* Guns are not used millions of times each year in self-defense
* Most purported self-defense gun uses are gun uses in escalating arguments, and are both socially undesirable and illegal
* Firearms are used far more often to intimidate than in self-defense
* Criminals who are shot are typically the victims of crime
https://www.hsph.harvard.edu/hicrc/firearms-research/gun-threats-and-self-defense-gun-use-2/
Are there "1 to 3 million" defensive gun uses per year in the US?
The "1 to 3 million" figure comes from one "random-digit-dialed telephone survey of 4,977 adults conducted from February through April of 1993" and reported by Kleck & Gertz in 1995. Kleck says the random sampling error of his survey is less than 1%, but if so, that would produce far smaller error bars than 1 to 3 million.
Others have tried to replicate Kleck's data; here's a survey of polling
https://www.hsph.harvard.edu/hicrc/firearms-research/gun-threats-and-self-defense-gun-use-2/
Needless to say, they produce much smaller numbers and likely have their own biases. But here's a useful quote:
We analyzed data from two national random-digit-dial surveys conducted under the auspices of the Harvard Injury Control Research Center. Criminal court judges who read the self-reported accounts of the purported self-defense gun use rated a majority as being illegal, even assuming that the respondent had a permit to own and to carry a gun, and that the respondent had described the event honestly from his own perspective.
What they're saying, in not so many words, is that guns are often used to intimidate. The guy who draws the gun may think he's preventing a crime. But putting another person in reasonable fear of bodily harm is the crime of assault (battery is when you make contact with the other person). So they're running these stories by judges, who say that many reported "defensive" gun uses are actually the offensive crime of assault, unknown to the gun user.
They also say "8. Criminals who are shot are typically the victims of crime; 9-10. Few criminals are shot by decent law-abiding citizens." i.e. they say it's almost impossible to find a criminal with a gun wound who was shot by a law abiding citizen.
And finally,
Victims use guns in less than 1% of contact crimes, and women never use guns to protect themselves against sexual assault (in more than 300 cases). Victims using a gun were no less likely to be injured after taking protective action than victims using other forms of protective action. Compared to other protective actions, the National Crime Victimization Surveys provide little evidence that self-defense gun use is uniquely beneficial in reducing the likelihood of injury or property loss.
I'm not saying this Harvard data is perfect, but it asks a lot of specific questions about defensive gun use; and the results do tend to completely undermine the Kleck 1993 polling result.
Here's the peer reviewed report from the two Harvard surveys:
https://injuryprevention.bmj.com/content/6/4/263Same author, an analysis of the National Crime Victimization Survey from 2007-2011 covering 14,000 incidents
https://www.sciencedirect.com/science/article/abs/pii/S0091743515001188Results
Of over 14,000 incidents in which the victim was present, 127 (0.9%) involved a SDGU. SDGU was more common among males, in rural areas, away from home, against male offenders and against offenders with a gun. After any protective action, 4.2% of victims were injured; after SDGU, 4.1% of victims were injured. In property crimes, 55.9% of victims who took protective action lost property, 38.5 of SDGU victims lost property, and 34.9% of victims who used a weapon other than a gun lost property.
Conclusions
Compared to other protective actions, the National Crime Victimization Surveys provide little evidence that SDGU is uniquely beneficial in reducing the likelihood of injury or property loss.
SDGU = self-defense gun use
