Households with firearms suffer more injury due to accidents than they prevent via deterrence. I think you're mistaking correlation for causation. Consider: a

Households with firearms suffer more injury due to accidents than they prevent via deterrence. I think you're mistaking correlation for causation. Consider: a household that's more likely to be in danger is also more likely to have a gun. For example, a drug dealer is frequently in dangerous situations, and so is likely to have a gun. He's also more likely to be injured or killed than your or I. Is that because he owns that gun? It seems unlikely. Many more legitimate professions and activities carry risks like that. So simply showing that correlation isn't anywhere near sufficient to demonstrate that the gun owner would be safer without his gun.