You don't see how a bank not doing KYC and thereby financing terrorists and Pyongyang could be seen as behaving recklessly?
> quite a fig leaf for any entity operating at scale
Sure. It's why a lot of anti-money laundering isn't prosecuted and why OP's complaint strikes me as silly.
> conduct mass surveillance
U.S. AML law is actually somewhat terrible for mass surveillance. Records are held at each private institution. This is why, if you've ever been party to wire fraud, it takes the Feds hours to days to gather records. It's pull, not push.
The practical arms of mass surveillance are the credit/debit card rails. (And, increasingly, crypto.)
Even if that were a reasonable premise, the analogy would then be automakers not doing KYC and thereby enabling drunk drivers to have cars. Which is a much more controversial claim, and it's not at all obvious why carmakers should be the ones deciding who gets to drive.
Which is the reason the premise is unreasonable. Is it reckless for a general store to sell ski masks without investigating their customers? That's not the right place to address the problem.
> It's why a lot of anti-money laundering isn't prosecuted and why OP's complaint strikes me as silly.
Laws that are commonly violated and selectively enforced are bad laws.
> U.S. AML law is actually somewhat terrible for mass surveillance.
Just having it inside of megabanks is already mass surveillance, because then the bank is using it for things and the user can't switch to a bank that doesn't collect the information because the law prohibits that.
Moreover, the data is dangerous as soon as it's collected even if it isn't currently being centrally indexed, because its existence allows the central database to be created ex post facto. Turnkey totalitarianism is bad.
> The practical arms of mass surveillance are the credit/debit card rails.
How are you distinguishing this? Those companies are financial institutions that have to do KYC/AML.
The crime of DUI is not contingent on the actual facts, risk, and danger presented in a given situation.
The association with realized harm is probabilistic in nature.
Drunk driving is an upstream pre-requisite to drunk crashing.
I’m trying to understand this. It is well documented that consuming alcohol impairs thinking, reaction time, motor function, etc. How is the decision to drive a multi-thousand pound vehicle in the proximity of other humans while in that state not itself reckless?
You seem to be arguing that the problem primarily exists when some percentage of impaired drivers start to drive recklessly. But given the baseline impairment, how is the act of driving not inherently reckless?
Drunk driving crashes don’t just happen because someone decided to go crazy and drive 130mph or something similar. They also happen because the impaired person fails at basic aspects of driving like staying in their lane, stopping in time, seeing pedestrians, etc.
> The crime of DUI is not contingent on the actual facts, risk, and danger presented in a given situation.
DUI is a crime based on the fact that people driving under the influence are significantly more likely to harm others, and the baseline risk is high enough to make driving in such a state unacceptable. Even with laws in place banning it, alcohol is a leading cause of vehicle related deaths and there are quite a few of them every year.
It seem problematic to argue that these laws aren’t based on actual facts and real danger when both are well documented.
> The association with realized harm is probabilistic in nature
The same could be said about almost every kind of harm you can think of. It’s unclear what someone is supposed to take from this statement.
A cop or the law makes no pretension of trying to evaluate if a driver actually put someone at risk in a given scenario. A judge doesn't care if you were alone on the road or how far you were going. A judge doesn't even care if the car is on or moving.
>alcohol is a leading cause of vehicle related deaths and there are quite a few of them every year. It seem problematic to argue that these laws aren’t based on actual facts and real danger when both are well documented.
I would say there is little systemic logic to the rate and severity of penalties. In 33% of pedestrian fatalities, the pedestrian was drunk. In 16%, the driver was drunk.[1]
Similarly, the common DUI threshold is 0.08 BAC.
Being awake for 17 hours (working late) is similar to having a BAC of 0.05%. [2]
Being awake for 24 hours (working a double shift) is similar to having a BAC of 0.10%
Simply talking on a phone while driving has been shown to be similar to a BAC of 0.08%.
Texting while driving is equivalent to a BAC of 0.19%! [3]
The fine for texting while driving starts $20, but the total cost can be over $162 due to court costs.
>The same could be said about almost every kind of harm you can think of. It’s unclear what someone is supposed to take from this statement.
Im just making the point that we we criminalize isnt as clear cut as one might think, and people often conflate actual harm, actual risk, and heuristically approximated risk.
More generally, I think the intersection with probability and criminal justice is fascinating, particularly the criminalizing risk factors. e.g. if we took a similar approach, what percent confidence would we need that someone will commit murder to preemptively lock them up based on risk.
https://www.ghsa.org/sites/default/files/2020-02/GHSA-Pedest...
https://www.cdc.gov/niosh/work-hour-training-for-nurses/long... https://facilities.uw.edu/blog/posts/2016/07/26/texting-and-...
On average, drunk driving obviously poses more risk than regular driving. However, drunk driving isnt always risky, and does not necessarily or uniformly put others at risk at the factual case by case level.
Edit: I think this makes it an interesting intersection between probability, heuristics, and criminal justice. It highlights the difference between a categorical model of the world, and reality, i.e. Probability is an abstraction and property of the model, not the object, in a deterministic universe.
Otherwise I’m not sure how to resolve the contradiction between “drunk driving … poses more risk than regular driving” and “drunk driving isn’t always risky.”
You’re right to wonder how a system based on civil justifiability principles that include concretized injury also allows criminal penalty for risk without harm.
In my excerpt you quoted, I meant what I said, and added an edit that might help.
On average, drunk drivers are more risky than sober drivers. However an individual is not the average. The more information you have about an individual circumstance, the more it can diverge from the average of the model.
For example, for the average smoker, the lifetime risk of developing lung cancer is typically estimated to be around 10% to 20%.
If you know the smoker is 80 years old, the chance they will get lung cancer in their lifetime is much lower.
If that 80 year old smoker is in an airplane barreling towards a mountain, their chance of getting lung cancer in their lifetime is 0% and less than an average non-smoker.
The same is true for anything described by risk (like drunk driving). If you plug in enough details around the circumstance, the risk changes. Sometimes it goes up, sometimes it goes down.
>You’re right to wonder how a system based on civil justifiability principles that include concretized injury also allows criminal penalty for risk without harm.
Thats a big part of what I was thinking. Drunk drivers can cause tremendous damage, and make up a clear portion of vehicle accidents, but I think the actual chance of accident per individual drunk trip is quite low in terms of percent. Lets say for argument it is 5%, which I think is very conservative.
If we were to use all data available, I wonder how well something like future violent crime could be predicted. At what point is that a rationale for incarceration based on risk alone? For context, the Average recidivism rate for the category of violent instant felons (spontaneous violence) is over 60%. Average recidivism rate for groups with particularly serious criminal records (Criminal History Category VI ) are over 85%![1 pg 14], and that is without all of the invasive but theoretically available information like psychological reports, medical records, race, income, and family history, ect.
What are the implications here for the viability of criminalizing risk without harm. I know these two cases are not the same, but I am still processing the differences, to see what distinctions can be made. Surely we shouldn't incarcerate based on risk alone. I suppose that is why I am posting these provocative but good faith questions.
https://www.ussc.gov/sites/default/files/pdf/research-and-pu...
I'm struggling to see how you could truly analyze the additional risk of being > .08 as zero but 1) I expect you're about to make an example and 2) I reckon the example is out there.
A one foot journey on a flat road during the day in good weather seems a bit degenerate, no?
Your "at what point is that a rationale for incarceration based on risk alone" reminds me a little of banning fireworks. We're not banning fireworks because we hate fun, we're banning them because of the risk of a deleterious effect that is not possible without an explosive. Do we incarcerate for fireworks possession? We usually do not. This is a thing that is worth thinking about.
I guess you could say we're not imprisoning because of the risk; we're imprisoning because drunk drivers are unable to follow a very bright line rule. Why would I be imprisoned for driving 180 mph? Same reason.
Hell, being drunk just sitting in a parked, no ignition driver seat is illegal in some places.
Not sure how the second one is reckless in this context.
But those laws are controversial for exactly that deviation from the norm, so it can't be held up as an instance of a widely agreed upon law which is uncontroversial to emulate.
If anything, the consensus is we need harsher DUI laws and enforcement.
Also, in general any kind of "people support harsher X" where X is Think Of The Children-related will be true, but that's a heuristic for discovering bad laws that should be repealed because of the known effect that Think Of The Children rhetoric causes people to support bad laws.
Nobody expects the intersection to be zero - not the banks, not the regulators. The rule is that banks must take prudent measures, and the definition of that has evolved over time (as financial crime has evolved over time).
Google "risk-based anti-money laundering approach" or if you'd like a more thorough introduction, listen to the Dark Money Files podcast.
I have libertarian sympathies too, especially since most AML rules were originally established to fight the war on drugs (my sympathies lie with the dealers, sorry). But I'm pretty happy to make life hard for terrorists, bad governments, and kleptocrats - who, unlike the characters in Reefer Madness, are not just bogeymen.
The premise of DUIs is to punish a person acting recklessly. The premise of KYC is to punish a third party for not conducting mass surveillance on predominantly innocent people in case one of them commits a crime. It's hard to see how these are analogous.
> Practically every prosecution of money-laundering services included evidence of knowledge by the operator that their customers included people committing crimes.
This is quite a fig leaf for any entity operating at scale. The customers of any major bank will obviously include people committing crimes -- they have millions of customers. They also have thousands of employees, and then some percentage of them will be criminals as well. Expecting the intersection of these to be literally zero is an unreasonable standard.
More to the point, the underlying problem is what it incentivizes them to do, i.e. manufacture some opaque and arbitrary indicators and then conduct mass surveillance and punish innocent people for running afoul of secret rules.