The anti-gun group Everytown for Gun Safety says someone is fatally shot or injured in a road rage incident every 16 hours.  Is that number real? Yes and no. The Gun Violence Archive they drew their claim from lumps criminals doing drive-by shootings and criminals being shot by police committing violent acts in with innocent victims. Even if it's only over half of the 500 they claim that is still over 300 per year.

Which means it could also be considered less than one-in-a-million risk and not really worth worrying about. If it is about to happen to you, of course, absolute and relative risk nuance go out the window(1), but on a policy level we do have to consider the whole nation. If 500 deaths demand action in every case, then we should certainly lower the speed limit in cars to 5 MPH and save 200,000 - yet we don't.

Autmobile deaths are pretty well quantified but clean data are hard to come by when it comes to guns and it seems obvious that should not be the case, so why hasn't it been fixed? A big reason is because many of the people who want the data most are also most likely to have a cultural agenda, and their target is not illegal guns, crimes, or anything except extinction of the Second Amendment to the US Constitution.

That prevents many from wanting to use federal funds for it.

Data show that drive-by shootings are so minor, in the "statistical noise" range, that nothing could prevent them. The 300-500 drive-by shootings they are talking about is alarmingly framed as "double" from 2018, but even at double more people are killed by hippos each year. They are down in the 'death due to cows' range if you don't yell at other drivers from your car the way Zack Snyder bots yell at people on Twitter.

The report notes incidents doubled during the pandemic, which has nothing to do with guns, any more than the "Destiny 2" video game is why so many pandemic teenagers lack social skills. Gun sales did go up, three solid months of violence in Portland and then a riot at the Capitol, coupled with reports of cops killing defenseless people while being unwilling to do anything to stop shooters in grade schools make many decide to be more self-reliant. Gun sales have went up in fantastic numbers in California the last two decades but murders have plummeted, so sales are not a valid correlation to anything.

Media accounts always begin with the personal - someone who had a family member shot - and I certainly get that journalists are trained to promote dead bodies high up in the story but doing so won't help us address how better data can prevent dead bodies from being exploited by activists.(2) Or save lives.

Government will not help

Would more rules, protocols, and regulations prevent gun purchase from being abused in the quest for data? History says that will not be likely. It is great to believe government union employees protect your medical data, your kids in school, or anything else, but that may not be true.(3) If government has a political ally and a lot of constituents on one side of an issue, that ally is going to do its work using government-provided aid. When 80,000 farm workers had their names and home addresses exposed to eco-terrorists, do you really believe the Obama administration did not understand a FOIA request and made a mistake? Even crazier, do you believe militant environmentalists gave the names and addresses back when his EPA was caught breaking the law and asked the group receiving them to return the names and addresses of those who might 'use pesticides' without making a copy? 

Let's just accept that government is a 'leaky bucket' when it comes to corruption and favoritism, the best we can do is mitigate that. Yet if we want there to be fewer shootings without states like California banning guns using regulations so arcane no firearm in the world matches the criteria(4), we do need data. And a new paper says data can predict patterns of behavior that may mean a shooter is about to occur - if they get access to who is buying guns.

Bring on the precogs

The analysis believes that patterns of purchasing behavior can predict harm, like in "Minority Report."(5) Some of their methodology is a giant red flag to the science community, like a crime analyst at the California Department of Justice creating a profile for violent criminals using media stories and transaction histories - of just 22 shootings in 22 years. That subjective framework is the basis for their claim that they can predict who will be violent based on their firearms purchases.

Their warning signs are; more gun purchases in the past year, being older for the first purchase, and a history of being unable to buy guns.

That last one shows the approval system works, so that's good news. To get around legal gun ownership requires criminal behavior. Being older before a first purchase is more prone to bias, since it may be that older people now have lost confidence in government they once trusted to protect them, especially if they live in a 'defund the police' city, but with such a small number of incidents it can be inferred.

The confounders make the discussion almost moot; one state, limited data, a lot of potential bias in the profiles, but the solution to that would also be better data.

Yet who can be trusted with such data, when we are dealing with purchasing behavior covered by the Bill of Rights? Academics? Not when the social sciences on the left, which means more anti-gun beliefs, outnumber the right by nearly 99 to 1. The government? When it comes to other science and health issues the government is going to force data into its goals, and in the White House we could be seeing our fourth administration since 2016 next year so name and address data could be safe one administration and then 'accidentally' turned over by the next.

It may be that until we attain more cultural maturity bad data that goes nowhere is better than good data that is manipulated to eat away at constitutional rights.


(1) Absolute risk is often meaningless, but it is what get media headlines. If you have a 1 in 10 million chances of X happening and it increases to 2, that will get "Risk Doubled!" headlines, whereas your relative risk is the only thing that matters.

Credit: Saturday Morning Breakfast Cereal by Zach Weinersmith.

(2) When political vigilantism is not the goal, reforms are made easily. In California,  police can now seize "ghost" guns - made from parts - if a person is red-flagged by a restraining order. 
No gun group opposed it, no legislator, not anyone.

(3) There are more heroic cops than bullies or cowards overall but, again, in relative terms that is little help if you are a parent of a dead Uvalde child who knows 376 law enforcement people were at the Uvalde elementary school not doing anything to stop a maniac - while an armed parent who tried to go in was stopped. And the parent is considered to be the problem by anti-gun activists. 

(4) Due to the hypocritically named Unsafe Handgun Act. no new models of semiautomatic handguns have been approved for sale in California since 2013 because government creating a regulation that was a ban. No handgun in the world existed that matched the new regulations and still does not. The Biden EPA creating a regulation/ban dictating that a weedkiller presence has to be so low it's impossible to use at all is bad, epidemiology can do anything if statistics are tortured enough, but weedkillers are not in the Bill of Rights.

You can't force poor people to pay a fee to vote, or create any regulations that prevent you from reading a newspaper, and that is what California did to guns. Worse, it forced consumers to buy older, less 'safe' guns under the ironic law.

(5) It sounds ethically suspect but there is precedent. If you threaten the President of the United States on Twitter, you may get a visit from the Secret Service. You may not, if there is no pattern of behavior and you are not on psychoactive drugs. If you think the government does not know if you are on psychoactive drugs, oh to be blessed with your naïve soul.