It's a popular argument among gun supporters that "guns don't kill people, people do." In the trivially reductive sense, of course, it's true, but that sense is irrelevant to gun laws. In the same sense, smog doesn't kill people -- after all, you have to live in the area and breathe it to suffer injury. The gun supporter might argue that someone has to breathe (at least assuming they don't move out of the area), whereas they don't have to pick up a gun -- but then, none of the kids or teachers at Sandy Hook were the one holding the gun.
That is, much like smog, just existing in the vicinity of a gun puts your life at risk. Or, more pointedly:
Guns do kill people. As sure as any other ambient environmental poison. Guns have legs: toss a hundred guns into a neighborhood, and you can be sure more people will die.
With that in mind, I searched the literature, wondering how good this comparison really was. As it turns out, the necessary calculations are out there. We can make a rough estimate, good enough for conversational purposes, of the comparative effect on mortality of increases or decreases in the density of guns, or the density of smog. After the link, I offer the supporting calculations, and the summatory conversation point for your use.