47

Joe Biden, in his speech on June 2, said that guns are the number one killer of children, even more than car accidents. Is this claim true?

According to new data just released by the Centers for Disease Control and Prevention, guns are the No. 1 killer of children in the United States of America. The No. 1 killer. More than car accidents, more than cancer. Over the last two decades, more school-age children have died from guns than on-duty police officers and active-duty military combined.

Think about that. More kids than on-duty cops killed by guns. More kids than soldiers killed by guns.

Transcript

Link of the speech.

Laurel
  • 30,040
  • 9
  • 132
  • 118
Amir reza Riahi
  • 543
  • 1
  • 4
  • 10
  • Number one cause of deaths when compared to what? – Joe W Jun 06 '22 at 16:14
  • 5
    "Over the last two decades, more school-age children have died from guns than on-duty police officers and active-duty military combined." Wouldn't school-age children dying from on-duty police officers and active-duty military largely be a subset of children dying from guns? – Acccumulation Jun 07 '22 at 00:36
  • 31
    @Acccumulation Not children being shot by on-duty police officers; it's how many on-duty police officers and military died to guns, compared to how many children died to guns. Of course, that can be misleading without comparing the populations of children compared to police officers and military - I'd guess even in the US, there's more children. – Luaan Jun 07 '22 at 07:06
  • 13
    @Luaan you would think they are not exposed to the same level of risk, though. – clabacchio Jun 07 '22 at 12:50
  • 6
    @JoeW If you're claiming that something is the "number one cause" of something, then you are comparing it against all other causes. I.e. if you ranked every cause of child deaths in a list, the claim is that guns would appear at position 1 in the list. Note that the quote does in fact give some examples of what are being compared (car accidents, cancer) – JBentley Jun 07 '22 at 13:37
  • @JBentley When this question was asked it just said number one cause of death and did not specify children. Also when you talk about causes of death do you include "natural causes" such as old age and health related causes to old age? When looking at all types of death things like heart disease and cancer lead the rest by a large margin but might not be counted when comparing other types of deaths. Honestly the types of death that need to be worried about change as you grow older. – Joe W Jun 07 '22 at 13:44
  • 4
    @JoeW As I said, when you say something is the "number 1 cause" you are comparing against *all other causes*. So yes, that includes natural causes. I'm not sure I follow your point about what you worry about as you get older. Even if the rankings change with age, you can still point to a given age group, list all the causes of death, and put them in a ranked order. (Acknowledged however that the original question wording was unclear). – JBentley Jun 07 '22 at 14:12
  • @JBentley And as I said the question was different when I placed my comment. Also I would argue that trying to compare gun deaths to people dying of "natural causes" is just silly because that is something that can't be controlled, the higher the population the higher the death rate will be. I mean the death rate from guns in America is under 50k a year but the death rate from heart disease (or cancer) is 600k+ a year which means it isn't even a question of which one kills more each year. – Joe W Jun 07 '22 at 14:43
  • 2
    @JBentley You can group deaths any which way you want. How you lump together or split the categories counts for everything. There is no definitive *list* which tells you what the groups are. – Araucaria Jun 07 '22 at 15:01
  • @Araucaria Agreed, but for the purposes of this question that doesn't matter since we are talking about government published statistics, where the rankings *are* derived from a fixed set of categories. – JBentley Jun 07 '22 at 16:24
  • 1
    @JoeW "*the higher the population the higher the death rate will be*". yes, but *in all cases*, so for gun death, cancer, strokes, ... meaning you still can compare *relatively* the rates. This is what is done here. "*death rate from guns in America is under 50k a year but the death rate from heart disease (or cancer) is 600k+ a year*" maybe it's like that (50K? sounds more like a conflict theater to me...), but what is compared is the death **for children**. Not overall. Children rarely die of cancer or of "old age" or similar. – Mayou36 Jun 07 '22 at 17:25
  • @Mayou36 And when I made my comment the question was asking about guns being the number one killer in the US and didn't make any references to the age group in question. – Joe W Jun 07 '22 at 19:42

4 Answers4

70

According to NBC News:

Child vehicle deaths, while falling, still topped firearm deaths in 2020

Nearly 2,400 children ages 1-17 died of vehicle-related injuries in 2020, compared with 2,270 firearm deaths

(Note that this excludes children under age 1, who have a high death rate from birth defects, low birth weight and SIDS and few gun deaths. 19,582 children under 1 died in 2020. Including children under 1 would add 11 firearm deaths and 75 motor vehicle deaths CDC source.)

However, if young adults (18 and 19 year olds) are included, and again children under age 1 are excluded, then Biden's claim becomes accurate.

Considering all children, ages 0-17, the correct information from the CDC source above is as follows:
Deaths in 2020 due to:

Congenital abnormalities: 4,860
Short Gestation: 3,141
Motor Vehicles: 2,462
Firearms: 2,281

DavePhD
  • 103,432
  • 24
  • 436
  • 464
  • 1
    Sorry: I tried to move the comments (which ranged across topics such as the culpability of children who sell drugs, the age range of active marines, stabbing deaths in the UK and how many cars were on the road in 2020, all of which are not about improving the answer) to chat, but I kept getting a cryptic error. – Oddthinking Jun 09 '22 at 00:06
  • 1
    Is this the just released recent data provided by the CDC that Biden was referring to? It's currently June 2022 and this is data for 2020. – Araucaria Jun 09 '22 at 02:06
34

A letter (dated mid April 2022) in the New England Journal of Medicine states that "The Centers for Disease Control and Prevention (CDC) recently released updated official mortality data..." and "...in 2020, firearm-related injuries became the leading cause of death in that age group" (1-19 years old).

Laurel
  • 30,040
  • 9
  • 132
  • 118
Alan White
  • 457
  • 2
  • 2
  • 3
    Recommend you include the graph from that article. – Daniel R. Collins Jun 07 '22 at 17:37
  • 18
    This is misleading. I can think of no place that considers an 18 or 19 year old a child. Maybe an adolescent, but for all intents and purposes, by law they are at least a probationary adult (unable to buy alcohol and gamble, and in some places buy nicotine). – Bardicer Jun 07 '22 at 20:10
  • 31
    The criticisms about the 18-19 age inclusion are off-base, because this is the standard the medical researchers have used for at least some decades. E.g., this article cites 2018 paper by Cunningham et. al. which likewise defines "children and adolescents (1 to 19 years of age)". Bibliography leads to 2003 Brenner article defining "US children younger than 20". Etc. I stopped there, I wonder how much further back? – Daniel R. Collins Jun 08 '22 at 03:22
  • 1
    Biden said "[school-age children](https://www.vocabulary.com/dictionary/school-age%20child)". In the US, most of those graduating high school are not minors, but 18 or even 19 years old. Yet it's common to still call them "kids". @Bardicer – Laurel Jun 08 '22 at 12:23
  • 1
    @Laurel The National Center for Education Statistics says that the median age of those who enroll in college directly after high school is 17. https://nces.ed.gov/pubs2005/2005152.pdf The president is making more than one statement is his speech. One claim is "guns are the No. 1 killer of children in the United States of America." and another is "Over the last two decades, more school-age children have died from guns than on-duty police officers and active-duty military" One claim is about children in the most recent data, the other claim is about school age children over 2 decades. – DavePhD Jun 08 '22 at 14:14
  • 4
    @DavePhD That's not the median age of high school graduates, because all the people who don't get higher degrees aren't included. And even then, so what? If the real percent of people graduating high school at 18+ is 30%, why should people their age be excluded? I interpret "school age children" as being a clarification of what was said previously, not another claim. There's no reason why your interpretation should be considered the one and only, especially when mine matches how he reported the stats. – Laurel Jun 08 '22 at 14:48
  • 2
    @DanielR.Collins: You seem to be arguing that "children and adolescents" is commonly defined as 1-19, so therefore defining "children" as 1-19 is reasonable. I disagree (and I hence think this answer should be updated to clarify it is a near-miss answer), but I fear I have missed your point. – Oddthinking Jun 09 '22 at 00:10
  • This is the answer, regardless of whether people agree with the question or with the source or with the definition of "children". I don't see any point in further pedantry. – Darren Ringer Jun 09 '22 at 19:24
7

First we need to define "children". There are a lot of possible demographic ranges we could use. A common age group is <1-14 for children, and if we include all adolescents, that's expanded to <1-19.

We'll used <1-19, meaning any infant, child, or adolescent under the age of 20; or, 1-19, meaning infants less than 1 are left out. This description is clear and concise, and frequently used in the literature as well as published by government departments. For instance, the CDC publishes for the age groups:

  • <1 (infants)[a]
  • 1-5 (child)[b]
  • 6-10 (child)[b]
  • 11-14 (child)[b]
  • 15-19 (adolescent)[c]

So we'd only need to sum the data from these groups to get a final answer for any sources of mortality. If we were to conduct O.R.

Luckily researchers at think tanks, universities, and non-profits compile data for us.

The next thing we'd need to discuss is how we'd group or cat the various types of disease, accidents, and intentional homicides. For instance, if we want to figure out the leading cause of death, we'd need to know if we should consider heart disease and cancer distinctly or under the same "poor chronic health" umbrella cluster. Also, to avoid conflating homicides with a gun and assault with some other deadly weapon or blunt/sharp object.

This has all been systematized by various bodies. For instance, the WHO use ICD-10 as a medical coding reference. The U.S. uses a slightly modified version ICD-10-CM. Here are the high level codes:

enter image description here

Any death by gun is part of V00-Y99 External causes of morbidity:

enter image description here

For simplicity, we could stick with ICD-10-CM, as it's the most widely-used, though obviously you can tell there's a taxonomic hierarchy here, and we have to go down several levels before we get to specific situations such as homicide with a gun, and even then those are subdivided into the different types of gun used in the assault, and so on. Cause of death can also get ruled as due to multiple codes in certain situations. It's complicated, and any assessment you read has made some assumptions and simplifications.

enter image description here

There are hundreds of different codes, so which ones should get classified together and which ones separate? as when researchers, statisticians, and media report the data. it's a tough question to answer when it comes to diseases and accidents, though the ICD makes it easier to classify it.

Around April-May, 2021, the CDC published data for leading causes of death for various age categories. The CDC used 10-CM.

Researchers at the University of Mich summarized them for "children and adolescents", in a paper submission called "Current Causes of Death in Children and Adolescents in the United States"[1], to the New England Journal of Med. So, this data is not including the infant group (<1). They found that the data says that fire-arm injury (homicide or self-inflicted) overtook motor vehicle — induced death for the first time in 2020 for age 1-19, based on how they categorized causes. They provide a graph:

enter image description here

Note few things, though:

  1. there is no distinguished between gun suicide and gun homicide as partitioned cat
  2. no distinguished between intentional gun vs accident
  3. infant mortality (<1) would affect this by increasing amount of congenital abnormalities, etc., as well as car crash; however, the gun amount would probably stay almost the same.
  4. motor vehicle deaths have depreciated precipitously since 1999, which is one of the reasons why gun cause has exceeded vehicle deaths to become the leading source for the first time. there are several factors[2] for this, including enhance driving assist technology, improved car performance and safety design, the digitization of car systems such as driver trains and braking, as well as lowered speed limits, improved roads and crosswalk availability, road sign increases, and better traffic light placement. less drunk driving (DUI) compared to the 1990s, and glaringly to the 1960s, has also contributed to the decrease. the point is, gun has overtaken driving both because gun increased AND vehicle has decreased.
  5. cancer, heart, and lung disease are all partition from one another and congenital defects. if these LT conditions were aggregated into one, it would enter the running for the #1, 2, or 3 spot.

CDC has yet to publish data for 2021, so we have zero way of telling from the official source as to if the amount is still above vehicle induced through 2021.

FluffyFlareon
  • 503
  • 3
  • 5
  • Please [provide some references](http://meta.skeptics.stackexchange.com/q/5) to support your about the causes of vehicle deaths dropping... or perhaps just remove it; the reasons aren't important for this answer as much as seeing that it has dropped. – Oddthinking Jun 09 '22 at 10:53
  • The CDC has release provisional data for 2021 here: https://wonder.cdc.gov/mcd-icd10-provisional.html – TimRias Jun 09 '22 at 22:40
5

There may be some conflation going on in the data being cited. Biden is also citing "new data" from the CDC, which suggests he's citing 2021 numbers, which aren't available on the CDC's website yet.

However, looking at 2020 data for the whole population, the statement certainly holds true.

CDC data for 2020 (You need to download the CSV in order to get this number, but I'm summing the total for all states in 2020) gives a total of 45,055 firearm-linked deaths in the United States in 2020. This includes homicides, suicides, and accidental discharges.

IIHS data for the same period gives a total deaths by automobile accident: 38,824.

So, for the population in 2020, firearms absolutely topped car crashes as a cause of death, by roughly +20%.

There's plenty of discussion in another answer about the sources of and information about juvenile deaths linked to either cause, but it's possible that there was conflation of the adult statistics.

William Walker III
  • 490
  • 2
  • 6
  • 4
    Are you sure your firearms data is for adults? or is it for all people? Also, for adults or all people, firearms are certainly not the number one cause of death. Drug overdose deaths are much more than firearms deaths https://www.cdc.gov/nchs/fastats/injury.htm and of course there are many more deaths from heart disease, cancer, and COVID. – DavePhD Jun 07 '22 at 13:47
  • 2
    @DavePhD Good catch, fixed miswording. As for #1, I didn't say it was #1, I just said that firearms deaths exceed those of automobile-related deaths. I'm not going to get into the semantics of "killer." – William Walker III Jun 07 '22 at 13:58
  • 3
    But you say "the statement certainly holds true" and the statement is "guns are the No. 1 killer of children in the United States of America. The No. 1 killer". – DavePhD Jun 07 '22 at 14:42
  • 1
    The statement said "guns are the No. 1 killer of children in the United States of America. The No. 1 killer. More than car accidents, more than cancer." So you're saying that the statement is true because you've disregarded the "children" part and the "cancer" part? – Sneftel Jun 07 '22 at 15:44
  • 1
    -1 for incorrect speculation on source of data. As other answers/comments point out, this is clearly from CDC letter to NEJM dated May 19, 2022, including data from 1999 to 2020. – Daniel R. Collins Jun 08 '22 at 03:12