Thank you for the input, I’m an engineer and phd student in engineering so I based a lot of this on intuition, there’s an order gap for me to write something traditionally axiomatic and rigorous, but my understanding is that algebra is based on logic and problems in general, and I would like to know any logical fallacies at the very least.
I feel like this starts with an agreeable premise. Some fraud is egregious, costly, and/or easy to detect. These low-hanging or high-impact cases are most worth pursuing. At some point you reach diminishing returns, where the amount of time / effort / capital you're putting in to eliminating fraud outstrips the losses from the fraud itself.
I don't know that I agree with the ethical conclusion that the optimal amount of fraud is therefore non-zero. The leap from "anti-fraud efforts are expensive" to these sentences in the final paragraph was not, in my opinion, convincingly made here:
>We should, as a society, accept non-zero amounts of benefits fraud. We should accept non-zero amounts of cheating on taxes.
I don’t know if that statement is backed by the article, which I will admit to not having read, but in general I agree. Completely eradicating benefit fraud will necessarily increase the burden on legitimate claimants to prove that they are in fact legitimate. Doing that is going to place enough burden on some people who should otherwise be able to claim that it results in them not doing so, or failing to do so because they were unable to provide the required evidence.
I’d much rather see a few people who didn’t need benefits manage to claim them than see people who do need them be left without. The first option costs tax payers a bit more money. The second results in people’s lives being made significantly worse, and in some cases in deaths.
Also, not means testing universal benefits means everybody appreciates them as just something their society does, so that reduces stigma for the beneficiaries and increases pride in your society. "We ensure children in this country have nutritious food" not "Why are my taxes going to feed this 10 year old whose mother has a full time job".
I grew up in an area where many parents could afford (maybe if they budget carefully, maybe just anyway) to privately educate a child. But they mostly didn't, because the government funded schools were pretty good. In fact, as children it was actually a minor stigma to be privately educated, because if your parents are spending a lot of money on the fancy school, either they don't know how to spend their cash (so they're stupid) or you're really stupid and they sent you to that school in the hope of making up for it. It was seen as like easy mode. Smart kids don't go to private school, why would they waste the money?
> The first option costs tax payers a bit more money.
The first option costs taxpayers significantly more than a ‘bit’.
Just look at how much it cost when they basically turned off all the checks in order to get covid relief into the hands of people who really needed it. In Arizona, after a while, they made it so you had to sit on (virtual) hold for 8-10 hours to verify your identity with a human or they would cut you off. Which worked well enough to ensure only the people who really needed it went through all the hassle. It really sucked for those people but they stopped sending billions of dollars overseas to people who just googled someone’s address.
> At some point you reach diminishing returns, where the amount of time / effort / capital you're putting in to eliminating fraud outstrips the losses from the fraud itself.
That's not quite what I got from the article. I read it as the more friction you put in place to prevent fraud, the harder it is for legitimate transactions to happen. Therefore, it's not so much about the cost of the fraud, but the opportunity cost of legitimate transactions which don't happen in the zero-fraud environment.
I appreciate how you phrased this. It has me thinking about how it might be similar for privacy and security in terms of information or even physical security. Yes, one can be super secure and safe from harm if one puts tons of locks on everything, but it also keeps out people who we might want to let in.
Actually, now I'm thinking about it emotionally as well. Best way to prevent myself from getting hurt is to close off as much as I can. Also the best way to prevent myself from feeling joy and all the other things I want to feel.
Emotions are a little different though. The more you know how deep the lows can go, the more you appreciate even the smallest highs. There is a little utility in getting hurt.
Good point. I agree with the overall thesis; there are a lot of things that get increasingly expensive as you approach perfection. (Perfection is still a useful guidestar, but each step toward it has to be made with costs in mind.)
However, I'm not nearly as breezy about $20 billion annually in fraud. Maybe that's fine from the perspective of the merchants and credit card networks. But from the societal perspective, that's subsidizing bad actors. People and groups who will not stop at one kind of crime as they try to grow. People who will divert other people into being parasitic. That's not healthy for society or for the individuals who end up living lives of crime.
So I think the society-optimal level of fraud is way below the merchant-acceptable amount of fraud.
The problem is not merely that the anti-fraud efforts are costly but that the anti-fraud surveillance apparatus will itself be value destroying. (In the tax case, it’s “people in democracies don’t enjoy their government having total visibility into their activities and society, in its judgment, says this is more important than tax collection at some margins.”)
This is not an ethical conclusion. This is a pragmatic and utilitarian conclusion where 'optimal' means minimising the cost/benefit ratio.
Incidentally, this shows that the 'perfect' ethical stance is not necessarily the one that delivers the most benefits at the least cost, aka when ideals meet the real world...
Yes, we should not accept the existence of fraud. We should simply be able to recognize the situations where fighting fraud is more costly than letting it exist.
Not that it really matters in most places since we are quite far from that point anyways.
It feels like a very subtle is-ought distinction, where the author is discussing something that unavoidably is the case and therefore concludes that it ought to be the case and therefore ought to be accepted if not even welcomed. The marketing example makes this pretty clear. Of course no one thinks the marketing directory could spend zero on marketing. But…surely they would love to spend zero if they could still get what they wanted for zero money.
not if letting 10 guilty esacpe breeds more and more bad actors. you can make the argument that a few innocent suffer is a net benefit for society in tne same way. in pursuit of 0 innocent suffering you will capture no bad actors
To put it another way you're forgetting the victims. The 10 fraudsters made 10 people suffer. their suffering needs to added to the equation
The current system from that perspective works great. You have to get pretty unlucky to get convicted as an innocent, especially in this day and age where a cop’s testimony isn’t worth nowhere near as much as it used to be.
And morally alone it’s far worse to wrongfully convict an innocent person than to let a guilty one go free.
Not an ethical conclusion but a pragmatic one. The ethical part is what you do after the fact:
1. Pass the cost towards self regulation of people, using client facing measures e.g. prove their innocence if they are an outlier
2. Catch a couple of cases and over market your policing ability to disadvantage the most gullible.
3. Catch a couple of cases, even minor infractions and destroy them with disproportionate fines or jail sentences, economy of randomness or economy of those who have the best lawyers.
Fraud against government, as above but add:
4. Add arbitrary constraints, you don't really want the system to work, you just fake it for political reasons
The ethical issues of accepting nonzero fraud are that striving for zero fraud creates program design changes that lock people out of benefits. If you design a health care system that aims for 0% fraud, some measurable number of people are going to be deprived of care because the registration and billing procedures are too onerous. With taxes, aiming for 0% noncompliance will prevent people from taking advantage of deductions and credits.
This isn't hypothetical; it's the issue underling the "program design" controversies about means-testing in public policy.
Not to mention that enforcement has rapidly diminishing returns. Even if your only goal was to maximize tax revenue (minus cost of tax administration), and you didn't care at all about people being able to take advantage of deductions, the optimal amount of fraud is almost certainly non-zero.
(And of course, if you did want to maximize tax revenue, you'd focus enforcement on the big fish.)
I would think the strategy would be to encourage low impact fraud with lazy compliance and making a customer whole (Credit card chargebacks). And then hunt out and destroy high impact fraud.
With the intent to incentivize and train criminals to stay small and low impact.
If you're a retail platform, and you have a few scammers making a few grand of 20-100 dollar scams. You can play wack a mole with them and then that keeps people doing that small fraud rather than leveling up and potentially doing crimes that could endanger the whole business with the exposure.
> I don't know that I agree with the ethical conclusion that the optimal amount of fraud is therefore non-zero. The leap from "anti-fraud efforts are expensive" to these sentences in the final paragraph was not, in my opinion, convincingly made here
It’s like saying that the optimal dirtiness after cleaning your house is non-zero (greater than zero) because cleaning it perfectly takes much more effort than it is worth!
That’s not counter-intuitive at all. It’s just an obvious fact stated in a silly way (for clicks or whatever else).
That doesn't mean that the optimal amount is nonzero. Taken in isolation, the optimal amount is clearly zero. The optimal amount doesn't change based on the cost, the optimal amount of effort to expend is a different answer.
It's not just stated in a silly way, it's stated in a way that's incorrect because they didn't mean what they said. "The optimal amount of fraud is nonzero" does not actually mean the exact same thing as "in an optimally-beneficial fraud prevention effort, the amount of fraud is non-zero".
>Taken in isolation, the optimal amount is clearly zero.
But the very point of the article is to not take zero-fraud in isolation and instead, explain how non-zero-fraud is an unavoidable tradeoff when balancing 2 simultaneous goals:
- (1) prevent fraud transactions as much as practically possible
- (2) make legitimate transactions as easy as possible
If one accepts the premise of pursuing those 2 goals at the same time, then by definition, we're no longer talking about "in isolation". You've now unavoidably entered non-zero fraud territory.
Perhaps it's the author's particular wordsmithing of what he's trying to convey that just rubs many readers the wrong way.
> Taken in isolation, the optimal amount is clearly zero.
The post makes it clear that the discussion is not about theory or taking anything in isolation - it's about fraud in the real world. In that context, the way it's stated is correct - if you have zero fraud in the real world, that means that you designed the tradeoffs wrong and that the cost of your fraud prevention (in terms of actual dollars as well as inconvenience to customers, etc.) is greater than the overall cost would be if you allowed a small amount of fraud to occur (looking at the total cost of that fraud as well as the cost of preventing additional fraud).
I suppose the problem is that whether or not the title of the post is true or not depends on the context in which it's taken, and the title itself doesn't have any context. Since the post does offer context, though, I think it's reasonable to take the title in that context.
It's like cleaning old painted metal with a scouring pad; You want to clean thoroughly enough to take off the grime, but if you scour too long or too hard you'll end up taking off the paint itself. You'll always either leave a littke dirt behind, or take off some paint, never perfection. You could strip all the paint and repaint it, but that's so much more costly in terms of time and materials that it's a whole different task.
And the argument that more stringent anti-fraud protections increase the burdon on legitimate claimants is absolutely spot on, and has parallels in all sorts of other legal, financial, and market situations c:
Targeting zero is an immature approach that is self-destructive in most cases.
If your incentive is to have zero fraud, the organization will find ways to not detect fraud or add so many controls and audits that the cost of doing whatever will go up.
There’s a balance. In the tax world, the de-clawing of the IRS for certain things have dramatically impacted compliance. You want enough enforcement that you’re discouraging median cheater, but not so much the cure is more expensive.
We can start with payment. What would someone pay with? Credit/debit numbers can be stolen. Checks can be stolen or forged. Cash can be counterfeit. What form of transaction has zero chance of fraud?
To make transactions available to people you need to introduce systems that can have fraud in them. There is a balance between availability/ease and fraud.
Bitcoin. It can easily be confirmed as valid (zero chance of counterfeit), and is otherwise a bearer instrument with no further settlement, and impossible to reverse (like cash).
The problem with accepting it is that people figure out repeatable tricks to get around the system.
If we view those repeated tricks as business as usual - we should probably make them accessible to everyone. Otherwise the small fraud becomes rampant.
I'm not quite sure how to phrase this, but here goes: during this procedure did you experience anything that you feel you could not have experienced absent someone physically poking your brain?
In short, no.
The operation itself lasted for over 12hrs, but I am only aware of (approximately) the first hour. So weird shit may have happened, but I have no memory of it.
Mostly because I had severe head trama! =) lol
At some point they put me out, so that messes with your memory too, but I think `brain surgery` is a good reason to not being able to remember.
If you get into a car accident and hit your head or get a concussion you don't remember everything either.
reply