How much crime it catches is directly related to how much crime it could prevent. If the measure is totally ineffective and criminals hardly ever get caught then there is no deterrent.
First, there are widely known techniques to beat AML. Criminals use them. Using them imposes costs. So you can reduce the profitability of crime, and therefore its frequency, without catching much crime directly.
Secondly, one of the goals of AML regulations is to make it easy to construct a money laundering case against a criminal caught another way. This is kind of like putting Al Capone in jail for tax evasion. Tax evasion isn't why you want him in jail, it is just the thing you can convict him of. Prosecutors see value in these easy convictions.
Are they worthwhile? That's above my paygrade. Certainly patio11 makes a case that they might not be. But both of the things that I just mentioned show that the rules are valued for reasons other than routinely catching a lot of crime.
> First, there are widely known techniques to beat AML. Criminals use them. Using them imposes costs. So you can reduce the profitability of crime, and therefore its frequency, without catching much crime directly.
This is only reasonable if the same measures don't also impose costs on innocent people, which is not the case.
And the value of doing this operates inversely with value: The crime you deter this way is the lowest value crime which is the easiest to deter through some other means, but the innocent people you most harm are the ones already at the margin who you don't want to deter/bankrupt, but you do.
Which damages the most competitive markets with businesses that were operating with the thinnest margins and forces them to consolidate into something that can absorb the compliance cost. The cost of inducing that kind of market consolidation is enormous -- as we've seen time and again.
> Secondly, one of the goals of AML regulations is to make it easy to construct a money laundering case against a criminal caught another way. This is kind of like putting Al Capone in jail for tax evasion. Tax evasion isn't why you want him in jail, it is just the thing you can convict him of. Prosecutors see value in these easy convictions.
Undoubtedly the lobby in favor of AML laws is lazy prosecutors who can't be bothered to prove their case honestly and would rather have a vague law that causes common behavior to be a chargeable violation. But that purposeful subversion of the rule of law isn't a legitimate reason even if it wasn't imposing major costs on innocent people.
Thing is, if you’re laundering money, you’re probably doing something incredibly profitable (importing kilos of something that cost you $2k and then selling for $20k, or just straight up stolen funds with no cost basis).
These groups have no issue with the few percent cost of added transaction friction.
But it really costs people running low-margin businesses or unsophisticated honest people just trying to move overseas, remit funds to family or buy a home overseas.
And the knock-on effects of the bypasses becoming things like “buy a front-business and care less about the legitimate competitors that can’t compete with someone unworried about profit” or “buy a house and let it sit empty”.
> How much crime it catches is directly related to how much crime it could prevent.
If you could actually prove that, I think you'd become quite famous in sociological/economic science.
So no, I think you're making an unwarranted assumption: "How much you can catch" is not necessarily proportional to "How many would have tried except for the fear of being caught."
In particular, actual/would-be violators have a distribution of different motivations and tolerances for risk.
Proportionality doesn't have to be linear, and in this case it's unlikely to be, but in the way that makes the AML rules even more worthless.
You might deter the large majority of crime if the chance of getting caught is one in three, because the cost of getting caught is also high. But if the chance of getting caught is only 0.2%, from a psychological perspective people are much more likely to see that and discount the possibility of it happening at all, and mathematically it would be unreasonable to impose a penalty high enough to compensate for such an abysmal rate of effectiveness. So ineffective rules aren't just useless, they're disproportionately useless.
This is clearly not true. Why else do organizations involved in large scale ML move to locations with more lax AML regimes? E.g. crypto in eastern Europe. If the risk of being caught was no deterrent, ML activity would be dependent on 'amount of launderable funds' and independent of AML regime - can you offer an argument showing this is true?
Ah the crypto boogeyman! That argument is ridiculous. The CIA factbook estimates that the proportion of the GDP linked to crime worldwide is 3% to 5%: that's 3 to 5 TRILLION USD directly linked to criminal activities.
Cryptocurrencies do not even register here. It's not even a drop in the bucket.
Moreover public ledgers are what law enforcement and IRS' employeees' wet dreams are made of.
One of the reasons it's so ineffective is that you can move to jurisdictions with different laws (or corrupt governments), so the law only negatively impacts innocent people in your own country, meanwhile criminals have stable alternatives.
The absolute magnitude of crime it catches is completely irrelevant. The earlier poster is incorrect saying is useless because
> It recovers 100x less than its compliance cost.
The proportion of crime that it catches is of interest to criminals and is what will have the deterrence effect, but whether it catches 10% or 90% is difficult to know as a layperson.
It's not actually that difficult to know, you can look it up. It isn't 90% effective, or 10%, but rather 0.2% effective.
The other major problem here is that what you're deterring isn't the underlying crime that generates the money but rather the activity that triggers AML scrutiny. So you have little hope of deterring the underlying crime, all you do is cause people to organize their finances in a different way. Which is hardly worth imposing significant costs on millions of innocent people.
It's "0.2% effective" by measuring in an arbitrary and frankly unrelated way. That figure comes from "what number do we get if we take the amount of illegal funds caught, and divide by the amount of money spent by other companies complying with the regulations?"
I'm sure you'll agree that this doesn't actually relate in any way to the specific number discussed above, aside from sharing the adjective "effective" used as a descriptor.
The cost of compliance to legitimate parties does not have any obvious mechanism to directly affect the deterrence effect.
