Guns are now the leading cause of death for children in the US?

Another in the Gunz Warz. From the sainted Obama, no less: "We are failing our children. Guns are now the leading cause of death for children in the U.S.". But is it true?

You'd hope it was true. Big O is a high-profile saint, with the best advice available to him, and certainly no fool. And yet, he appears to be wrong. He appears to be so badly wrong that I suspect I've made some obvious mistake; if so, please point it out to me¹.

My source is the CDC.

Here's a picture (the one below, not the one to the right...) from the CDC site. As expected, the wrinklies die of disease, but for the moment we don't care about them, we care about the young folk.

As sort-of expected, homicide peaks for late-teens / early adult, presumably amongst young men who make regrettable life choices. If I split by sex, we'd probably se something different for the rather more sensible female types <tries it> Ah yes, the sensible women are more likely to kill themselves: H+S = 27% for 15-24, vs 44% for males.

I don't think switching attention to the 10-14 group would help the argument. I wonder if you adjusted the breakdown, you could get something that looked like Gunz-are-Number-One: for 1-14, UI leads, but contains 27 "firearms"; Suicide (#2) has 189 F; Motor Vehicles leads UI as 294... ah, but if you broke MV down into all its constituents, the leading number would be 87, which is less than F. I think that's cheating, though.

You'll be wondering, of course, if "Unintentional Injury" hides some Gunz Deathz.

But no. UI is mostly Carz, with somewhat oddly "Poisoning" coming a very close second. But that's because P includes Drugz. For 10-14, F-in-UI comes up, but only to 5%.

Update

P et al. point me at Current Causes of Death in Children and Adolescents in the United States in defence of O's claim. For the purposes of a pol, that's good enough: for the single year 2020 (though if I painfully search by year, 2021 looks similar to 2020 at least for X95), O's claim is "true". Although you'll notice that is "children and adolescents", not the "children" that O claimed. That's because rather than the what-looks-like-default age binning, that study uses 1-19.

Apart from the age issue, there's another, which is that is all death-by-gun; in the context of school shootings, that's misleading.

But to return to the age issue: why would you use "children and adolescents", if you mean "children"? The obvious reason is that death-by-violence goes up a lot into the 15-24 age group, and including older folk lets you tap into that.

Can I face quantifying that? Oh go on... and so (from the kindly provided https://wonder.cdc.gov/ucd-icd10-expanded.html) I managed to drag out that graph (which if I've got it right is for X95 - which is most of the death-by-firearms-violence, but doesn't include suicide I think, for 2018-2021 (blue) and then I updated it to add just 2018-2019 in red; and remembered to multiply those latter by two to make them comparable, since they are numbers-of-the-fallen. Data in disorganised form here). This rather clearly shows why mixing up "children" and "those old enough to go out and get into trouble" isn't a good idea. Unless you have propaganda in mind, obvs (PF notices the same problem but is more forgiving than me). It is also rather clear that 2020-21 are anomalous; or at least, not like 2018-2019. The "obvious" "explanation" for that is Covid, but YMMV.

Notes

0. The gravestone is from SEH.

1. There was no shortage of kind volunteers. See the comments. And see my update.

Refs

* I’m not sure people fully appreciate how dire the US life expectancy / mortality situation has got.

* War and subsidies have turbocharged the green transition.

* On the Rise of the "Economic Style of Reasoning".

* WILL CLIMATE MITIGATION END UP IN A REAL PICKLE?

* Hourly modelling of conversion of USA48 to wind/solar, with costing and optimisation - Moyhu.

* The Dominion lawsuit showed the limits of Fox’s influence over its audience