logoalt Hacker News

ninkendo • today at 3:04 PM • 1 reply • view on HN

The data on school shootings is very inconsistent. In most of the sources, a shooting between gangs that happens on a weekend at 2am that happens to be in school grounds is reported as a school shooting. So is any other murder between anyone on campus including staff, etc, even if they're not "mass" or "random" killings. Even suicide is often counted in these lists.

The closest analog we have for "person goes into a school while it's in session to kill people at random" is the FBI's list of "active shooter" incidents, and only some of those are counted as "Mass killings", and the data there says we've has as little as 3 (in 2024) and as many as 15 (in 2023) per year in the past 5 years: https://usafacts.org/articles/what-is-considered-a-mass-shoo...

And that 3-15 number is all mass killings, not just in schools (I don't know how many of those were in schools, but we could stipulate maybe 1 in 3?)

So, the 250-350 number is a couple of orders of magnitude higher than the type of school shootings parents are actually worried about when it comes to needing to text them during the day.


Replies

charted • today at 6:35 PM

I spent an hour or so analyzing these data a little while after we had kids, to see if school shootings were something I should spend any effort or money trying to hedge against, or any brain-space worrying about, whatsoever.

The TL;DR is "no, not at all, it's not a close call".

Even with extremely generous assumptions (and no attempt at all to adjust for differences in distributions of incidents, geographically) I landed on a pretty small, though very much not zero (this was years ago, but IIRC ~2.5%) chance a given child of mine would even be on on campus for the most generous definition of a "school shooting", which can include negligent discharge anywhere on school property, any kind of murder or suicide (the murders are almost always targeted at one person, often against an adult, like jilted-lover or angry-spouse stuff, or targeted gang/drug-trade related violence, plus this can include incidents that involve nobody attached to the school but happen on e.g. a street corner that's school property), and loads of other stuff that aren't "person with gun starts randomly murdering anyone they see in a school", at any time, for their entire k-12 education.

Since the kinds of shootings people think of when they hear "school shooting" are a teensy, tiny sliver of all "school shootings", the odds of being present for that (again, over 13 entire years) would fall into the "vanishingly small" category.

And that's present-during, which I judged as the broadest reasonable category for "suffered significant harm". Actually shot in a school mass shooting, over an entire 13-year school career? Such low odds it may as well be 0%, for the purposes of individual risk assessment.

On the one hand, it's shameful these (the kind of mass, or attempted-mass, indiscriminate shootings people mean when they say "school shooting") happen with the frequency they do, in the US. On the other hand, they're nonetheless rare enough that they're not, on an individual level, really worth worrying about at all.

Policy-wise, the only shift performing this analysis made for me is that I'm now definitely opposed to active shooter drills in schools. I think they're doing quite a bit more net-harm than good.

(Part of why the numbers work out the way they do is that there are an absolute shitload of k-12 schools in the US; it's a big, high-population place. A quick search [I've forgotten the numbers from my original analysis] turns up ~100,000 public k-12 schools, and another 33,000 or so private schools, though IIRC I ignored the existence of private schools for my analysis, as I wasn't sure whether some of the stats I'd turned up included them or not, which was one of the ways I erred on the side of caution)