Why Does the U.S. Have More Gun Deaths?

On the Global Burden of Disease study’s age-standardised rates for 2019, the U.S. firearm death rate is nearly five times France’s, the next highest of twelve other high-income countries. It is roughly 50 times the United Kingdom’s and 100 times Japan’s. Its firearm homicide rate, 4.1 per 100,000 people, is eight times Canada’s, the next highest. Its firearm suicide rate, 5.8, is more than three times France’s and Switzerland’s. On the same study’s later estimates, the U.S. rate was 13.2 per 100,000 in 2023 (Pew Research Center).

The chart and table below give every country’s rate on that one dataset.

How to read it: each bar is one country’s firearm deaths per 100,000 people in 2019, counting homicides, suicides and unintentional deaths together. The rates are age-standardised so that countries with older or younger populations compare fairly. A longer bar means more deaths per person. The red bar is the United States. Every value is in the table below.

Source: Institute for Health Metrics and Evaluation, Global Burden of Disease 2019, as charted by the Commonwealth Fund (April 2023). Totals include unintentional firearm deaths, so the homicide and suicide columns do not always add up to the total. Australia is in the Commonwealth Fund’s charts but is left out here because its three figures there do not agree with each other.

Most U.S. firearm deaths are suicides. In 2024, on final data from the Centers for Disease Control and Prevention (CDC), gun deaths fell for the third year running, to 44,447. Gun homicides dropped 14% to 15,364. Gun suicides rose to a record 27,593, 62% of the total.

The gap with other countries is about half suicide and half homicide, though. In 2019 the U.S. rate was 8.2 per 100,000 higher than France’s. Of that gap, 4.1 was suicide and 3.8 homicide.

If you are thinking about suicide, or worried about someone who is, helplines in the United States and other countries answer calls and texts at any hour.

The U.S. firearm homicide rate is where the international gap is widest. On the Global Burden of Disease study’s age-adjusted rates for 2021, it was 19 times France’s, 33 times Australia’s and 77 times Germany’s (Institute for Health Metrics and Evaluation).

Set against two peers whose police statistics record the method of each homicide, Canada and England and Wales, the homicide gap is a gun gap. In 2024, 76% of U.S. homicides involved a gun, 15,364 of 20,162, on Centers for Disease Control and Prevention (CDC) counts via Pew. The remaining 4,798 did not.

Crude rates, not age-adjusted. United States: CDC counts via Pew, over the Census Bureau’s July 2024 estimate of 340,003,797 (Vintage 2025). Canada: Statistics Canada, 788 homicides and 287 firearm-related in 2024, 1.91 per 100,000, shooting homicides 0.69 per 100,000. England and Wales: the Office for National Statistics’ 8.6 homicides per million from the Homicide Index, and the 32 homicides by shooting it reports from the same index. The two smaller figures are computed from those and the 522 victims in all. The U.S. and Canadian years run January to December; England and Wales runs April to March.

On these rates, Americans are killed without a gun at about 1.2 times the Canadian rate and 1.7 times the rate in England and Wales. With a gun they are killed at 4.5 per 100,000, six and a half times Canada’s 0.69. Against about 0.05 in England and Wales, it is more than 80 times.

Download this conclusion: the gun gap

This card takes apart the idea that Americans are simply more violent. Without a gun the U.S. homicide rate is 1.4 per 100,000, against Canada’s 1.2 and England and Wales’s 0.8. With a gun it is 4.5, against 0.69 in Canada and about 0.05 in England and Wales. Anyone may download it and post it.

Download the card (PNG, 1,200 × 630)

The U.S. gun suicide rate was 7.6 per 100,000 people in 2024, about equal to the record 7.7 of 1977 (Centers for Disease Control and Prevention, age-adjusted, via Pew). In 2024, 57% of all U.S. suicides were by firearm, 27,593 of 48,824.

The question that decides gun suicide policy is whether people denied a gun switch to another method or do not die. The evidence runs against full substitution.

When the Israeli army required soldiers to leave their weapons on base during weekend leave, in 2006, the total suicide rate fell 40% (Lubin et al., 2010). Most of the fall was in weekend firearm suicides, and weekday suicides did not rise to make up for it.

After Australia’s 1996 National Firearms Agreement and buyback, firearm suicides fell faster than before (Chapman et al., 2006; Leigh and Neill, 2010). Both studies also found little sign of a lasting rise in suicides by other means. Suicides by other means fell too in the years after 1997, so how much of the fall the reforms caused is disputed; the policy section below gives the dispute.

In a national study, Conner, Azrael and Miller (2019) looked at suicidal acts from 2007 to 2014 that ended in death or in emergency or hospital care. Of those acts, 8.5% were fatal. Of acts with a firearm, 89.6% were, the highest of any method. Drug poisoning was 59% of acts and 13.5% of deaths.

The explanation the studies offer is the crisis window: the acute period of heightened risk is often only minutes or hours long (Harvard’s Means Matter project, citing Hawton, 2007). An attempt made in that window with a method other than a gun is far more likely to be survived.

On a consistent definition, the U.S. is an outlier among its peers. The definition counts shootings in a public place that killed four or more, excluding the shooter, with victims chosen at random or for what they represent. It excludes family, felony and organised terrorist attacks.

Jason Silva’s count for the Regional Gun Violence Research Consortium found 109 public mass shootings in the United States between 2000 and 2022. Across 35 comparable countries it found 35. The United States held a third of those 36 countries’ population. It had 76% of their public mass shootings and 70% of the deaths. France, Germany and Canada each had four or more over the 23 years; more than half of the comparison countries had none.

Looser definitions count far more: the Gun Violence Archive counts incidents in which four or more people are shot, killed or not, and recorded 502 in 2024. On any definition, mass shooting deaths are a small fraction of gun homicides (Pew, from Federal Bureau of Investigation and Gun Violence Archive data).

Australia’s 1996 gun laws followed the killing of 35 people at Port Arthur, Tasmania, that April. Chapman, Alpers and Jones (Journal of the American Medical Association, 2016) counted 13 fatal mass shootings, five or more killed, in the 18 years before the 1996 laws. They counted none from 1997 to May 2016. The policy section below has the rest of the Australian evidence and what it does not show.

At the end of 2017 the United States had 120.5 civilian firearms per 100 residents, more guns than people and by far the highest rate in the world. Yemen was the next-highest country, at 52.8. The U.S. civilian stock was about 393 million, 46% of the world’s 857 million. U.S. civilians acquired at least 122 million new or imported firearms between 2006 and 2017 (Small Arms Survey, 2018).

About four in ten U.S. adults live in a household with a gun, and 32% own one themselves (Pew, June 2023); the owner share has changed little since 2017.

Set beside the firearm death rates in the international table, the country with by far the most guns per person has by far the most firearm deaths. Among Canada, France, Germany and Japan the pattern is loose. A comparison of five countries cannot separate guns from anything else that differs between them, such as poverty or inequality.

On everyday gun homicide, the weapon data point at handguns.

Campbell and colleagues’ 11-city study of abused women (2003) compared women killed by a partner with abused women who were not. In it, the abuser’s access to a gun raised the risk of being killed by more than five times, over and above the abuse itself.

After the Port Arthur killings, Australia banned semi-automatic rifles and pump-action shotguns and bought back more than 650,000 firearms, about 20% of the civilian stock.

Firearm suicide and homicide were already falling before 1996, and the studies disagree about how much the reforms added. Chapman’s 2016 study found that suicides and homicides by other means also turned downward after 1996, by more than the change in firearm deaths. So it concluded that it was not possible to determine whether the change in firearm deaths could be attributed to the reforms. Leigh and Neill found that the buyback cut firearm suicides, and that its effect on firearm homicide was similar in size but less precise.

The RAND Corporation, a nonprofit research organisation, reviewed the studies. Only one study found convincing evidence that firearm homicide changed, a fall in women killed. RAND judges the evidence consistent with the reforms reducing firearm suicides, mass shootings and killings of women, with doubts remaining for suicide.

The RAND Corporation’s systematic review, “The Science of Gun Policy”, published its fifth edition, the last RAND has scheduled, in January 2026.

RAND’s tiers:

• supportive is the highest;
• moderate means two or more studies found significant effects in the same direction and no study with equal or stronger methods contradicted them;
• limited means at least one qualifying study found a significant effect and none with equal or stronger methods contradicted it;
• inconclusive means studies of comparable rigour disagree, or a single study found only suggestive effects.

Inconclusive describes the studies and says nothing about the law.

The review’s evidence ratings for eight of its 18 policies:

Source: RAND Corporation, “The Science of Gun Policy,” fifth edition, January 2026 (rand.org/research/gun-policy). This site’s companion article, Does Gun Control Reduce Violence?, goes through all 18 policies and 8 outcomes.

The gap is where the guns are (see the homicide section). Whatever explains the gap without a gun, the weapon explains the gap with one.

On suicide, the method used largely decides how often an attempt ends in death. Taking guns out of reach lowered total suicide in the Israeli army (see the suicide section). An abuser’s access to a gun raised a woman’s risk of being killed more than fivefold in Campbell’s study. The RAND Corporation finds supportive evidence for child access prevention laws and minimum ages for buying a gun, and moderate evidence for waiting periods and background checks.

Defensive gun use is real and its scale is unknown. Estimates built on the government’s crime survey run from about 70,000 to 116,000 a year. Kleck and Gertz’s 1993 survey produced 2.2 to 2.5 million. RAND’s review calls the highest estimates implausible and the crime survey’s almost certainly too low. The range is a dispute about how to define and count a defensive use, and researchers and gun-rights advocates tend to cite different estimates.

About a third of American adults own a gun (Pew). A policy that restricts all of them to reach the people who misuse guns asks the law-abiding majority to carry the cost.

The guns used in crime mostly did not come through a shop. In a 2016 survey, 287,400 state and federal prisoners said they had a gun during the offence that put them inside. The table shows where they got it (Bureau of Justice Statistics). Nine in ten did not get it from a retail source, so a check at the counter touches the one in ten who went to a counter.

Gun homicide is concentrated, and the programmes that target it need no new gun law. In focused deterrence, police and community groups confront the small groups of people responsible for a disproportionate share of crime. Across 24 evaluations it was associated with a moderate, statistically significant fall in crime (Braga, Weisburd and Turchan, 2019). But none of the 24 was a randomised trial, and the effects were smaller in the better-designed studies.

International comparisons are confounded. Inequality, poverty, drug markets and incarceration may drive violence on their own, and the 4,798 U.S. homicides without a gun in 2024 are the part no gun law would touch. Switzerland has 27.6 civilian guns per 100 people, among the 20 highest rates in the world, and 0.2 firearm homicides per 100,000.

The Constitution sets the field. In Heller (2008) the Supreme Court held that the Second Amendment protects an individual right to keep and bear arms for self-defence. Bruen (2022) extended it to carrying a handgun in public. It also requires any regulation to be consistent with the nation’s historical tradition of firearm regulation. A policy that works in Britain is not on offer here.

Most of the case that it is not the guns is true and beside the question. Defensive gun use, however common, is about what guns prevent, and the question here is why more people die. The law-abiding majority is real, and the largest share of the deaths involves no crime against anyone else: 27,593 of 2024’s gun deaths were people shooting themselves.

The prisoners’ answers are the argument for checks beyond the dealer’s counter, and the RAND Corporation rates universal background checks as having moderate evidence of reducing total homicide. The Constitution bounds what can be done about the gap and says nothing about why the gap exists.

The confounding argument is the serious one, and the page’s own numbers answer part of it. If poverty and inequality made Americans more violent, the violence would show without guns too. It does, a little: the homicide rate without a gun is 1.2 times Canada’s and 1.7 times England and Wales’s. It does not show at six and a half times or more than eighty.

On the suicide side, the U.S. firearm suicide rate is more than three times any peer’s. Conner’s study found that the method used largely explains how often attempts end in death. Switzerland fits the pattern that more guns go with more gun deaths, rather than breaking it. With 27.6 guns per 100 people, it has a firearm suicide rate of 1.7, the highest of the thirteen countries after the United States and level with France’s. Its total firearm death rate of 2.1 matches Canada’s, which has 34.7 guns per 100.

The evidence cannot yet put a number on how much of the gap the guns cause. The correlation across countries is a pattern, the natural experiments are on suicide and on one country’s reforms, and RAND’s ratings are for individual laws. RAND’s reviews find more guns linked with more violent crime without showing which causes which, and find the evidence on suicide consistent with guns raising the risk. How much poverty, inequality and culture add on top of the guns remains unsettled.

Mostly the guns. Beside Canada and England and Wales, Americans are 1.2 to 1.7 times as likely to be killed without a gun. With one, they are six and a half to more than eighty times as likely to be killed. No other country has as many guns per person. A suicidal act with a gun ends in death far more often than one by any other means. The evidence cannot yet measure the guns’ share of the gap: the studies the RAND Corporation reviewed link more guns with more violent crime but cannot show which causes which.

Download the conclusion

This card gives the answer, mostly its guns, and the reasons. In 2024 Americans were killed without a gun at 1.2 times Canada’s rate and with one at six and a half times. At the end of 2017 the country had 120.5 civilian guns per 100 people. And of suicidal acts that ended in death or hospital care, nine in ten with a gun were fatal, against one in twelve overall. Its bars are the 2019 firearm death rates of the United States, France, Canada and the United Kingdom. It is 1,200 by 630 pixels, the size social networks use for link previews, and anyone may download it and post it.

Download the card (PNG, 1,200 × 630)