The Medium article covers why using that correlation to assume causation is a problem, using other experimental variables that were shown to have a correlation with homicides.

No it doesn't. Because there are other studies that back up the causal relationship. He's taken one that shows a clear correlation and explained why that's not enough to craft policy from, and that's true. But studies that show cross time correlation between homicide and gun ownership also exist.

It might not be the only cause, but it being a cause makes sense: another poster showed a news story of a man who had killed 3 black kids who were trying to break into his house. Had he not had a gun, it's unlikely anyone would have died (most home intrusions, even with armed assailants, don't end up killing anyone). Guns provide means for homicide that basically no other tool provides.

> But studies that show cross time correlation between homicide and gun ownership also exist.

So do you believe we could apply that criteria with all of the other correlated variables with firearm homicide from the AJPH article, concluding that they cause firearm homicide?

Do you suppose those intruders would have preferred being cut to death? Why or why not?