Similarly, while the idea of new laws gets most of the attention, some projects have focused on enforcing existing laws more effectively, or changing policing strategies the way Boston's Operation Ceasefire did in the 1990s. In fact, launching community-based programs has proven to be one of the most effective strategies for reducing gun violence. (See: What has worked, below.)
There have also been programs based on other principles, such as public safety education and gun buy-back campaigns. The White House proposals (see below) address both gun access and gun use, and include both new laws and enhanced enforcement of existing laws.
(More: Aiming for Evidence-based Gun Policy, Phillip Cook and Jens Ludwig)
Would fewer guns result in less gun violence?
Suppose it were possible to reduce the number of guns in circulation, or make it harder for people to get a gun. Would gun violence go down?
Although countries that offer easier access to guns also have more gun violence, at least among developed nations, this doesn't necessarily mean that more guns cause more deaths. People may own more guns in dangerous places because they want to protect themselves. It's also possible that gun ownership is a deterrent to crime, because criminals must consider the possibility that their intended victim is armed.
Economist John Lott did extensive work on this question in the late 1990s, culminating in his 1998 book More Guns, Less Crime. He studied the effect of right-to-carry laws by examining violent crime rates before and after they were implemented in various states, up until 1992, and concluded that such laws decreased homicides by an average of 8%. Lott's data and methods have been extensively reviewed since then. A massive 2004 report by a 16-member panel of the National Research Council found that there was not enough evidence to say either way whether right-to-carry laws affected violence. In 2010, different researchers re-examined Lott's work, the NRC report, and additional data up through 2006, and reaffirmed that there is no evidence that right-to-carry laws reduce crime.
Meanwhile, other studies have suggested that reduced access to guns would result in less crime. These studies compared homicide rates with gun availability in various states and cities. The most comprehensive estimate is that a 10% reduction in U.S. households with guns would result in a 3% reduction in homicides. Internationally, the effect of reductions in gun ownership might be much larger. This might have to do with the large number of guns already available in the U.S.: Any reduction in gun violence hinges on whether gun control laws would actually make it prohibitively difficult to get a gun.
(More: Gun Rhetoric vs. Gun Facts, Factcheck.org; The Impact of Right to Carry Laws and the NRC Report: Lessons for the Empirical Evaluation of Law and Policy, John J. Donohue III, Abhay Aneja, and Alexandria Zhang)
Does gun control result in fewer guns?
In principle, it's not necessary to keep guns away from everyone, just those who would misuse them. Background checks are promising because a high fraction of future killers already have a criminal record. In one study in Illinois, 71% of those convicted of homicide had a previous arrest, and 42% had a prior felony conviction.
Yet current federal gun regulation (see above) contains an enormous loophole: While businesses that deal in guns are required to keep records and run background checks, guns can be transferred between private citizens without any record. This makes straw purchases easy. In other words, these laws may generally make guns harder to come by, but those who really want them can still obtain them through private sales.
Also, although it's generally illegal to sell guns across state lines, in practice this is very common. There's abundant evidence that under the current system, guns flow easily between legal and illegal markets. Washington, D.C,. banned all handguns in 1976, and Chicago did the same in 1982. In neither case did the percentage of suicides using firearms -- considered a very good proxy for general gun availability -- fall significantly.
Similarly, Illinois had a background check requirement before 1994, so the local gun market was not affected by the passage of the Brady Act, but gun trace data shows that criminals simply switched from smuggling guns from out of state to buying them illegally in state.
(More: Where 50,000 Guns Recovered in Chicago Came From, New York Times)
How often are guns used in self-defense?
There are no comprehensive records kept of incidents where guns are used in self-defense, so the only way to know is to ask people. Data from the National Crime Victimization Survey suggest that a gun is used in self-defense about 60,000 to 120,000 times each year. Several other surveys confirm this estimate. By comparison, each year about a million violent crimes involve guns. This means guns are used to commit a crime about 10 times as often as they are used for self-defense.
A few surveys in the early 1990s suggested that there are millions gun self-defense incidents each year, but there are very good reasons to believe that these estimates were improperly calculated and these numbers are way off, more than 10 times too high. If the numbers really were this high, this would imply that pretty much every gunshot wound in America is the result of somebody protecting him or herself.
Even among the more accurate surveys, according to a panel of criminal court judges who reviewed survey respondents' stories, about half the time the gun use was "probably illegal," even assuming the gun itself had been purchased legally.
(More: Gun threats and self-defense gun use, Harvard Injury Control Research Center)
Won't criminals kill with other weapons if they don't have guns?
The crux of this question is whether most homicides are planned, or whether killers more often confront their victims with no clear intention. In the second case, adding a gun could result in a fatal shooting that would otherwise have been avoided.
The evidence that weapon does matter goes back decades. In 1968, Franklin Zimring examined cases of knife assaults versus gun assaults in Chicago. The gun attacks were five times more deadly. Moreover, the two sets of attacks were similar in all other dimensions: age, sex, race, whether the victim knew the assailant beforehand, and so forth. A few years later, he repeated his analysis, this time comparing small and large caliber guns. As expected, the victim was much more likely to die from larger caliber guns.
Although it is surely true that a determined killer cannot be stopped by the absence of a gun, this type of evidence indicates that many homicides are unplanned. The outcome depends, at least partially, on the weapon at hand. In that restricted sense, guns do kill people.
(More:Crime is Not the Problem: Lethal Violence in America, by Franklin E. Zimring and Gordon Hawkins)
What has worked to reduce gun violence?
This is not an easy question to answer, because crime rates can decline for a wide range of reasons. For example, violent crime rates declined sharply all across the country in the mid-1990s, regardless of whether a given area had tightened its gun laws. So based on a naive interpretation of the numbers, any attempt at reducing gun violence in 1995 would have appeared successful by 1998. Then there is the problem of comparing different states or cities: Circumstances differ, and what works in Memphis may fail in Detroit.