Your comment is actually a perfect example of why I have issues with using per capita as the basis of the analysis.

I adjusted the OCED numbers to be on a per-gun basis as opposed to a per capita basis and non US OCED countries combined were more likely to have rampage shootings per gun than the US.

There are not issues in the US with social programs, mental health, etc as compared with other countries when looking at mass shootings. The only number that plays any significance is the number of guns per capita to influence the number of shootings per capita.

If other countries had as many guns per person as the US, the data indicates they'd proportionally have the same amount of shootings per person or more.

To a certain extend it makes sense that more readily available guns cause more shootings, but I highly doubt that it's a simple linear relationship.

If I don't have access to a gun I can't go on a rampage shooting. If a friend of mine has a gun I could go on a rampage shooting, but I'm probably less likely to do it on impulse. If anybody in my household owns a gun I can likely go on a rampage shooting whenever I like. But if I own 10 guns I don't see how that makes a rampage shooting any more likely.

The proper statistic would be rampage shooting per household with at least one gun. But even that's probably still very flawed (are dangerous people more likely to own guns?)

In the USA, people who own guns are likely to own more than one gun. They may be hunters or collectors. Per-gun is fairly meaningless.

I recently read statistics that the number of households owning guns has decreased from 50% to ~30% since 1977, while the number of guns increased over that time period.

I did some back of the envelope checks on those statistics, and they would require us gun owners to have arsenals worth $100,000 or so.

Note that one's willingness to tell an anonymous voice on a phone that one owns guns changes with the political climate towards guns.

How do you get $100,000? If 30% of households own guns, that's about 37 million gun-owning households. If there are 300 million guns, that's about 8 guns per gun-owning household. If that's $100,000, that means a gun is worth on average $12,500. I'm hardly an expert on guns but that seems about an order of magnitude too high from what I do know.

I see. Is there some context in the gun world that I'm missing out on? Since the comment you replied to said nothing about AR-15s or indeed anything beyond the number of gun-owning households and number of guns overall changing over the years.

The number of households who answer in the affirmative when asked by pollsters has certainly dropped in that time period.

I have serious doubts that the actual number of gun owning households has dropped that much.

The problem here is that you're assuming there's a correlation between guns per capita and shootings per capita, which is similar to the (potential) fallacy that the OP made: that there's a correlation between gun restriction laws and shootings per capita.

Why is the USA then the extreme outlier then when it comes to the sheer number of shootings also highest concentration of guns anywhere in the developed world?

Correlation does not imply causation. As noted elsewhere in this discussion, there are plenty of other potentially relevant correlations, and as also noted, if you cast the net wider than the ODEC we no longer are an "extreme outlier".

I'm not sure how we've deluded ourselves into thinking that the two are not directly related. As other commenters have mentioned, it isn't big scary guns, but the hundreds of millions of handguns everywhere and easily available in the USA that do most of the damage.

See where I just commented on the fact that our murder rate has halved since the 1970-90s period while gun ownership rates have not changed, strongly suggesting we should look elsewhere, perhaps, as I note, demographics: https://news.ycombinator.com/item?id=9838517

The murder rate overall has gone down, but we are still an outlier when it comes to the number of guns and gun homicides.

This is a strange definition of correlated, where one variable changes and the other does not.

Our outlier high gun homicide rate is dependent on our extremely high number of guns easily available to commit those homicides.

So you're saying these two variables are correlated, even though they are not, in fact, correlated?

They are correlated.

Who's on first?

Perhaps people afraid of being the next shooting victim buy lots of guns.

I have a lot of guns, and I know lots of people with lots of guns.

None of us buy them beause we're "afraid of being the next shooting victim".

Ah, so per resonators comment, you must have bought them because you want to be the the next shooting perpetrator.

But drug violence has plummeted from its high point. The drop in drug prices (engineered by the CIA?) meant folks no longer want to get killed over $50.

Do these counts of the number of guns include the number of illegally held guns as well?

The blog concludes that the abuse of statistics is politically motivated by a vested interest in keeping gun-safete legislation minimal. This is not the same as implying that there should be more gun-safety legistation.

I know this is just pissing in the wind (and you just used the term from the article), but "gun safety" already means something. <https://en.wikipedia.org/wiki/Gun_safety> This type of language engineering annoys me.

Unless social programs turn out to be unexpectedly cheap or gun law enforcement unexpectedly expensive, gun laws are probably the cheapest, smallest-delta intervention against mass shootings.

Gun control cost the Democratic party control of the Congress in 1994 and the Presidency in 2000. That's about as "expensive" as policies get in the US....

And as noted in this discussion, any attempt to squeeze tight enough that there might actually be an effect on mass shootings of the current type would replace them with fantastically more "mass" shootings of the civil war type, which tend to be only exceeded by foreign invasions as "expensive".