Whether it's reasonable to use poison as a disincentive needs to depend on how often people consume it, accidentally or purposefully. If it's extremely rare, then this is a reasonable use case, even if we don't think poisoning is a reasonable punishment for tax avoidance (because, indeed, the poison is not intended to be a punishment).
Likewise, we put up physical barriers like metal posts and concrete walls to dissuade drivers from crossing certain lines even though it would obviously be outrageous to punish a driver for crossing a lane line by smashing their car. As long as drivers strike these very rarely, it's reasonable to use these as disincentives.
By all means, if we find disincentives (like bittering agents) that are just as effective with even less risk, we should use them. But claiming that poison is being used as a punishment for tax avoidance is unjustified.
Your situations are not comparable. It's impossible for a human to tell the difference between methanol and ethanol using our readily available senses. Once the methylated spirit is removed from its container there's nothing to warn someone the alcohol is poison that will probably intoxicate, then kill or blind them. Methylated sprits substantially increase the risk of death for no other advantage than to prevent tax evasion. It's the equivalent of deciding the punishment for tax evasion is to execute one in every hundred tax evaders.
Physical barriers on the other hand are both obvious to car drivers, there's no situation where a car driver believes a physical barrier might be safe to drive into, and more importantly, physical barriers are used to protect other road users, not punish people. So you're trading an increased risk of death to prevent bad driving, you're increasing the risk of driver death to reduce deaths caused by bad driving. If drivers regularly strike a barrier that protects a school, it would be completely unreasonable to remove that barrier, and allow drivers to strike children instead.
Even more so, counterfeit drinks contaminated with methanol have killed people who accidentally bought and unknowingly consumed said drinks, even in bars! Becherovka incidents come to mind.
so you're not killing tax evaders, but rather random innocents.
> During summertime, the patient had earned his living by fire eating at different Spanish locations.
> According to the patient, a sudden episode of hiccough during fire eating caused accidental ingestion of denatured alcohol containing methanol.
> Our patient demonstrates that accidental ingestion of even small amounts of denatured alcohol containing methanol can cause irreversible blindness with intracerebral lesions.
> As a clear, colourless, volatile liquid with a weak odour, methanol is difficult to differentiate from other forms of alcohols such as ethanol.
> Methanol is rapidly absorbed not only after oral ingestion but by inhalation or after cutaneous exposure and becomes oxidised in the liver to formaldehyde and to formic acid, metabolites which are more toxic than methanol itself
This highlights the fact that the average person has no idea how seriously methanol should be taken. I started distilling my own moonshine / hand sanitizer at the beginning of the pandemic, and before that I really hadn't heard the horror stories and had no idea how common accidental poisoning was.
Horrifically, the US government actually increased methanol in industrial spirits during an prohibition with what seems to have been clear intent that more people would die as a result[1].
> It's impossible for a human to tell the difference...
You are arguing that the risk of accidental poisoning by methanol is in fact high. Thats an empirical question, and I explicitly agreed in my comment that if the risk is high then this is not a good method of disincentive. Importantly, your argument is not the one I was rebutting, that poisons are bad disincentives because they are inhumane punishments for tax evasion.
But more to your point: methanol poisoning is often used in situations where the potential drinkers are highly informed, e.g., people ordering chemicals out of a catalog. The risk, ultimately, is an empirical question, and you are invited to check how often accidental poisonings of this type happen.
> It's the equivalent of deciding the punishment for tax evasion is to execute one in every hundred tax evaders.
It's not, both in terms of numbers and in terms of principle.
> physical barriers are used to protect other road user...
You have misinterpreted my comment. I did not refer to protective barriers. There are plenty of examples where metal posts and concrete barriers are used to stop cars from getting to places where they would be a nuisance but not dangerous (e.g., pay parking lots, dedicated express lanes, and "no thru traffic" barriers).
> You are arguing that the risk of accidental poisoning by methanol is in fact high.
I’m arguing that no level of introduced “deterrent” fatality risk is an acceptable trade off for preventing a crime that itself isn’t punishable by death (and I believe that no crime should be punishable by death).
The only situation I would consider deliberately introducing the risk of fatality as a deterrent, is when there’s extremely clear evidence that such a deterrent lowers the overall fatality risk associated with the crime, by reducing harm to victims, or to the perpetrators themselves.
> methanol poisoning is often used in situations where the potential drinkers are highly informed, e.g., people ordering chemicals out of a catalog.
It’s also used in plenty of situations where potential drinks are completely uninformed. Such as when bought for the purpose of paint thinning, camping stoves, and as a cleaning product. Plenty of situations where a completely uniformed individual may purchase the denatured alcohol, or have easy access to denatured alcohol.
> Importantly, your argument is not the one I was rebutting, that poisons are bad disincentives because they are inhumane punishments for tax evasion.
I can provide a pretty simple rebuttal to idea that poisons are good disincentive to prevent tax evasion. If you’re not educated on the process of methylating spirits to make alcohol undrinkable, then by the time you’ve drunk enough of it to figure it out, you’re dead.
Now in a an extremely pragmatic way, that makes poisons extremely effective at preventing tax evasion. Those educated enough to know better don’t try to evade the tax, those that do die. This all evaders are either discouraged or eliminated. Of course one issue with this is assuming the person consuming the substance is the same person that evaded the tax. There have been plenty of cases of people diluting normal alcohol with methylated spirits to bulk up their stock, in the process making the assumption the dilution is great enough to avoid poisoning people, or that they won’t be around to face the consequences.
I think as a society we’ve agreed that this sort of absolutist approach to law and order is pretty barbaric and unbecoming of a more civilised society.
> It's the equivalent of deciding the punishment for tax evasion is to execute one in every hundred tax evaders.
> It's not, both in terms of numbers and in terms of principle.
You’re right. It’s closer to killing a random small proportion of tax evaders, and also killing a random small proportion of anyone who’s had the misfortune of doing business with them.
> You have misinterpreted my comment. I did not refer to protective barriers. There are plenty of examples where metal posts and concrete barriers are used to stop cars from getting to places where they would be a nuisance but not dangerous (e.g., pay parking lots, dedicated express lanes, and "no thru traffic" barriers).
Again, it’s obvious to anybody what a physical barrier will do to their car, protective or otherwise. It’s pretty hard to conceal a physical barrier in a manner that would cause someone to unknowingly drive into it. Unlike methylated sprits where’s it trivial to conceal its nature.
> I’m arguing that no level of introduced “deterrent” fatality risk is an acceptable trade off for preventing a crime that itself isn’t punishable by death
All deterrents have a non-zero chance of causing death. People die in prisons, catch fatal viruses in court, etc. Indeed, even bitterants can cause allergic reactions, and so definitely have a non-zero probability of causing death.
> People die in prisons, catch fatal viruses in court, etc.
Prisons and courts aren’t designed to kill. Methylated spirits are. This is very simple concept you seem to be struggling with, governments shouldn’t seek to kill their citizens, or introduce deterrents that operate entirely on the principle of making people fear for their lives if they fail to comply with the law.
A bitterent that can cause allergic reactions (and most substances can!) is on a completely different scale to something that is an actual poison and consuming a small amount can cause permanent damage/kill you, regardless of your immune system's sensitivity.
> If it's extremely rare, then this is a reasonable use case
I don’t really see the point of risking poisoning people as a way of dissuading tax dodging especially when alternatives exist. In the same way we don’t put solid barriers on roads unless people walking need to be protected - when a barrier is there to dissuade a driver it’s designed so a car runs through a barrier without hurting the driver and passengers.
> I don’t really see the point of risking poisoning people as a way of dissuading tax dodging especially when alternatives exist.
You're not disagreeing with what I wrote.
> In the same way we don’t put solid barriers on roads unless people walking need to be protected
Not true. We also put up barriers when people driving over the line would merely be a nuisance, e.g. in parking lots or to keep people from entering closed areas.
Those nuisance barriers you mention are often intended to cause limited damage. For example, water barrels in a gore point or the gates of a parking garage which are made out of wood and intended to snap off.
The problem with the first sentence of your initial post is that you seem willing to justify some degree of death if it serves as a deterrent, even though you later claim that if other options are equally good deterrents then they should be used. This leaves open the possibility that if bittering agents don't deter as well, poison may be a reasonable choice. Yet the only thing we're talking about deterring is tax avoidance. In which case poison seems completely unreasonable. This leads one to suspect that you believe there are other public goods brought about by deterrence which would equal the weight of an occasional poisoning.
> Those nuisance barriers you mention are often intended to cause limited damage. For example, water barrels in a gore point or the gates of a parking garage which are made out of wood and intended to snap off.
Some are, and they may indeed be preferable to ones that cause worse damage, but some are not. The correct choice depends on the risks involved and the costs of the various methods (e.g., concrete requires less maintenance than wood).
> This leaves open the possibility that if bittering agents don't deter as well, poison may be a reasonable choice. Yet the only thing we're talking about deterring is tax avoidance. In which case poison seems completely unreasonable.
This is a good line of inquiry, but your blanket conclusion is wrong. Again it depends on the details.
If bitterants are 90% effective but methanol is 99.9% effective, corresponding to $100M vs. $100k worth of alcohol in the black market, then this must be weighed against the statistical value of life, which is ~$8M in the US, and the number of people killed by the poisoning.
The poison isn't a punishment because it isn't being drunk by people who we intend to have drink it (which is no one). The poison is being accidentally drunk by people who don't know it's there, which is exactly the opposite of what's intended. That's what makes those poisonings accidental side effects, rather than penalties.
In contrast, "man traps" (booby traps intended to harm or capture trespassers) are generally illegal because their effect (bodily injury and/or, effectively, kidnapping) are illegal because they unjustly inflict consequences on their intended targets.
> this must be weighed against the statistical value of life
Must it be? There are a great many regulations concerning safety devices and procedures that government and industry shell out for which cost more to industry than the statistical value of lives saved. Safety compliance is an added cost to virtually every aspect of commerce. You seem to suggest that any cost of compliance beyond the actuarial value of lives taken by noncompliance is unwarranted. But in this case we're going much further, since the addition of methanol is not merely a lax safety regulation but an added cost to manufacturing which serves no purpose other than to make the product less safe.
Consider safety caps on pill bottles, or seat belts in cars. Both those things had to be forced upon manufacturers, because the cost to manufacturers of fighting civil litigation - which we can assume both operates upon and inflates your $8M figure - didn't provide sufficient economic downside to spur costly implementation. Their existences are both examples of life being valued by society as more than its actuarial value, or value provable in court.
If, as a libertarian would, you argued that companies should never pay for any safety regulations beyond the dollar value of the internal and (maybe) external costs inflicted on the public by their products, implying that the value of human life should be reduced to a dollar figure, even then it would make no sense to pay more just to make a product more dangerous.
I feel like you mean your statement to be objectionable. And I feel like we're getting to the root of why your statements in this thread have generated such a bad feeling; your disdain for life is coming out. We could get into a discussion of the actual value of human life versus its average value in economic output - including the externalized costs manifested in social disorder and rebellion when life is snuffed out by neglect or malpractice. But again, this is why I think the motive for adding poison to products is suspect, and can't be attributed only to economic motives. It implies a hatred of humanity. Indeed, the value proposition for many governments to add methanol to industrial alcohol isn't based at all on actuarial tables, but on "moral" judgments. In Iran, the lacing of all industrial alcohol with methanol is part and parcel of the religious ban on personal consumption of ethanol, and it led to more deaths during the pandemic than the virus itself. So perhaps there, it could be justified economically by saying that the life of someone who indulges in drink is worth a negative actuarial value; but this is economics shaded by fanatical belief, and must not be taken as a proxy for the true value of life.
> If it's extremely rare, then this is a reasonable use case
You never actually gave a reason for that assertion.
> Whether it's reasonable to use poison as a disincentive ...
> (because, indeed, the poison is not intended to be a punishment)
What is the difference between "disincentive" and "punishment" here?
> "Likewise, we put up physical barriers*
So, would you be OK with replacing the physical barriers with, say, spikes that would tear up a car's tires and instantly cause an accident, should the car try to cross the line? Just to make the disincentive more effective?
> You never actually gave a reason for that assertion.
As discussed elsewhere in this thread, all interventions carry some risk of bad unintended things happening. You can get killed in car accident driving to court, but we don't eliminate court.
> What is the difference between "disincentive" and "punishment" here?
Traditionally, punishment is seen as having up to three components: deterrence, retribution, and/or incapacitation.
I'm using "disincentive" here to emphasize that (1) only deterrence is sought, (2) the deterrent is created before the deterred action rather than implemented later, (3) the deterrent is "bound up" with the deterred action in a particularly tight way. There is more to say about (3), but it would take a while to unpack, and I think (1) and (2) are sufficient for this discussion.
(I'm happy to use better or more standard terminology if it exists.)
> would you be OK with replacing the physical barriers with, say, spikes that would tear up a car's tires and instantly cause an accident, should the car try to cross the line?
Only insofar as the effectiveness went up so much that the total harm caused was less.
The spikes sound scary and bring a grisly accident to mind, and intuitively we know that in the real world drivers would sometimes hit them by accident. So let me try an alternative hypothetical that may help illustrate my position better.
Suppose we are trying to build a humane prison and we need to find a way to keep people inside. First note that, traditionally, a fence with razor wire on top is used as a disincentive even though we all agree it would be terrible to punish escapees by cutting their hands with a razor. And say there is some rate of prisoners escaping (1 per decade, or whatever) with the fence. Now, imagine we could replace the fence with an extremely deep pit around the prison that would be fatal to fall into. (For the sake of the thought experiment, we assume counterfactually that the pit is as cheap as a fence and introduced no other issues.) Suppose the pit has a railing around it so no one falls in accidentally, and also that it is much more effective at preventing escape because essentially no one bothers to try.
Death by falling is even more unreasonable as a punishment for trying to escape, yet the situation seems better than the razor wire. The reason is that the pit is a very effective disincentive (a deterrent created prior rather than after the fact) and the total harm (number of escapees plus injury to attempted escapees) is much less.
Likewise, we put up physical barriers like metal posts and concrete walls to dissuade drivers from crossing certain lines even though it would obviously be outrageous to punish a driver for crossing a lane line by smashing their car. As long as drivers strike these very rarely, it's reasonable to use these as disincentives.
By all means, if we find disincentives (like bittering agents) that are just as effective with even less risk, we should use them. But claiming that poison is being used as a punishment for tax avoidance is unjustified.