One of the main arguments against gun control is that people use guns defensively to fight crime, so guns offer a net benefit to society. Reducing or banning guns will actually increase crime. Let’s investigate the numbers surrounding this argument.
Violent crime rates: The violent crime rate in 2011 was 386.3 (all crime rates in this article are per 100,000 inhabitants) . Despite the low probability, it adds up over time. In 70 years, this translates to a 24% chance of being a victim of violent crime ! As a point of comparison, consider vehicular fatalities, which are a good measure of our baseline tolerance for risk. Deaths in car crashes occur at a rate of 10.6 , over 100 times lower, resulting in a chance of death of less than 1% over a 70 year span.
Regarding guns specifically, according to some slightly older numbers (1993-2001), about 11% of violent crimes involved guns . Using the average crime rate of 386.3, this suggests that the rate of violent crime involving handguns is about 42.
Defensive gun use: There is vast disagreement on rates of defensive gun use. Estimates range from 100,000 to 2.5 million per year . With 313 million inhabitants in the US, this results in rates somewhere between 31.9 and almost 800! The high value may be inflated due to a number of reasons, including the following:
- Poor survey design. One of many problems is “telescoping,” i.e., counting older events as having occurred recently.
- Perceptual biases. Just because you think your gun prevented a crime, doesn’t mean a serious crime was ever going to occur. “Whether one is a defender (of oneself or others) or a perpetrator … may depend on perspective. Some reports of defensive gun use may involve illegal carrying and possession … and some uses against supposed criminals may legally amount to aggravated assault… Protecting oneself against possible or perceived harm may be different from protecting oneself while being victimized.” 
- Poor definition of what defensive use even means. “Imagine, for example, measuring defensive gun use for a person who routinely carries a handgun in a visible holster. How many times has this person ‘used a handgun, even it was not fired, for self-protection?’ … In this regard, much of the debate on the number of defensive gun uses may stem from an ill-defined question…” 
Because of these and similar arguments, I am throwing out the upper limit and replacing it by 300,000 defensive gun uses per year (a rate of 95.6). I’ll discuss the implications shortly.
Imagine if there were no more guns: So what if, magically, there were no more guns? Guns would no longer be used in crimes. Whether this would have any effect on the crime rate is unclear. Many perpetrators would probably just choose a different weapon. We can expect the crime rate to drop anywhere from zero to 42 with perhaps a most likely value of 10. The benefit of defensive gun use would disappear. The range of increase is between 31.9 and 95.6 with a most likely value in the middle. I converted these ranges to probability curves and computed the possible change in crime rate. The result is shown in the two figures below.
Based on this calculation, the most probable change is an increase in crime rate of about 50. Thus, we conclude that, even throwing out the suspect value of 2.5 million defensive gun uses per year, and given the existing gun culture in the United States, a complete gun ban would probably result in a modest increase in violent crime rates.
Of course, guns will not really disappear overnight – what we are facing today is stricter laws on purchasing guns. If stricter gun controls make no change on defensive gun use, but do remove the guns from criminals, the increase in crime rate may be much smaller. Furthermore, if the value of 2.5 million counts incidents that are actually criminal, then we should expect a noticeable decrease in crime rates. If, however, the value of 2.5 million is credible and real, then it would constitute a very strong case for allowing, even encouraging, widespread weapon ownership (by “law-abiding citizens” of course).
The conclusion, then? Given the strong disagreements in the rates of defensive gun use, the possible implications of those disagreements, and the importance of defensive gun use to determining changes in crime rate, it is impossible to draw any satisfying conclusion. The various sides will, of course, cherry pick facts to make their case, but this kid ain’t gonna believe them. He’s just waiting for the experts to offer a believable value for defensive gun use.
If there is any conclusion to draw it’s that this matter isn’t in the end all that important. There is no way that there will be a complete gun ban in the US any time soon. The strictest outcome might involve universal and strict background checks at all gun outlets with prohibition on some types of weapons and ammunition. None of this can be expected to hinder “law-abiding citizens” from using handguns defensively nor will it decrease crime much. As far as an argument for or against current gun control legislation, self-defense doesn’t provide a strong case for either side.
Notes and References:
- This number is oversimplified. Crime rates are nowhere close to uniform; they vary across ages, sex, and socioeconomic status. Most victims of violent crime are minority men between 18 and 24 (http://bjs.ojp.usdoj.gov/content/pub/pdf/cv11.pdf). Based on age-specific rates, if you are 40 years old your chance of being the victim of a crime in the future is only 8%. If you have never been the victim of violent crime, you probably belong to a demographic that has an even lower crime rate than the average for your age. Nonetheless, the chance of being the victim of a violent crime is sizeable, given the seriousness of the consequences. Self-defense is not a consideration that we can just ignore.
- http://bjs.ojp.usdoj.gov/content/pub/pdf/wuvc01.pdf. Even though the figure is a bit outdated, it’s probably OK since the percentage has been fairly steady.
- National Research Council, Firearms and Violence, p 102, http://books.nap.edu/catalog.php?record_id=10881.
- ibid, p 106
- ibid, p 108