I went searching for some scientifically rigorous data on this question today, after being bombarded by conservative relations, friends, acquaintances and SLOG readers today regarding a gentleman who defended his garage (undoubtedly filled with mountains of useless junk that should be thrown out) from two burglars, by killing one and seriously injuring another human being with his vaunted handgun.
Citing this gun owners' actions—right now costing state and federal taxpayers tens if not hundreds of thousands in medical bills for his victim, but saving that junk in his garage—as proof that guns make you, me and everyone safer is a bit like saying, "I totally know this guy who smoked until he was ninety, and then died of a broken hip. Smoking is totes safe, man."
I'd feel better with some data—pointing us in one direction or another. Here's what I found.
4. Guns in the home for personal protection were frequently used to threaten or intimidate other family members. This most often was done by a male gun-owning family member, to a female member of the household. This family-on-family threatening of gun violence occurred roughly ten times more often than the gun being used to threaten or intimidate an intruder.
Statistics are reviewed which show that a gun in the home is far more likely to lead to the death or injury of a family member or friend than to the death of an intruder. Data on victimizations and the use of firearms for self defense are then examined for the crimes of burglary, robbery, assault, and rape. In each case the effectiveness of guns in preventing or deterring the crime is analyzed, and compared to the effectiveness of other self defense methods. The data presented in this report indicate that private handgun ownership provides no significant deterrent to burglary and violent crime. It may, in fact, escalate the severity of the violence if offenders believe they must be more heavily armed than the citizenry. The statistics also showed that the use of a weapon in resistance to a criminal attack usually results in greater probability of bodily injury or death to the victim. Other methods of resistance, such as flight or verbal resistance, were found to be more effective in aborting the crime while having less probability of causing harm or death to the victim. In circumstances where the offender is armed, non-resistance most likely resulted in the minimum amount of harm to the victim. The authors conclude that because of the surprise nature of most violent crime and the fact that it is likely to occur between strangers, it is improbable that the victim would have time to use the handgun in any event. They argue that in light of the risks of handgun ownership - the possibility of escalating the violence of the crime, and the risk of accidents and suicides among family members - other safer methods of crime prevention must be adopted.
Among pro-gun advocates, the idea carrying a gun makes one safer is pretty much sacrosanct—the cornerstone of the Second Amendment religion that dominates our country. Per the available data, it's totally and completely wrong; carrying a gun, objectively, make one less safe.
Updated: One last point: I'm a civil libertarian, and believe all of the rights extended in the Bill of Rights (and more) should be defended—including gun ownership. To state something that should be obvious to everyone, but apparently isn't: Pointing out the statistically demonstrated risks of exercising those rights isn't the same as stating they aren't or shouldn't be rights. Be noted.