The study starts with the questionable premise that broadband deployment is associated with digital payments and ATMs per capita are associated with cash payments, but ecommerce sites are some of the least bandwidth-intensive and fully usable on ancient networks, and ATMs per capita is going to have all kinds of arbitrary confounders like population density, average income or number of local banking institutions that make it a poor proxy for cash use.
They then proceed to exclude the types of tax fraud that don't use cash. Of course, this invalidates the result: If Alice and Bob are both cheating on their taxes and Alice accepts credit cards and does the cheating through cross-border transfer pricing and Bob does it through not reporting cash payments then you're going to find that red cars are associated with crashes involving red cars, but that doesn't even imply that most crashes involve red cars, nor does it imply that you can reduce overall crashes by banning red cars.
Then it gets even worse. They try to compare it to data on tax evasion, but such data is by definition speculation because tax evaders don't disclose that they're doing it and you only know about the ones that get caught. So then you have a study implying that the tax evaders who use cash are more likely to get caught, which leads to a different conclusion than the one they're advocating.
> Your mechanic example essentially requires a very specific business scenario to execute.
Not really. The general business model of "you hire someone for a job that results in scrap and they dispose of the scrap" is extremely common in everything from equipment repairs to construction to trash hauling.
> The landlord example is also one that requires a very specific business scenario.
It isn't required that every type of business use the same method; various businesses would use a method tailored to their business.
> E.g., I run a video game studio that is spending the next 3 years working on a game, how is that landlord going to pull off that scam?
E.g. landlord gets shares in the company at a discount in exchange for the owner getting a discount on their personal rent.
> In your example the landlord eats at the restaurant every day, but they can't really barter in this way with all their tenants.
They don't have to barter with all their tenants. Unless you're talking about a huge skyscraper, the rent on one unit will be of approximately the right magnitude to erase the net profit on the whole building.
> The main point is that you with cash you can commit the fraud in the first place, not that you're going to necessarily survive an audit. The IRS from what I can tell audits a fraction of a percent of taxpayers every year and dedictate most of their resources to higher earners. The IRS going to your establishment requires that they dedicate resources to you in the first place.
That's the same as any other type of tax fraud. If you make up and deduct entirely fictional business expenses, someone has to ask for the receipts before you get caught.
> But there's a really simple mechanism for defeating this cash bar audit example: only commit tax fraud with transactions from the regulars. The IRS agent walks into the bar, is a stranger, and by default their transaction is one of the legit ones. You only run the scam with customers you recognize coming in multiple times over a span of months or years.
By which point you've much limited the number of transactions that could go unreported, and you still have the possibility to get caught if you have a dispute with any of the regulars and they turn you in, or the government puts in the work to do a long-term investigation.
Also, how is this anything specifically to do with cash? If the landlord is charging lower rent because they get to eat in the restaurant, how is that any easier for the IRS to detect unless they turn on each other?
> however, a business committing social security tax fraud with paid wages can do so much more easily using cash, so really the argument is similar – in that case it just involving checks and direct deposit instead of cash.
Except that employee wages are business expenses and they'd be losing more by failing to claim the deduction than they'd be paying in FICA.
> or exchange a mix of goods and money for labor.
Which is exactly what happens with electronic payments. The employee is nominally paid $1000 but is actually paid $1000 and an undisclosed amount of goods, with only the $1000 showing up in the bank transfer.
> And this idea that a large amount of W2 employees are using company assets for personal use is pretty silly, most probably don't even have any assets from the company.
Historically it was extremely common to have things like company cars and comped meals. The current rules generally now require that stuff to be counted as employee income and taxed, so it has largely disappeared from large corporate employers that don't openly engage in tax fraud, but you can guess what happens in a lot of small businesses where the family has a vehicle registered to the business and deducted as a business expense etc., or places where employees have access to company vehicles or other equipment and managers DGAF.
The study starts with the questionable premise that broadband deployment is associated with digital payments and ATMs per capita are associated with cash payments, but ecommerce sites are some of the least bandwidth-intensive and fully usable on ancient networks, and ATMs per capita is going to have all kinds of arbitrary confounders like population density, average income or number of local banking institutions that make it a poor proxy for cash use.
They then proceed to exclude the types of tax fraud that don't use cash. Of course, this invalidates the result: If Alice and Bob are both cheating on their taxes and Alice accepts credit cards and does the cheating through cross-border transfer pricing and Bob does it through not reporting cash payments then you're going to find that red cars are associated with crashes involving red cars, but that doesn't even imply that most crashes involve red cars, nor does it imply that you can reduce overall crashes by banning red cars.
Then it gets even worse. They try to compare it to data on tax evasion, but such data is by definition speculation because tax evaders don't disclose that they're doing it and you only know about the ones that get caught. So then you have a study implying that the tax evaders who use cash are more likely to get caught, which leads to a different conclusion than the one they're advocating.
> Your mechanic example essentially requires a very specific business scenario to execute.
Not really. The general business model of "you hire someone for a job that results in scrap and they dispose of the scrap" is extremely common in everything from equipment repairs to construction to trash hauling.
> The landlord example is also one that requires a very specific business scenario.
It isn't required that every type of business use the same method; various businesses would use a method tailored to their business.
> E.g., I run a video game studio that is spending the next 3 years working on a game, how is that landlord going to pull off that scam?
E.g. landlord gets shares in the company at a discount in exchange for the owner getting a discount on their personal rent.
> In your example the landlord eats at the restaurant every day, but they can't really barter in this way with all their tenants.
They don't have to barter with all their tenants. Unless you're talking about a huge skyscraper, the rent on one unit will be of approximately the right magnitude to erase the net profit on the whole building.
> The main point is that you with cash you can commit the fraud in the first place, not that you're going to necessarily survive an audit. The IRS from what I can tell audits a fraction of a percent of taxpayers every year and dedictate most of their resources to higher earners. The IRS going to your establishment requires that they dedicate resources to you in the first place.
That's the same as any other type of tax fraud. If you make up and deduct entirely fictional business expenses, someone has to ask for the receipts before you get caught.
> But there's a really simple mechanism for defeating this cash bar audit example: only commit tax fraud with transactions from the regulars. The IRS agent walks into the bar, is a stranger, and by default their transaction is one of the legit ones. You only run the scam with customers you recognize coming in multiple times over a span of months or years.
By which point you've much limited the number of transactions that could go unreported, and you still have the possibility to get caught if you have a dispute with any of the regulars and they turn you in, or the government puts in the work to do a long-term investigation.
Also, how is this anything specifically to do with cash? If the landlord is charging lower rent because they get to eat in the restaurant, how is that any easier for the IRS to detect unless they turn on each other?
> however, a business committing social security tax fraud with paid wages can do so much more easily using cash, so really the argument is similar – in that case it just involving checks and direct deposit instead of cash.
Except that employee wages are business expenses and they'd be losing more by failing to claim the deduction than they'd be paying in FICA.
> or exchange a mix of goods and money for labor.
Which is exactly what happens with electronic payments. The employee is nominally paid $1000 but is actually paid $1000 and an undisclosed amount of goods, with only the $1000 showing up in the bank transfer.
> And this idea that a large amount of W2 employees are using company assets for personal use is pretty silly, most probably don't even have any assets from the company.
Historically it was extremely common to have things like company cars and comped meals. The current rules generally now require that stuff to be counted as employee income and taxed, so it has largely disappeared from large corporate employers that don't openly engage in tax fraud, but you can guess what happens in a lot of small businesses where the family has a vehicle registered to the business and deducted as a business expense etc., or places where employees have access to company vehicles or other equipment and managers DGAF.