That depends whether you're a politician or an economist. ;-) Through a politician's lens the goal of a tax is to raise revenue for government and tax policy exists to do that efficiently. Through an economist's lens efficient trade and alignment of incentives are fundamental and governments exist only incidentally to ensure those. There are various schools of economics that suggest limiting government's role only to the enforcement of private property rights and kneecapping its ability to do anything else.

It's much like how to a physicist, vectors are real objects that live in the abstract concept of a vector space, while to a mathematician vector spaces are fundamental axioms and the existence of vectors is just a consequence that falls out of these axioms.

You missed out an important perspective that is neither purely politician nor purely economist: taxpayer.

Taxes are paid on money earned, whether through labors, investments or rents. They are a taking, a legal one, of something that rightfully belongs to someone else, so if you are going to take it, then it had better go back to government services.

Economists have found ways to reinterpret taxes and exploit their second Order effects, but this does not change what a tax is or how it should properly be used, nor excuse the misuse of surplus revenues.

Now the reason I bring this up is because if a tax were enacted for the express purpose of shaping public behavior and the resulting revenue was greater than what the government needed, what ought to be done with the surplus revenue? Should the tax be lowered to match expected expenses? What if lowering it to that level negates the benefit of shaping public behavior? These are the conflicts I would expect to see come up.

This is one of the reasons why the discussions around taxes tend to fall along the lines of what is taxed, how much, and where the burden should proportionally fall. Leading with second order effects muddies the waters and hinders productive discussion.

> [taxes] are a taking, a legal one, of something that rightfully belongs to someone else

IMO a LVT holds up (relatively) well under this sort of moral reasoning. I'm sure many arguments can be made around what's fair, but it's intuitive that income from labor, investments, etc should naturally go to who performs the labor, who makes the investments etc.

The (unimproved) land itself is a bit different. It can't go to who created the land, since no one did. In the US, the closest thing is maybe who first took it from the Native Americans. The moral argument for strong property rights (on the land without improvements) to begin with seems weaker than the other cases, since there's no contribution to society tied to making the land (of course switching from one tax regime to another is another matter, since people certainly buy land that is valuable under our current system with their hard-earned money).

That also leads to its (relatively) small effects on incentives for a tax. Taxes on labor and investment income of course discourage labor and investment, but the amount of land does not decrease no matter how much it is taxed. Even a huge tax probably does not affect the monthly payments to use land very much (since the supply and demand still remain similar), it would probably just lead to lower land values and more of the monthly payment going to taxes instead of interest.

> Now the reason I bring this up is because if a tax were enacted for the express purpose of shaping public behavior and the resulting revenue was greater than what the government needed, what ought to be done with the surplus revenue?

I think this is an interesting question especially with things like carbon taxes being considered seriously. It seems like two natural things to do would be to either reduce other harmful taxes that exist only to raise revenue, or to distribute the funds to everyone equally. I could see an argument for the latter if the incentive-shaping tax was something that would raise living expenses for most people (for example, a carbon tax that increases electricity costs), since otherwise it would negatively affect people living off savings or others with low incomes.

I'm aware of the taxpayer perspective and it's a big problem with getting adoption for an LVT. It's the type of policy that makes most people better off but you can never really convince people to adopt, because it sounds worse than alternatives.

There's actually an economic theorem (the "Henry George Theorem", discovered by Joseph Stiglitz in 1977 [1]) that states that under conditions of democracy & perfect information an LVT will exactly "right size" the government, spurring investment in only those public goods that pay for themselves in increased citizen welfare. That's because the tax revenues from public good expenditures come back in the form of increased rents: a playground nearby, a subway stop, better schools, low crime, stable currency, a strong military, etc. all increase the amount of rent that a landlord can charge for a property within their jurisdiction. That rent then comes back to benefit the exact same people who are paying the tax, which removes the principal-agent problem inherent in most taxes (where you pay the taxes but it benefits your neighbor). Landlords have an incentive to raise taxes on themselves right up to the point where the public improvement fails to raise rents by more than its cost. Tenants have the same incentive - as long as the cost is less than the public benefit, they get to enjoy the benefit, their rent will raise by the cost (as landlords pass along the tax), and they end up better off. When costs rise more than benefits (as they seemingly have in SF), people leave the expensive jurisdiction and move to areas that have more smartly allocated their tax dollars.

The main problem in practice is that you can always direct people's attention to small problems that don't raise their welfare in aggregate much - saving the burrowing owls, for example, or banning abortion. The economist's answer is that this will eventually even out across many people and many municipalities, but I think we have pretty ample evidence that people are not necessarily rational beings, and an LVT isn't all that robust to irrationality.

[1] https://en.wikipedia.org/wiki/Henry_George_theorem

Is there a textbook that teaches these sorts of theorems? I'd love to read economics from a theorem-proof perspective

> The primary reason for any tax at all, regardless of type or preference is to raise revenue for government. That fact shouldn’t be lost when discussing which kinds of taxes and what percent to raise them at is discussed, even when optimizing for political outcomes.

That's a political statement speculating at intentions. For instance, the number one purpose of a carbon tax is to disincentivize carbon emissions. This is exemplified by the fact that carbon tax proponents generally support returning the money to all taxpayers as a 'green dividend', i.e. it's a net-zero redistributive policy.

Raising government revenue isn't that important compared to the prime purpose of the carbon tax.