Lane Kenworthy, The Good Society
November 2017

“The right of the people to keep and bear arms shall not be infringed” says the US Constitution’s second amendment (given the need, its authors presumed, for a militia). Yet we probably would be safer if our society had fewer guns. While Americans are justifiably horrified by mass shootings, they’ve killed, on average, fewer than 20 people per year over the past half century.1 The big numbers are in non-mass shootings: around 10,000 Americans are murdered with guns each year, accounting for about two-thirds of all homicides. And 20,000 kill themselves with a gun.

No sensible person disagrees with the notion that access to guns should be restricted. But to what degree? And in what ways?


The over-time pattern in the US since 1960 suggests little or no effect of guns on the homicide rate. As figure 1 shows, the share of homes that have a gun has been relatively flat throughout the past half century, while the number of guns per person in the country has increased significantly. Neither trend looks to be useful in predicting the homicide rate, which jumped sharply in the 1960s and 1970s and then fell just as sharply beginning in the early-mid 1990s.

Figure 1. Guns and homicides
Homes with a gun data sources: Gallup; General Social Survey,, series owngun. Guns per person: data are available for only two years, 1968 and 2009. Data source: William J. Krouse, “Gun Control Legislation,” Congressional Research Service, 2012, pp. 8-9. Homicides data source: Bureau of Justice Statistics, using FBI Uniform Crime Reports data.

Compared to other rich nations, the United States has a very high prevalence of guns and a very high rate of gun killings, as figure 2 shows. This seems to support the notion that more guns results in more homicides. However, countries differ in lots of ways, and America is particularly idiosyncratic. If we remove the US from the picture, the apparent correlation between guns and gun homicides weakens considerably, which suggests reason to doubt the importance of guns.

Figure 2. Guns and homicides
Solid line: all 21 countries. Dashed line: US excluded. Gun homicides: per 100,000 population. Data source: United Nations Office on Drugs and Crime. Guns: per 100 persons. Data source: “Gun Homicides and Gun Ownership Listed by Country,” The Guardian, 2012, using Small Arms Survey data.

In 1994, the Clinton administration passed a crime law that included a ban on some types of assault weapons. It expired after ten years and hasn’t been reinstated. The ban appears to have had little impact on homicides.2 This is partly because manufacturers made some alterations to assault-type guns that allowed them to get around the law, and also because assault weapons, although scary, are used in only a small share of gun killings.

A more comprehensive change in gun policy in Australia seems to have had a larger impact. After a mass shooting that killed 35 people, Australia in 1996 banned semiautomatic and automatic rifles and shotguns, instituted a mandatory buyback program to take existing guns out of circulation, and instituted background check and mental health requirements for purchases of other types of guns. In the subsequent decade, gun homicides dropped by 59%.3

Comparison across states might be informative. Figure 3 shows a measure of the strength of a state’s gun regulations on the horizontal axis and gun deaths (homicides, suicides, and accidental deaths) on the vertical axis. The association is quite strong. A recent, thorough cross-state analysis concluded that universal background checks are associated with lower gun killings, while many other regulations aren’t.4

Figure 3. Gun regulations and gun deaths
Gun deaths: per 100,000 population. Homicides, suicides, and accidental deaths. Data source: Nicholas Kristof, “How to Prevent Shootings,” New York Times, 2017. Gun regulations: restrictiveness of state laws. Data source: Giffords Law Center to Prevent Gun Violence, The line is a linear regression line. The correlation is -.77.

We also can compare across cities. A study published nearly three decades ago compared Seattle with Vancouver. These two cities were similar demographically and economically and just 140 miles apart, but Vancouver had much more restrictive gun laws. The analysts found that burglary and assault rates were similar in the two cities, but gun killings were much more common in Seattle.5

The best evidence is quasi-experimental: comparison of changes in gun prevalence and changes in gun killings across countries, states, or cities. Studies pursuing this type of approach have yielded mixed findings.6

If reducing the prevalence of guns would in fact decrease homicides, how large might the effect be? The average gun homicide rate in other rich democracies is one-tenth of the US rate. Total gun homicides in the US are about 11,000 a year, so reducing our rate to the average among comparator nations would translate into about 10,000 fewer gun homicides per year. That’s a lot of lives, though it pales in comparison to, say, the 600,000 Americans killed by heart disease each year, the 580,000 killed by cancer, the 120,000 by accidents, the 100,000 by alcohol, and the 25,000 by illegal drugs.7 What we don’t know is whether a significant reduction in guns actually would get us such a large reduction in homicides.

We may be on firmer ground when it comes to suicides. About 20,000 Americans kill themselves with a gun each year, approximately half the total number of suicides. Only 11% of all suicide attempts succeed, but the rate is 90% for those who use a gun.8 Reducing the prevalence of guns thus might yield a sizable drop in the number of suicides.


There is little chance the federal government will adopt a policy of prohibition. As figure 4 indicates, fewer than half of Americans favor a ban on assault weapons and only a quarter support a ban on handguns. Even if there were more public support, the National Rifle Association (NRA) could probably ensure blockage of legislation in the House or Senate or via a presidential veto. And in the unlikely event that a ban became law, the courts might well rule it unconstitutional.9

Figure 4. Favor banning guns
Share of US adults. Handguns question: “Do you think there should or should not be a law that would ban the possession of handguns, except by the police and other authorized persons?” Assault rifles question: “Are you for or against a law which would make it illegal to manufacture, sell, or possess semi-automatic guns known as assault rifles?” Data source: Gallup, “Guns,”

Figure 5. Favor stricter requirements for gun purchases
Share of US adults. More strict question: “In general, do you feel that the laws covering the sale of firearms should be made more strict, less strict, or kept as they are now?” Data source: Gallup, “Guns,” Police permit question: “Would you favor or oppose a law which would require a person to obtain a police permit before he or she could buy a gun?” Data source: General Social Survey,, series gunlaw. Background check question: “Would you favor or oppose a law which would require universal background checks for all gun purchases in the US using a centralized database across all 50 states?” Data source: Gallup, “Guns,”

Another option is to more tightly restrict gun purchases.10 As figure 5 shows, this tends to have the support of a healthy majority of Americans. Nearly 90%, for instance, support requiring a universal background check before a person is allowed to buy a gun, and 75% support requiring a police permit.


The United States is unique among rich democratic nations in the pervasiveness of guns. There are approximately 300 million guns in the country, nearly one for each person, and roughly 40% of homes have one or more. Reducing the number of guns in America very likely would reduce homicides and suicides, but we don’t know by how much.

Public support for banning guns is low, and opponents have other means of blocking a ban in any case. Prospects seem better for tightening restrictions on the purchase of guns.

  1. Bonnie Berkowitz, Lazaro Gamio, Denise Lu, Kevin Uhrmacher, and Todd Lindeman, “The Math of Mass Shootings,” Washington Post, 2017. 
  2. David Cole, “The Terror of Our Guns,” New York Review of Books, 2016. 
  3. Julian Santaella-Tenorio, Magdalena Cerda, Andres Villaveces, and Sandro Galea, “What Do We Know About the Association Between Firearm Legislation and Firearm-Related Injuries?,” Epidemiologic Reviews, 2016. 
  4. Bindu Kalesan et al, “Firearm Legislation and Firearm Mortality in the USA: A Cross-Sectional, State-Level Study,” The Lancet, 2016. 
  5. John Henry Sloan, et al, “Handgun Regulations, Crime, Assaults, and Homicide: A Tale of Two Cities,” New England Journal of Medicine, 1988. 
  6. Philip J. Cook and Jens Ludwig, “The Social Costs of Gun Ownership,” Journal of Public Economics, 2006; Santaella-Tenorio et al, “What Do We Know About the Association Between Firearm Legislation and Firearm-Related Injuries?” 
  7. The numbers for alcohol and for illegal drugs are from Mark A.R. Kleinman, Jonathan P. Caulkins, and Angela Hawken, Drugs and Drug Policy, Oxford University Press, 2011, p. 19. 
  8. Kim Soffen, “To Reduce Suicides, Look at Guns,” Washington Post, 2016. 
  9. The second amendment of the US Constitution: “A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed.” On one interpretation, this guarantees the right of individuals to own guns. On another, it guarantees such a right only if the person is part of a state militia. 
  10. Nicholas Kristof, “A New Way to Tackle Gun Deaths,” New York Times, 2015.