The US Is a Violent Nation
Donald Smith / ThinkerFeeler

 (September 17, 2026) — “Study finds more than half of US adults have experienced physical violence and nearly half sexual violence.”

This might be related to the continued prevalence of wars.

I asked ChatGPT “Is there reliable data comparing levels of violence in the U.S. with levels in other countries?”

ChatGPT said the U.S. has substantially more homicides per capita than other industrial countries (e.g., we have nearly three times the rate of Canada). But “Many countries—particularly some in Latin America, the Caribbean, and conflict-affected regions—have much higher lethal violence. UNODC estimates, for example, put homicide rates around 19 per 100,000 in the Caribbean and 28 in Central America, versus roughly 6 in ” the U.S.  It’s harder to compare non-homicides violence rates between countries, because countries use different definitions.

Here’s ChatGPT’s Full Response:
Yes. There is reliable international data, but the strongest comparisons are for homicide and, to a lesser extent, carefully harmonized measures of intimate-partner and sexual violence. The new U.S. survey in the Guardian is important, but its headline numbers should not be compared directly with superficially similar percentages from other countries unless the definitions and questionnaires match.

The clearest cross-national measure is homicide, because a death is much less dependent on willingness to report, survey wording, or legal definitions. On that measure, the U.S. stands out sharply among other wealthy democracies. In 2023, the UN reported a U.S. intentional-homicide rate of 5.8 per 100,000. World Bank/UNODC data for roughly the same period put Canada around 2, the UK, France, Germany and Italy around 1, and Japan below 1 per 100,000. (UNSD)

So a rough comparison among affluent countries looks like this:

Country	Intentional homicides per 100,000, ~2023
United States	5.8
Canada	~2
United Kingdom	~1
France	~1
Germany	~1
Italy	~1
Japan	<1

That is a very substantial difference: the U.S. homicide rate is approximately 3 times Canada’s and around 5–6 times those of several large Western European countries. It is vastly higher than Japan’s. (World Bank Open Data)

For violence against women, WHO has done something especially useful: it takes hundreds of national studies, attempts to standardize definitions, and statistically models them to improve comparability. Its latest estimates draw on population surveys from 168 countries and areas conducted between 2000 and 2023. WHO estimates that globally about 32% of women have experienced either physical/sexual intimate-partner violence or non-partner sexual violence during their lifetimes. (World Health Organization)

There are large regional differences. WHO’s latest estimates find particularly high lifetime intimate-partner violence in Oceania, sub-Saharan Africa and South Asia. Conversely, recent past-year intimate-partner violence is substantially lower in Europe and Northern America: about 5% of ever-partnered women, compared with 11% globally, 17% in sub-Saharan Africa, 18% in Central/Southern Asia, and 38% in Oceania excluding Australia and New Zealand. (World Health Organization)

But here is an important wrinkle concerning the article you read. The 49% lifetime sexual-violence figure in the new U.S. USVEX study uses a fairly broad conception of sexual violence. The VEX research program includes categories such as sexual harassment as well as sexual assault; related VEX work explicitly includes verbal, cyber, physically aggressive and quid-pro-quo sexual harassment alongside sexual assault. (Newcomb Institute) Therefore you cannot validly say something like:

U.S. sexual violence = 49%, world sexual violence = 32%, therefore America is 50% worse.

Those aren’t measuring the same thing.

The Guardian article reports that USVEX surveyed 2,720 adults in all 50 states, using a nationally representative NORC sample, and found lifetime prevalence of 52% for physical violence and 49% for sexual violence. Women reported considerably more sexual violence—68% versus 29% among men. (The Guardian) Those numbers tell us something important about how widespread violence is within the U.S., but USVEX wasn’t designed as a standardized multinational survey.

There’s also an interesting paradox in international sexual-violence statistics. Countries with strong institutions and greater social acceptance of discussing sexual assault sometimes appear to have higher reported prevalence than societies where sexual violence is highly stigmatized. WHO explicitly warns that all such surveys underestimate prevalence and that underreporting varies substantially across societies; sexual violence is particularly vulnerable to this problem. (World Health Organization) Thus homicide is much safer for statements such as “the United States is more violent than Germany.”

The overall picture I’d take from the evidence is therefore somewhat mixed, but not reassuring. The United States is not exceptionally violent compared with every country in the world. Many countries—particularly some in Latin America, the Caribbean, and conflict-affected regions—have much higher lethal violence. UNODC estimates, for example, put homicide rates around 19 per 100,000 in the Caribbean and 28 in Central America, versus roughly 6 in Northern America. (UNODC)

But among affluent developed democracies, the United States really is unusually violent, especially in lethal violence. That conclusion is considerably firmer than any comparison based on sexual-assault or ordinary-assault surveys.

One striking way of putting it is: Americans probably aren’t uniquely prone to getting into conflicts, abusive relationships or assaults—but conflicts in America are considerably more likely to end in someone dying. The international homicide data establish the latter part very strongly; determining how much of that difference is attributable specifically to firearms versus other social factors requires another comparison.

If you’d like, I can dig into that next and make a U.S. vs. Canada/UK/Germany/France/Netherlands/Sweden/Japan/Australia comparison for homicide, assault, rape/sexual violence, domestic violence, and gun violence using the best comparable datasets. That would give a much better answer to the broader question: “Is American society actually more violent than comparable societies, and if so, in what ways?”

I asked it to respond, but it said the data were unreliable, due to measurement differences. But one thing was clear: “the non-gun homicide rates were almost identical. The enormous difference in total homicide was overwhelmingly firearm homicide.”  So, guns kill.