Nonetheless, there are some plausible methods for isolating the different factors, using comparison groups or other controls. The most thorough summary is a 2008 meta-analysis where the authors reviewed every prior American gun violence reduction study, examining both the reported effectiveness and the strength of the statistical evidence. Here are some approaches that don't seem to work, at least not by themselves, or in the ways they've been tried so far:
- Stiffer prison sentences for gun crimes.
- Gun buy-backs: In a country with one gun per person, getting a few thousand guns off the street in each city may not mean very much.
- Safe storage laws and public safety campaigns.
We don't really have good enough evidence to evaluate these strategies:
- Background checks, such as the Brady Act requires.
- Bans on specific weapons types, such as the expired 1994 assault weapons ban or the handgun bans in various cities.
These policies do actually seem to reduce gun violence, at least somewhat or in some cases:
- More intensive probation strategies: increased contact with police, probation officers and social workers.
- Changes in policing strategies, such increased patrols in hot spots.
- Programs featuring cooperation between law enforcement, community leaders, and researchers, such as Project Safe Neighborhoods.
There is no obvious solution here, and there's a huge amount we still don't know. But it's possible that combinations of these policies, or variations in a different context, might work better. For example, background checks would probably be more effective if they were also applied to private sales. Also, of course, this list does not include policies that have not yet been tried.
There are no feasible policies that would reduce the rate of gun violence in the United States to that of Western Europe. But we believe there are ways to make a substantial dent in the problem.
(More: The Effectiveness of Policies and Programs That Attempt to Reduce Firearm Violence: A Meta-Analysis, Journalist's Resource. Project Safe Neighborhoods and Violent Crime Trends in US Cities: Assessing Violent Crime Impact, Edmund F. McGarrell, Nicholas Corsaro, Natalie Kroovand Hipple, Timothy S. Bynum)
Are the White House proposals likely to be effective?
There is no way to know whether the recent White House proposals will be effective in reducing gun violence. How can there be, when it's so difficult to assess the actual policies that have already been tried, let alone vague plans? But the White House proposals do at least plausibly target several components of the gun violence problem.
Probably the most significant proposal is the idea of requiring background checks for gun sales between private individuals, not just from licensed dealers (with some exceptions, such as transfers within a family). Private sales are currently the main way guns move between legal and illegal owners. However, the White House has yet to specify how a private seller would perform such a check.
There is less evidence for the effectiveness of banning assault weapons and large-capacity magazines. During the 1994-2004 assault weapons ban, the use of assault weapons in crimes fell, but use of large-capacity magazines increased. This is thought to be largely due to the huge number of already-owned LCMs, which were exempt from the ban on manufacturing and sales. If the new law does not address the LCMs already in private hands, it may be decades before it has any real effect.
Removing legal restrictions that prevent the Centers for Disease Control and other agencies from tracking and researching gun violence is also a sensible idea, and follows a long history of calls from scientists (see: what don't we know).
The U.S. has one of the highest rates of violent crime and homicide, per capita, of any developed country. According to 2008 figures compiled by the United Nations Office on Drugs and Crime, the U.S. homicide rate for 2010 is 5.1 per 100,000 people. Only Estonia's is higher, at 6.3. The next most violent country is Finland, which has a homicide rate of 2.5, half that of the U.S. The remaining 28 developed countries are even lower, with an average of 1.1 homicides per 100,000 people.
But many less developed countries have much higher homicide rates -- for example Columbia (35.9), South Africa (36.8) and Sudan (24.2). This analysis uses the 2012 IMF list of developed countries.
The U.S. also has the highest rate of civilian gun ownership in the world, by far. The best data is from the 2007 Small Arms Survey, which notes:
With less than 5% of the world's population, the United States is home to roughly 35-50 per cent of the world's civilian-owned guns, heavily skewing the global geography of firearms and any relative comparison.
U.S. gun violence has had several decades-long cycles over the past three centuries, but shows a long-term downward trend. Overall homicide rates were similar to Western Europe until the 1850s, but since then violence has declined more slowly in the U.S.
It's tempting to plot the relationship between gun ownership and gun violence across countries, but recent research suggests that gun violence is shaped by "socio-historical and cultural context," which varies regionally, meaning that it's not always possible to make direct comparisons. However, it's still reasonable to compare places with similar histories, and more guns still correlate with more homicides in Western nations. Meanwhile, in developing countries, cities with more guns have more homicides.
(More: Chart: The U.S. has far more gun-related killings than any other developed country, The Washington Post; Facebook post says the U.S. is No. 1 in gun violence. Is it?, Politifact' Gun homicides and gun ownership listed by country, Guardian Data Blog)
What don't we know yet?
A lot! We lack some of the most basic information we need to have a sensible gun policy debate, partially because researchers have been prevented by law from collecting it. The 2004 National Research Council report discussed above identified several key types of missing data: systematic reporting of individual gun incidents and injuries, gun ownership at the local level, and detailed information on the operation of firearms markets. We don't even have reliable data on the number of homicides in each county.
For such sensitive data sets, it would be important to preserve privacy both legally and technically. There have been recent advances in this area, such as anonymous registries. But the Centers for Disease Control, the main U.S. agency that tracks and studies American injuries and death, has been effectively prevented from studying gun violence, due to a law passed by Congress in 1996.
Similarly, anonymized hospital reporting systems are the main ways we know about many other types of injuries, but the Affordable Care Act prevents doctors from gathering information about their patients' gun use. A 2011 law restricts gun violence research at the National Institutes of Health. The legal language prevents these agencies from using any money "to advocate or promote gun control."
This may not technically rule out basic research, but scientists say it has made the issue so sensitive that key funding agencies will not support their work. They point to grant data as evidence of an effective ban. The White House has recently proposed lifting these research restrictions (see above)
(More: NRA Stymies Firearms Research, Scientists Say, The New York Times)