It is accepted as fact by most gun control advocates that simply having a gun in your home nearly triples your chances of being murdered.  The reality is that this number (from Dr. Kellerman’s work of 3-4 decades ago) is nothing more than deceptive propaganda.  Here’s the evidence that supports this claim:

1) Kellerman Refused To Release His Underlying Until For Four Years After He Published His “Study”

In the world of research, this is very suggestive of deception.  At the very least, it suggests that the researcher does not think that his work can stand up to real peer review.  As we will see, this is clearly the case here, as Kellerman’s claims about the risk of gun ownership do not hold up to examination.

It also enabled the use of this “data” for years before it could be examined critically – which is why it is so widely accepted today.

2) Association Does Not Equal Causation – Therefore This Study Does Not Prove That Owning A Gun Increases Your Risk Of Being Murdered.

This is a mistake that, to be honest, people on both sides of the gun debate frequently make.  However, Kellerman (the M.D. who did the “study”) is a professional researcher who definitely knows better than to make this mistake.  He knows very well that association proves absolutely nothing.  AT MOST, association is an indication for further study.  Please do not take my word for this – allow me to prove it.

From Kellerman’s own study, here are some of the “risk factors” associated with increase homicide risk, along with the amount of increased risk:

***************************************************
Variable                     Adjusted Odds Ratio
—————————————————
Illicit drug use                5.7
Home rented                     4.4
Any household member hit or
hurt in a fight in the home   4.4
Case subject lived alone        3.7
Gun or guns kept in the home    2.7
Any household member arrested   2.5
***************************************************

Let’s look at each of these factors in order of “risk”.

Anyone who has watched “The First 48” understands that illicit drug use, and more significantly sales, increases your chances of being murdered.  In fact, the associated risk is more than double that of owning a gun.  How many of the drug dealers included in that top number also own guns?  My guess is close to 100%.  Yet, Kellerman makes no effort to determine if this likely association is distorting his “gun ownership risk factor” number.

The next highest risk factor is renting your home verses owning your home.  Here we see clearly that association does not equal causation.  Is it plausible that renting a home increases my chances of being murdered compared to owning the home next door?  Could we reduce the murder rate by giving everyone a home?  Of course not.  Such a conclusion is both illogical and absurd.  What may be a very real factor is what neighborhood you live in.  In poorer, high crime areas, your chances of being murdered and of not owning your home are both higher.  This is an excellent example of association without causation.

The next risk factor, “Any household member hit or hurt in a fight in the home.”, makes logical sense.  Clearly, previous domestic violence likely is a risk factor.  However, take note that this is a much higher number than gun ownership – indicating that many of these murders took place in homes without firearms.  More about this later.

Next, we see that living alone increases your risk of being murdered by a lot more than owning a gun (3.7 times verses 2.7 times) – yet here again, is this association or causation?  Are the same people who want to discourage gun ownership telling people to get roommates or get married?  Of course not – because they don’t believe that this statistic is a meaningful indicator of risk.  Why not?  Well, my opinion is that it doesn’t fit their preconceived beliefs.

Now we come to the intent of the “study” – proving that gun ownership increases your risk of being murdered.  But does it really prove that?  Let’s look at what this number doesn’t tell us.

First, it doesn’t tell us if the gun was purchased because the person was at higher risk of being murdered.  It also doesn’t tell us if the person was murdered with that firearm (more about that later).  It doesn’t tell us if the person owning the gun is a criminal, such as a drug dealer.  It doesn’t tell us if there has been domestic violence in that home.  It is certainly possible, IMHO even likely, that criminals and domestic abusers are more likely to own a gun for evil purposes.

Another entire area of missing data is the legality of the firearm.  I’m certainly not endorsing every gun law in the US, or even any of them (save perhaps felon in possession), but I do think it very likely that illegally owned firearms are much more likely to be used to murder someone.  After all people buying firearms legally at gun stores are required to undergo a federal background check.  In many states, there are additional requirements.  I assume Kellerman supports these efforts – yet his “2.7 times” statistic makes no effort to determine if they are working.  I think the reason that gun control advocates don’t want to know if the gun was owned legally is that they know that the increased risk posed by legally owned guns – if any – is far lower than “nearly 3 times greater”.

Finally, note that the “increased risk” of someone in the home being arrested is nearly identical to that associated with firearms ownership.  It would be very interesting to know what the overlap between these two groups is – but again, we do not know.

3) How Did They Die?

The short answer is: We don’t know.  That’s right, Kellerman made no effort to determine if the firearm owned had anything whatsoever to do with the murder.

Here is the long answer….

Whose gun?  Anti-gun rights activists would like you to believe that the association is a result of people being murdered with their own firearms, yet Kellerman’s own data suggests that for all gun homicides of matched cases no more than 34% (probably much less) were murdered by a gun from the victim’s home (source).  This actually makes perfect sense when one considers that many (likely most) homes have firearms carefully stored away.  These are counted as “gun owning homes” in his “study”.  So, when a household with a shotgun in the attic experiences a homicide, Kellerman counts it as being associated.

Was a gun used at all?  Unsurprisingly, Kellerman made no attempt to answer this question either.  As a result, we know that his numbers include 100% of murders committed with knives (the 2nd most common murder weapon), clubs and other blunt instraments, and bare hands.  All of these methods of murder are used more commonly than rifles and shotguns – which compose about 50% of the firearms owned by Americans.  Clearly, a lot of the people included in that “2.7 times” stat were killed with something other than a firearm.  How many?  We simply do not know – but it is a lot.

What we do know is counting murders that don’t involve the gun owned by someone in the household inflates Kellerman’s numbers.  Excuse me for believing that this was intentional.

Would the murder have happened if no gun was available?  Again, we do not know.  What we do know is that murders have been happening as long as humans have been around – and for nearly all of that time there were no firearms.  Even today, a huge number of murders are committed with weapons of opportunity – knives, blunt interments, bare hands.

In cases of violence against women resulting in murder, if the man wants to kill the woman, he likely doesn’t need a gun (although he may indeed use one).  A knife, club or his bare hands will do just fine.  He is stronger and larger in most cases.  (On the other hand, a woman with a firearm can defend herself – even against an armed attacker.  The news is filled with such events.)  It is therefore likely, even in cases where a gun was used, that many of the murders would have happened anyway.  Only the method would have changed.  The exact number is unknowable, but this clearly inflates Kellerman’s “2.7 times” statistic.

Conclusion:

Everyone has bias.  We all have opinions.  That said, I think it is very clear that Kellerman allowed his opposition to gun ownership to overcome his knowledge of what good research is and is not.  He could have answered some of the preceding questions and come up with a solid number for what the risk of gun ownership (if any) really is.  Instead, he produced a so called study that is “junk science” at best and pure propaganda at worst.

What to dive deeper into this issue?  Read THIS