OSD 291: Who controls the gats controls the future
Last week, a local legislator in New York City introduced legislation to require city gun stores to display cigarette-style warning posters:
There are a few interesting things about this.
First, it’s a good example of the “I can tolerate anything except the outgroup phenomenon”. Of all things that might be needlessly killing people in New York City, “The city’s gun stores are insufficiently regulated” is not one of them.
Second, reducing murder and suicide by regulating NYC’s gun stores is like solving alcoholism by regulating Saudi Arabia’s liquor stores. There are six gun shops in all of NYC. Two are in Manhattan — one is Westside Rifle & Pistol in Rep. Bottcher’s district, and the other is the Beretta Gallery on the Upper East Side. NYC overall has 0.07 gun stores per 100,000 people. Wyoming, the state with the highest density of gun stores, has 1391x the density of gun stores at 97 per 100,000 people. Even New Jersey, the state with the lowest density of gun stores, has 3.5 per 100,000 people — 50x the gun store density of NYC.
Third and last: people will argue about the way the messaging harasses gun owners, or the First Amendment violations of the compelled speech. But even ignoring that, there’s one big problem left: is the messaging even true?
The core claim:
A firearm in the home significantly increases the risk of suicide, homicide, death during domestic disputes, and unintentional deaths to children, household members, and others.
This is politics in scientists’ clothing. It models a gun not as a tool, but as an entity that causes outcomes on its own. It’s like saying that having a bottle of whiskey in the house increases your risk of cirrhosis. In a purely correlational sense, sure, you can make a graph that shows a relationship. But the whiskey’s presence isn’t what caused the cirrhosis. Most people with a cabinet full of liquor will never have a problem with it. A small percentage will find a way to have a problem with liquor no matter where they encounter it.
From “OSD 241: Violence is like getting into Harvard”:
The worldwide average homicide rate is roughly 6.1 murders per 100,000 people. But that doesn’t tell you anything about your own personal murder risk. You can keep going and make it even more useless — zoom out further to the whole universe, whose murder rate asymptotically approaches zero. Equally useless for telling you whether you should walk outside alone at night.
This applies to any process where you try to translate group-level outcomes into your own personal odds for seeing the same outcome. A story from memory (because Google isn’t turning it up): Paul Graham was talking about Y Combinator acceptance rates on a forum once, 10+ years ago. He said something to the effect of, “Y Combinator having a 2% acceptance rate doesn't mean that you have a 2% chance of getting in. It means that 2% of applicants have a near-100% chance, and 98% of applicants have almost zero chance.”
…
This misapplication of group-level statistics drives a lot of false ideas. Especially around violence. When you hear about St. Louis or Baltimore being dangerous cities, what you’re actually hearing is that you need to zoom in. In most parts of those cities, your risk of being murdered is effectively zero. Then there are a handful of neighborhoods where murder rates are many times higher than the already-very-high city-wide number, and those neighborhoods pull the average up.
Group-level stats vs. individual stats also explains the cultural reaction to mass shootings, a vanishingly unlikely cause of death. But that is violence that randomly targets people who in all other respects have managed their risk of murder down to effectively zero, so it would be surprising if they didn’t react strongly.
This idea of guns as inherently harmful is always going to be common in places where guns are uncommon. We’re cherry-picking a bit to point at what an NYC legislator is doing. Of course their ideas will view guns as alien. Because in that environment, guns are alien — and the point of the posters is to reinforce that.
This is about controlling the narrative about guns. Because whoever controls that will determine the questions that people think to ask. But there’s no way to stop people from learning about guns on their own. If we do a good job at that, the posters won’t matter.
This week’s links
Bypassing airport security via SQL injection
The employment status check is the most critical component of these processes. If the individual doesn’t currently work for an airline, they have not had a background check and should not be permitted to bypass security screening or access the cockpit. This process is also responsible for returning the photo of the crewmember to ensure the right person is being authorized for access. So how does this work, when every airline presumably uses a different system to store their employee information? That is what we were wondering, and where it gets interesting…
About Open Source Defense
Merch
Grab a t-shirt or a sticker and rep OSD.
OSD Discord server
If you like this newsletter and want to talk live with the people behind it, join the Discord server. The OSD team is there along with lots of readers.
> A firearm in the home significantly increases the risk of suicide,
I am definitely preaching to the crowd with this, but, I have to say it every time it comes up.
This is a very very archetypal example of a thing I notice all the time, that doesn't have a good name. The best I have is a folksy saying I heard growing up: "The average American has one breast and one testicle".
Our benevolent social planners do their social science and come up with their highly abstract, aggregated statistics. Then, everybody acts as if those averages and aggregates apply to individual cases. Averages _cannot_ apply to individual cases.
In the case of firearms suicides, and apologies in advance to all the Bayesians in the audience for oversimplifying statistics, but, firearms in the home may increase the _aggregate_ suicide risk. But for any given individual, their suicide risk is either 0 or 100%. And, most people know themselves well enough to know which category they fall into. The people whose risk is 0, it doesn't matter if they own guns, owning a gun won't change that risk. And the people whose risk is 100%, they probably know that about themselves, and so the study telling them it increases the risk is not useful, because they already know-or-should-know their own proclivities, and self-select out of owning a gun.
Like, social science aggregate statistics are good as starting points when you have no other data, but, people have extra information _about themselves_ which completely supersedes any social science data. Like, is there really someone out there who's like "yeah, life is pretty good, but I read a study that says that owning guns increased the suicide rate, so now I want to kill myself.". I don't believe that such a person exists on this specific subject, but I have personally encountered people who are like this for lower-stakes issues, and I fundamentally don't understand it.
Something Something Seeing Like A State
I’d rather them have six of these posters up in the six affected gun stores than have them shove another piece of paper in the box of every gun sold.