The ultimate put up or shut up gun study (hypothetical)
Let's say we wanted to do a study that directly addresses the question about guns and safety that people are most likely to have: "If I bring a gun into my household, will it make me and the other people there more or less safe?"
Here's the study: we start with a very large sample that is representative of the American population (say 100,000) and we follow them for, say, five years. Along the way, we record every serious injury and every death (including suicide). At the end of the five year period, we tot up the figures for gun and non-gun households (pro-rating for households that acquired or disposed of guns along the way).
Now, either the gun-possessing households are more safe, or less safe than the non-gun households, or else there's no significant difference. Should settle the issue once and for all.
Why, you might ask, include all deaths and injuries? Gun people point out, quite correctly, that just because someone committed suicide with a gun, that doesn't mean they wouldn't have committed suicide by some other means. I'm willing to go further than that. Maybe someone who would have committed suicide with a gun but didn't have one still lives self-destructively and dies (or is injured) in an incident that isn't clearly intentional. (Picture someone who goes out and gets into drunken brawls on a regular basis and one day dies of a beating.)
What about defensive gun uses (DGUs)? Well, if they're successful, they should have an effect on the deaths and injury stats, right?
What about DGUs in defense of property? The study isn't about that. Might be a good subject for another study.
Isn't it pointless to include natural deaths and injuries unrelated to violence in the results? That would open up a whole can of worms about which injuries and deaths would be unrelated to violence that would be affected by the presence of guns. Gun people seem to think that just having a gun wards off attackers. Maybe they're right. Besides, that's the reason for the very large sample size -- the unrelated deaths and injuries should be equal in both groups.
On edit: one other thing: what about people outside the household who die or are injured as a result of some sort of encounter with people inside the household. Well -- this is weird, but I wouldn't count them in this study. A hit man could have a very safe home, I suppose.
I'm expecting to hear from the CDC and the NRA any day.