America Has 120 Guns Per 100 People: New Model Suggests Fear Is Pushing That Number Higher

When someone buys a gun out of fear that their neighbor might have one, and that neighbor buys one for the same reason, something goes wrong at a scale bigger than any one household. A new mathematical study published in Science Advances argues that this cycle, people arming themselves in ways that are individually logical but collectively harmful, is pushing American society into a state researchers call “overarming.” And the problem, the model suggests, gets significantly worse depending on who your social connections are.

Dartmouth College researchers used a type of mathematical modeling also applied to nuclear arms races between nations to model how gun ownership spreads through a population. In the model, as the perceived threat of conflict between any two people rises, the rate of gun ownership that emerges from individual choices quickly outpaces the rate that would minimize harm. Decisions that make perfect sense on their own add up to a collective outcome that leaves everyone worse off.

Real American numbers anchor the study. A 2023 Pew Research Center report cited in the paper found that 30% of U.S. adults own a gun and 42% live in a household with at least one. As of 2026, the United States has roughly 120 guns per 100 people, nearly four times the rate of Canada or the most gun-dense country in Europe.

How Fear Fuels America’s Gun-Buying Arms Race

At the heart of the model is something the researchers call the “provocation rate,” essentially, how likely any two people are to end up in a confrontation. When that rate rises, the model consistently shows that individuals rush toward gun ownership faster than is collectively good for them.

When guns are rare in a community, owning one gives a meaningful advantage if a conflict arises. But as more people arm themselves, those encounters become increasingly dangerous for everyone. Rational individuals keep buying guns anyway because the fear of being the only unarmed person in a conflict overrides the growing collective risk. Researchers call this the “concession cost,” the disadvantage of being caught without a weapon when the other person has one.

A policymaker focused purely on reducing harm across the whole population would set a much lower rate of gun ownership than what actually emerges when everyone acts in their own self-interest. That gap is overarming, and the model found it grows wider as perceived conflict rises and as people become more afraid of being caught unarmed.

Your Social Network Changes the Gun Equation

When the researchers applied their model to real-world social networks, including a gang network, a village network, and a campus network, they found the shape of those connections matters enormously. Tightly clustered communities, where like-minded people tend to associate with one another, can actually slow the initial spread of gun ownership when perceived threat is low. Flip the threat level to high, and the same clustering effect accelerates overarming sharply. Researchers call this a “double-edged sword” effect of social network structure.

Perhaps the most counterintuitive finding involves the most socially connected people. In the model, community hubs with the most ties to others are actually less inclined to own guns than people on the fringes. When a well-connected person buys a gun, it tends to trigger purchases among all of their many neighbors, rapidly erasing any advantage the original purchase provided. Being socially central creates a natural incentive against arming, one that also happens to align with what minimizes harm for the group.

A Self-Reinforcing Spiral

Beyond the base model, the researchers also developed an extended version where perceived threat itself shifts over time based on how many guns are in circulation. More guns increase perceived danger, which encourages more buying, which increases perceived danger further.

When this extended model was fitted to U.S. gun sales data, including the surge during the COVID-19 pandemic, the results suggested that in the American context, individuals perceive the benefits of gun ownership as outweighing its costs. Under certain conditions, this spiral can lock a society into a high-gun, high-fear equilibrium that is difficult to escape. In theory, targeted community interventions and public information campaigns could help dampen the feedback loop, though the authors say future work is needed to test what any of this would look like as real-world policy.

Model Puts Gun Ownership Gap Between Personal Choice and Public Good Into Numbers

Political forces, media influence, cultural history, and policy environments are largely outside the model’s scope, and the authors are upfront that it has not been validated against real behavioral data. What it does offer is a cost-benefit framework capable of generating testable predictions for future research, one that tries to reframe the debate around costs, benefits, and collective outcomes rather than rights alone.

America’s gun ownership rate is nearly four times that of its closest large-nation peers. Fear, social pressure, and network dynamics may be driving that number well past the point where it minimizes collective harm, and the structure of who we know could be making it worse.

---

Disclaimer: This article is based on a theoretical mathematical model and does not constitute empirical proof of any specific cause of gun ownership behavior in the United States. The study has not been validated against individual behavioral data, and its findings should be understood as hypothesis-generating rather than definitive.

---