Foundational Principles and Derivatives

One question that I've been mulling over since posting this is the issue of deadweight loss. A simple example is that I have a product over which I have a monopoly, allowing me to cheaply create a widget that can massively improve human welfare. I could sell it cheaply enough for a million people to buy it and benefit from it, or I could set the price so high that only a few rich people can afford it. But in the latter case, they'll pay so much money that I'd be better off. But this causes a net welfare loss.

Mulling over this, it feels like it actually behaves like a positive externality. That is, if we were to subsidize the widget producer to make it slightly more profitable to sell to the masses, then we'd increase net utility. The issue then is simply whether that money could be more effectively spent on some other positive externality, or by direct transfer to people.

This came up in discussions with advocates of Land Value Tax, one of whom argued it effectively addressed deadweight loss caused by certain individuals essentially having monopoly rights over land.

Example argument:

Demand curve is currently moved to the right compared to a perfect market (where everybody rents their immovable property thus paying twice as much for their housing expenses) causing a deadweight loss. Evidence. We see excessive vacancy, under occupation

He seems to be saying that landowners will essentially horde land, because e.g. it would be better to sell one parcel of land for a million dollars than two lower the price and sell two parcels of land for 499,000. This seems a flawed argument to me but I'm still thinking on it.

Not sure exactly what the question is - would you like a discussion on how tax affects monopolies, or one specifically focused on Land Value Taxes?

I think this thread explains it best. I could summarize it - but it speaks for itself.
https://www.quora.com/How-does-a-land-value-tax-not-create-deadweight-losses

There's actually no deadweight loss in a monopoly market in the short term. Everything that can be supplied is supplied, whether you tax it or not. Just because I'm getting taxed 10% of my rent doesn't mean I'm going to supply my apartment any less. It's because of this lack of dead weight loss that LVTers say that it's the perfect tax. You can tax land owners to 90% and as long as they're still making their nominal profit they're going to supply the land to the market.

The author of that Quora response is the same person I'm responding to here. I wasn't talking about whether LVT creates deadweight loss, but whether it fixes some existing deadweight loss, which is what he seemed to be arguing elsewhere.

A simple thought experiment: suppose you have a tax of 100% of the market value of a parcel of land. Now just call that a "rent", because it is. Now realize that a rent is equivalent to a one-time payment of its net present value, so the government should be equally willing to sell land outright instead of the 100% rent/LVT. Especially given people want to be able to build immovable property on top of it.

Government definitely shouldn't institute a rent/LVT on land that was already purchased. When an LVT is instituted, it only harms the poor individual who randomly happens to be holding that "hot potato" at that point in time, by reducing the value of the land by the NPV of that LVT stream. If he sells it, the original owner has to discount the price by that amount. So it feels like a one-time tax on the current owner, which isn't justified by anything. It can't be, because it has nothing to do with his wealth.

He said:

only the individual or firm able to put it to it’s most productive use will pay the most to exclude others from an irreproducible factor of production.

This makes no sense. The current owner already has an incentive to make the maximum profit from the land. Adding a tax to it doesn't change that, aside from making him poorer and thus increasing the marginal utility of his dollars which may incentivize him to work more.

Also it doesn't matter that land is irreproducible. All resources are finite, including human labor. (And actually, you CAN make more land. See Holland and San Francisco.)

I'm still not sure what exactly you're arguing for here.

To level set: There is no deadweight loss in the existing land owning market as the supply curve is very close to straight up and down. Normal theory on Dead Weight Loss says that when you introduce a tax (say 10%) on a normal good, such as soda ($100M market), the consumer will consume less, and the supplier will supply less. You might think that a 10% tax on a $100M market would lead to $10M. But the tax will end up shrinking the market to $70M, and the tax will be less effective than expected $7M. The dead weight loss here is $30M which simply vanishes into 'inefficiency'.

The reason that LVT is a good tax is that the supply of land is very nearly always fixed. Suppliers/Landlords have no marginal costs that they would save by holding back their land when they are taxed - they just have to eat the tax (and still earn a nominal profit).
It doesn't fix deadweight loss because deadweight loss isn't a factor in the market.

LVT isn't a good tax because it fixes a negative externality or corrects a market. LVT is a good tax because it can't distort the market. In an ideal world, we'd put in place taxes to correct for market failures such as congestion and pollution taxes first - but when the market obeys those incentives and reduces pollution tax revenues will fall.

LVT taxes make a great backstop because the taxes can't be avoided, and will consistently generate revenue so the government can provide services.

I'm still not sure what exactly you're arguing for here.

Foundational principles on "what to tax or subsidize", as well as the particular use of these principles to show that the arguments for LVT are incoherent.

But the tax will end up shrinking the market to $70M, and the tax will be less effective than expected $7M. The dead weight loss here is $30M which simply vanishes into 'inefficiency'.

The only reason to tax a thing/activity is if it's a negative externality in the first place, in which case there's absolutely no dead weight loss. The opposite. You tax soda if you think it produces a negative externality, in order to avoid dead weight loss.

The idea of a Land Value Tax doesn't make sense as a tax. If you think about a carbon tax by contrast, it's there to shape incentives, to encourage people not to produce carbon emissions. But if the amount of land really is fixed, then there's nothing to disincentivize by taxing it. Actually, technically, you can produce more land by "terraforming"—so LVT actually would disincentivize doing that, as if it's a negative externality. Which... it's not, in any clear sense. (Yeah, there are complex environmental impacts but that's different.) So it just makes no sense as a tax.

The deadweight loss argument is saying that the point of LVT is for government to make money, in a way which seems to have minimal downside. But actually, that inability to shape incentives is precisely why LVT is irrational. It effectively is just a tax on the owner. But when you tax people (rather than things/activities) that should be based purely on their wealth. Richer people have a lower marginal utility per dollar. But that's entirely orthogonal to land ownership. So LVT makes no sense.

You're confusing the economic reality of supply and demand curves with morality. Let's ignore if a tax is just or not, and let's just look at how a tax effects the price, supply, and demand.

A deadweight loss is defined as "a loss of economic efficiency that can occur when equilibrium for a good or a service is not achieved."

This could be in the trade of harmful drugs or in helpful medicine. There isn't a moral connotation to deadweight loss. Take for example a price floor on medicine. https://en.wikipedia.org/wiki/Deadweight_loss#/media/File:Deadweight-loss-price-ceiling.svg The government will intercede to lower prices, and producers will be affected by that. Producers will be less likely to supply into that market, leading to scarcity. That scarcity is measured as "dead weight loss."

Now let's look into the other issue you're discussing, which is Pigouvian taxation. You establish a few criteria about "sensible taxes" but you ignore the #1 reason for taxes: to fund the government. Any benefit besides funding the government is auxiliary in the current mindset. You're right - Pigovuian taxes are better for society and should be the primary form of taxation.
But let's say that you put in place all the Pigouvian taxes you can. But it's not enough! The carbon taxes brought carbon emission rates down. Everything is running on green energy, so nobody has to pay the carbon tax. Congestion charges make transit flow effectively, and most people take transit, dodging the tax.
But now the government has no money to pay for the services that don't pay for themselves. Where does it raise money from?
Now you need to setup arbitrary taxes. Which arbitrary tax is the least bad?

Let's ignore if a tax is just or not, and let's just look at how a tax effects the price, supply, and demand.

Whether it's "just" (maximizes utility) is precisely the thing at issue here.

There isn't a moral connotation to deadweight loss.

It's a moral issue in that it affects welfare. You want to maximize net wealth to maximize net utility.

you ignore the #1 reason for taxes: to fund the government

No. Funds to the government meet the same criteria as taxes. They are useful to either directly give to individuals to improve their welfare, or to address positive externalities. If you run out of externalities to tax, but you have an opportunity to improve welfare, you can then start taxing wealth (people, not activities) to spend on that welfare improvement subsidy. But it's virtually inconceivable for it to make sense to tax activities that aren't externalities.

And my other point was that LVT effectively is a tax on people, not activities. So you should just tax the richest people, not people who happen to own land.

Deadweight loss is an economic definition. I agree with you that it's better for markets to be efficient - and therefore deadweight loss should be avoided. LVT is a tax which does not cause deadweight loss. Because of this efficiency, it is the least distorting tax. As you were saying it's a moral issue to maximize net wealth to maximize net utility. Because LVT maximizes net wealth by not causing deadweight loss, it also maximizes net utility.

How else would you tax the richest people, in a nondistorting fashion?

LVT is a tax which does not cause deadweight loss. Because of this efficiency, it is the least distorting tax.

But it's effectively just a tax on a person rather than on an activity. It's like saying, "We need to raise 1000$, let's have a lottery and tax some random person 1000$." You'll be better off taxing the richest people, rather than random people who happen to own land.

It's a tax on people who invested in real estate. The activity is "land ownership." Not all people own land.
1. All economic activity is ultimately derived from land, so if you tax land, you tax all economic activity on an even basis.
2. Land owners are monopoly rentiers - their wealth earned from investment in land is not in proportion to the risk that they would take as if they were loaning their money to the capital markets. Because of this, it is far more ethical to tax them then a capital gains tax or an income tax.
3. A random lottery tax has no progressive nature. Land taxes tax wealth. A tax on wealth is inherently progressive and better for society. If they don't want to get taxed, they should sell their land.