentages of deaths that involve firearms, are irrelevant because there is no reason to think that death rates (overall) go down just because firearm death rates go down.  It is entirely possible that anything that causes firearm death rates to drop might also cause overall death rates to rise.  Focusing on firearm deaths is just a gun controller way to sway the uninformed by impressing them with numbers and taking advantage of their inability to understand statistical information.

Even if firearm death rates were accepted as a valid measure of violence, the firearm death rate history for the decade preceding the report does not consistently support the claims made in the press release (for the three states for which has been able to find information about effectivity of laws and amendments that were claimed to have reduced firearm deaths).


Thirteen years after the ban/buyback there is now plenty of data from which to evaluate the effectiveness of the measures put in place.

A lot of the peoples' money was spent buying firearms.  A lot of Australians became alienated from their governments and institutions.  Irrational fears of people unfamiliar with guns may have been eased.  Gun accident deaths inexplicably rose.  Robbery rates popped up for several years (until '04).  There was no perceptible, definite impact—good or bad—on burglary, serious assault, homicide, gun homicide, suicide or gun suicide.

The gun controllers continued to do something similar regarding handguns, which rose in prominence since long guns had been reined in.  If handgun crime becomes less significant, some kind of handgun or long-gun crime will gain in significance and the gun controllers will target that category, and so on as long as there is a firearm.

See the detailed analysis (2009) of results of the ban/buyback and/or concurrently implemented policies.


By '87, a lot of government and community leaders and activists were already convinced—based on ignorance, assumption and a bit of junk science— that guns were a problem they needed to fix.  They were able to add a lot of like-brained people to their ranks.  They eventually included much of the medical community, educators, social workers, churches, woman advocates, child advocates, the press (to sell papers), Duncan Chappell, the AIC, police leaders, John Howard, television media (to deflect attention from their violent programming), and the Commonwealth government.

At some point before the Port Arthur incident, at least some of the leaders among them became part of a conspiracy, planning for gun control.  After Port Arthur, Australians were conditioned from the beginning of the campaign to reject anything from the USA unless it came from gun controllers.  This was done by astutely denegrating the USA and it's "frontier gun culture" with a public relations theme of "We don't really want to be like America in this."  This was continued with a campaign attacking the "evil NRA" abouts its criticism after the buyback.

In the USA, gun controllers work at keeping people from knowing about criminologists.  They try hard to ignore criminologists and keep the media full of reports on the junk studies done by the medical gun controllers.  In Australia the AIC years ago assumed the mantle of "criminologists."  The role of the AIC in connection with gun control in Australia has been to use statistical methods, the government's prestige, and the peoples' money to produce scientific-looking propaganda. From the quality of some documents at the AIC web site, and the fact that they reference the junk science by such as Killias and Kellermann, one would think that the staff of AIC are medical doctors who "know best what is best for the common people," or just statisticians with prejudice against firearms.

AIC staff and others at the 1997 symposium of gun control advocates to plan "public health approaches to firearm violence" thought that the Kellermann "43 times" junk study should be replicated in Australia, and that something needed to be done to get accurate information about the numbers of firearms Australians have.  They're too ignorant of the body of criminological knowledge about firearms to realize that the Kellermann study is worthless except as propaganda, and that the gun laws they supported have made it impossible for anyone ever to know how many guns Australians have (without sending authorities to exhaustively search all properties) because the laws cause people to falsely deny possession of firearms except registered firearms.

The AIC web site has a lot of useful information.  But use it with caution.  Be alert for anti-gun bias.  Much useful information is also available at the research archive of the Sporting Shooter's Assn of Australia.

The Killias Analysis

The Killias analysis examined the correlation between the national rates of homicide and suicide, and private firearms possession, for several countries.  Killias found that there was indeed a correlation.

Unfortunately, there is no way he could have known accurately how many households had guns in even one of the countries.  And he seemed unaware of basic principles of statistical analysis.  For one thing, the number of data points (14) was too small for any significant results.  And the correlation was entirely the result of including in the analysis one country —the U.S.A.—that had an extra-high level of violence and an extra-high level of (estimated) firearm possession.  When the analysis is done without the U.S.A. data in the mix, the correlation disappears.13  (Removing the N. Ireland data point as well would bring back some correlation.)

In effect, Killias' result was simply a scientific-appearing restatement of the simplistic idea that guns must cause violence since the U.S.A. has both extra-high violence and extra-high firearm possession rates.  The U.S.A. data point "overwhelmed" all the others.

Besides the fact that the analysis was not valid, a correlation would not have proved that firearm possession caused the violence.  Correlation only means two things occur together.  It says nothing about which of the two is the cause and which is the effect, or if maybe each has an effect upon the other—or if both are just results of some other thing(s) in common.  The study results could just as easily have been interpreted that when violence is high, or the risk of homicide is high, in a society, a higher portion of that society obtain firearms to protect themselves, their families, and their neighbors.

The three parliamentary library authors of the brief provided to parliament should not be blamed for the fact that the Killias "study" was junk science.  After all, they were doing a report on things they basically knew nothing about—and they might not even be able to understand an explanation of the deficiencies of the study.  But people whose jobs are to do research on available information for guidance of ignorant politicians should be expected to know that any one study cannot be considered proof of anything, especially before qualified people have critiqued the study.  The authors should have searched out and summarized all available data on the topic, which has life and death consequences.  If little evidence is available, that should be stated.  If such meager evidence is summarized, a warning of the inconclusive nature of the evidence should be included.


In all the following "NCV" signifies National Committee on Violence, and "AIC" signifies Australian Institute of Criminology located in Canberra, Australia.


  1. NCV, Violence:  Directions for Australia (Part 1) (AIC, 1990)
  2. David Indermaur, Violent Crime in Australia:  Interpreting the Trends, (AIC Oct, 1996)
      (Trends and Issues No. 61)
  3. Australia Parliamentary Library Historical Supplement
      to the Monthly Economic and Social Indicators (MESI)
  4. NCV Violence:  Directions for Australia (App B) (AIC, 1990)
  5. NCV Violence:  Directions for Australia (App C) (AIC, 1990)
  6. NCV Violence:  Directions for Australia (App D) (AIC, 1990)
  7. Jenniver Gardiner (minister) in 12 Nov 91 speech to NSW parliament
  8. Jenny Mouzos, Firearm Related Violence:
     The Impact of the Nationwide Agreement on Firearms,

      (Australia, AIC, May, 1999)
      p/o "Trends & Issues" series (T&I 116) (PDF file)
  9. Jennifer Norberry, Derek Woolner & Kirsty Magarey,
      After Port Arthur - Issues of Gun Control in Australia,
      Current Issues Briefs, 16 1995-96 (Canberra, ACT, Australia:
      Australian Parliamentary Library, May 7, 96)
10. Compton's Interactive Encyclopedia 2.01VWF ('94)
11. Peter Saunders, "Poverty and Deprivation in Australia," Year
      Book Australia
reproduced as p/o "Australia Now" series at ABS
      web site (Canberra, Australia: Australian Bur. of Stats, 1996)
12. Australian social trends 1997, ABS (cat. No. 4102.0)
13. Gary Kleck, Targeting Guns—Firearms and Their Control
      (New York: Aldine de Gruyter, 1997) p.253-254.