Over the last 60 years, the number of guns in private hands in the United States has risen sharply. At the same time, during the last two decades, the number of murders and non-negligent homicides have fallen sharply. More guns do not equal more crime. In response, opponents of an armed civilian population have shifted their focus and changed the terminology they use. Instead of gun control, now they label restrictive legislation as promoting “gun safety”.
Instead of focusing on violent crime, they’ve re-framed the issue, using the term “gun violence.” Most people consider gun violence to be “gun crime.” But anti-gunners use the term to include suicides and accidents as well as murders. They lump them all together as “gun deaths.”
Their theory is that more guns equal more deaths. Fatal gun accidents have fallen to all time lows; murder has fallen sharply to rates not seen since the middle 1960s. Most “gun deaths” are suicides. Their claim has become “more guns equal more suicides”.
The claim ignores the fact that other cultures with few guns have higher suicide rates than the United States. Substitute methods of suicide are readily available.
If more guns equals more suicides, you would expect the fraction of suicides committed with guns to increase as the per capita number of guns increased.
An increased fraction of suicides with guns does not necessarily mean that more guns result in more suicides. Suicide levels probably rise or fall independently. But if more guns does not correlate with a greater fraction of guns used in suicides, the theory that more guns result in more suicides becomes implausible.
Looking at the fraction of suicides with guns eliminates the various economic ups and downs and other factors that contribute to the total number of suicides.
Comparing the fraction of gun suicides to the per capita number of guns in the country eliminates many of the problems with surveys and estimates of gun numbers in various states, cities and regions. State and local data about gun ownership is unreliable and incomplete.
The ATF has good data about increases in private gun numbers.
The Center for Disease Control has records for suicides from 1981 to 2014. The fraction of suicides with firearms has followed the general trend of homicides. The fraction of suicides with firearms increased slightly in the 1990s, then decreased over the past two decades.
Guns were chosen for .58 of all suicides in 1981. The fraction of suicides committed with guns rose to high of .61 from 1990 to 1997 (the high point of homicides was 1993-95), then started a long decline ending at .49, the record low, in 2014.
Per capita numbers of private firearms increased from .754 in 1981 to 1.176 in 2014. While the fraction of suicides committed with guns fell 16 percent, the per capita number of guns in the country rose 56 percent.
More guns have not increased the fraction of suicides committed with guns over several economic cycles.
The rates of suicide used are those from the Center for Disease Control WISQARS data base. The CDC adjusted the rates to account for changing age distributions. The age correction does not make a large difference. It removes the variation associated with changing percentages of various age groups over the decades examined.
The number of private firearms was calculated from the Bureau of Alcohol, Tobacco, Firearms, and Explosives (ATF) numbers, using the method developed by Newton and Zimring, expanded on by Gary Kleck in “Point Blank: Guns and Violence in America”. Census figures were used to determine the per capita numbers.
©2016 by Dean Weingarten: Permission to share is granted when this notice is included.
Gun Watch
Throwing logic at this is like tossing bread crumbs to a cockroach. You can’t feed them to death.
In the cases of instant rice to pigeons, instant oatmeal to ants, or generally any large quantity of food to a pug, I disagree!
Data is important in our fight to restore Second Amendment rights. There are only a very small number of dedicated disarmists that are immune to logic. Dedicated Second Amendment supporters outnumber them by about 5 to 1. Together, they likely make up less than 10 percent of the population.
The other 90% can be swayed by facts, if they are presented with them. They have generally not been presented with facts from about 1965 to 1994. 1994 was the peak of the disarmist push. Facts are perculating out though sites such as TTAG, and the alternate media. The five to one ratio of Second Amendment activists to disarmists is telling in the comments sections of articles on the net. We are gradually getting the information out to the uninformed, and it is changing attitudes in our favor.
More drama in Dallas.
Sooooo…even Bury Soetoro admits crime is falling- while bemoaning there’s too many guns. And countries like Korea and Japan have much higher rates of suicide without those evil gunz. Of course y’all never hear this study from the dumbocrats…
To be fair, it’s probably more relevant to compare the fraction of suicides committed with guns to the fraction of households with guns, rather than to the number of guns per capita.
We simply do not have the data on households with guns. There are surveys, which seem to rise and fall with the level of vitriol expended on guns in the media. we have much better data on the overall private stock increases than we have on household gun numbers.
There are many possibilities. Women tend to use guns less for suicides than men, and the number of women gun owners is has been rising.
Households with guns probably has either not changed or risen in the past few decades, . Using direct question polling and surveys is likely a massive and increasing undercount. All the peer reviewed work, all, that looks at truthfulness responsive in surveys and polling on privacy or confidentially related direct questions shows large undercounts with direct questions. Ask people if they are gay and you get large undercounts especially in areas of the US that are less tolerant, Ask people if they have a gun at home and you get the same thing. Indeed we know from anonymized health facility information on many women have had an abortion, that only about 60% of those who have will tell a telephone survey and under 50% of those who have will tell a face to face survey. Even women who will tell will underreport the number they have had.
Why do you think GSS (the survey cited by the gun control advocates) gets on average 11% less households with guns than Gallup? GSS interview is face-to-face, Gallup is by phone. We know in the case of direct question on abortions both get undercounts, but face to face get the most severe undercounts. Same thing with gun ownership.
On firearms in the home, we know that the proportion of people who have a firearm for self/home defense has risen profoundly in relation to the number who have a firearm for hunting. For self/home defense, all modern training includes the instruction to never tell a stranger you have a firearm at home. so not only is there a large undercount of homes with guns it is likely that undercount has been increasing as the reason for gun ownership has been shifting – and privacy concerns have therefore increased.
The peer reviewed science on surveys shows the much more accurate methods on high confidentiality issues are to use indirect or “veiled” questions related to the subject for both baselines and trends.
What do the indirect questions pointing to gun ownership rates look like?
1) Gallup and Pew 25 year question record on “does owning a gun make a home safer?” is a the central one, that has risen from 32% and 35% saying it does in 1990 in gallup and pew, to 58% and 63% saying it makes a household safer in 2015. ‘
2) Gallup, Pew, ABC, NBC, NYTimes, CNN, WSJ polling on same question support for more substantive gun control over 35, 30, 25, 20, 15, 10 and five year trends shows a decrease in support.
3) Approvals of NRA are around 58% which is up on the 20 year trend and up despite massive amounts of money spent by the gun control lobby directly attacking the NRA.
Also The Opiod Epidemic Has Made Pills An Easy Alternative To Firearms Over The Past Few years. If u took the pill increase out u may find firearm suicide rates are fairly static, but that’s just a hypothesis. For suicide by firearm to go down during the last 15 years of economic downturn is interesting. I guess people pawned there guns…
Us people of the gun already know this. However, I think a link of this most certainly belongs in the “facts about guns” section.
This graph is very strong evidence that “more guns does not equal more crime,” which is often touted by Democratic Party gun haters.
Dean I pointed out above we simply do not know the proportion of homes with guns, but that contrary to the claims of the gun control lobby, it likely has risen in the past 25 years, not declined as they will claim.
We should also look at the science in Australia. Suicide is a core factor in the US gun control debate, since US gun murder and US gun fatal accident rates have been on a long and profound decline. Suicide is now about 2/3 of gun death and this is why the gun control lobby uses the terms “gun deaths” and “gun violence death” so they can pack in suicide and avoid discussion the sustained long term decline in gun murder rate.
The fact is after a sharp and immediate near halving of households with firearms, and access to firearms through mass confiscation, it initially appeared that Australia had not only reduced gun suicide, but reduced overall suicide. Essentially the raw data showed for every 100 gun suicides reduced, other methods increased by only 50 (substitution effect) and it could be claimed gun confiscation regime of Australia was responsible for overall reduction in suicide of the other 50 reduced suicides.
It is that initial conclusion that is touted by US and Australian gun control advocates.
But we know now, from every Austrian academic study that has looked *how* suicides are tabulated that in fact there was no reduction is suicide whatsoever. Half of the reduced gun suicide moves to other means correctly tabulated as suicide, and half the reduced gun suicide moved to suicide by other means incorrectly recorded (“suicide misclassification as accident”).
In short not only were half of decreased gun suicide immediately replaced by other means known suicide – Australia also saw a sharp and immediate increase in self caused fatal “accident” death associated with suicide. Unless one posits another reason for an immediate increase in Australians driving alone into abutments and trees, taking “accidental” fatal overdose, falling onto train tracks, falling out of windows, in 1997s and after, it is clear there was no drop in suicide, just a increase in suicide by replacement methods and an increase suicide means that were incorrectly tabulated.
Guns access simply increases gun suicide, and all the examinations that control for demographics and underlying misclassification as accident of other means, find NO increase in overall suicide among individuals, households, jurisdictions, states or regions or countries due to guns.
Coroners and medical examiners have been proven to presume suicide when prevented with self inflicted fatal gun shot or self hanging, and do the opposite — presume accident — with virtually all other self caused death that are commonly also suicide.
Here are references and quotes on ten years after the mass confiscation that are ignored by the gun control lobby, the US media and academic researchers who have a political bias:
“These studies have shown that ABS has seriously been under reporting the number of suicides. The Queensland study reported that this under reporting had increased during the period under study and that the under count negates any apparent decrease in suicide deaths shown in ABS statistics.”http://www.aph.GOV.AU/~/media/wopapub/senate/committee/clac_ctte/completed_inquiries/2008_10/suicide/submissions/sub42_pdf.ashx
“These findings confirm that routine mortality data have underestimated suicide mortality in Australia in recent years probably due to incomplete coroner data being available to ABS coders. Certain types of unintentional injury deaths have been over-estimated [meaning recorded as ‘unintentional’ when they were suicide]. Incomplete linkage of NCIS and ABS data in the source data used for this project complicates calculations of adjusted estimates and trends.” http://www.researchgate.net/publication/24034093_Suicide_decline_in_Australia_where_did_the_cases_go
“AUSTRALIA has dangerously miscalculated its suicide statistics – by as much as 30 per cent in NSW and Queensland – leaving a silent and growing epidemic of mounting deaths.
The figures are in stark contrast to years of backslapping by state and federal governments, congratulating themselves for reducing suicide rates from a peak of 2700 in 1997.
The Herald can reveal the suicide toll is as high now as it was in the 1990s – if not higher”
http://www.smh.COM.AU/lifestyle/diet-and-fitness/revealed-australias-suicide-epidemic-20090820-es3p.html
Thank you for the detailed post. A good friend and I were speculating about exactly this effect as I was writing the article. I appreciate the links. I have asked for a full text copy of the last paper. We will see if the authors are willing to send it to me. It is an excellent point that suicides with guns are clearly identified, but others, such as with cars, are not. I wonder if anyone has done a comparable study in the U.S.
Dean, Sadly papers making direct assertion on the relative over-count of gun (and self hanging) suicide would be politically incorrect in most of academia and hence a dearth.
But what you want to do is go to google scholar and start looking for the term “suicide misclassification.”
You will start to see the underlying data and causes: coroner, social, family, and even central health authorities reticence to identify them as such, most affecting any self caused death causes that are ambiguous
Japan and South Korea have some of the highest acknowledged suicide rates in the world. In Japan, suicide is considered honorable, a social norm. This implies that reported suicide rates all over the world are much more a function of reporting than they are of actual rates. Is that noted in the literature on misclassification? We see the same problems with crime reporting.
By the graph more guns does equal more suicides (in a universe where correlation equals causation), it just means more suicides that don’t involve guns. Seems to suggest that if we flooded the streets with guns that suicides which use guns as the tool would drop to statistically insignificant numbers but somehow would make suicide by any other means much more attractive to the suicidal. Hmmm… Food for thought.