An Introduction to Land Value Taxes

Let’s review. As I’ve discussed before, it’s important in tax policy to measure how people respond to higher taxes. For example, if you tax people’s wages more, they will work less. But this is just one example of a larger idea: barring tax avoidance, if you tax something, the cost of supplying it goes up, so less of it will be supplied.¹

If people don’t change their behavior much in response to a tax, then that’s a good tax to focus on. We could get a lot of money out of it without much economic distortion (or deadweight loss, as economists say). Now, what if taxing something leads to no less of it being supplied? Well, then the revenue-maximizing tax on that good would be 100%. In other words, we could collect all of the value of that good in taxes without any distortions.

The Perfect Tax

This could only happen if changing the supply of something was impossible. In other words, if the supply was constant. There is only one thing in the world for which this is true. As noted by economist Henry George in the late 1800’s, you can work more, or make more food, or invest more-but you cannot make more land, nor the natural resources in that land. So, land rent (the money landlords make from owning land over time) is entirely unearned. The land will be there whether or not landowners get payment from it.

Crucially, these land rents don’t just include income from things like oil or being by a nice lake. Think of a property in the suburbs versus a property near New York City. Let’s say both properties have the same structures: houses, stone walls, whatever. And, say they have the same square footage and type of land. Which property will be more expensive? If you’ve ever looked at home prices there, you know the answer would absolutely be “the one in the city”. But how could this be? The structures, amount of land, and type of land are exactly the same. Well, there is one difference: the location. Demand for locations near the city are higher, so of course the property will be more expensive. The amount of locations/space is, of course, fixed.

Econ 10… something

Land is valued highly in cities in the first place because of what are called agglomeration effects. These are cost savings and productivity increases that come from having high population density, which gives access to a large, diversified workforce to employ, many different suppliers to buy from, and low transport costs.

Again, it’s important to emphasize that these agglomeration effects do not come from the landowner. To quote Adam Smith, “the landlords, like all other men, love to reap where they never sowed”. If I own land in New York City, its value comes from all of the people, business, and infrastructure around it. This framing shows why landlords can’t deserve land rents: there is no sense in which the value is due to them. Now of course, I could make my land more suitable for construction by flattening it for example, and this would increase its value. So I should be clear that when talking about land value taxes, I mean the unimproved value of land: value that comes from location and natural resources alone.

A common response to a land value tax (LVT) is: well, won’t landlords just raise rent if we tax land? It’s counter-intuitive, but the answer is no. Think of it this way: taxing land doesn’t change the demand for it. And the supply of land is constant. So, as long as supply and demand truly determine price, then the price for renters must remain the same too. The tax is therefore entirely paid by land owners. In other words, the tax doesn’t raise the profit-maximizing price, and it doesn’t lower land rents.

Rent vs. Value

I’ve been using “land rents” and “land value” interchangeably here, but that can be unclear. Land rents are the economic value you get from owning land over time. But land value, in the common lexicon, means what your land can sell for. Land value is based on land rents and the discount rate (how much you value money now vs. later). For example, say that owning a plot of land would make the owner $10,000 a year, and that people value a dollar now as much as 90 cents a year from now. Then the land would be worth $10,000, plus (0.90 * $10,000 = $9,000) next year, plus (0.90 * $9,000 = $8,100) the next year, and so on². A full LVT would be set equal to land rents. So on net, a landowner would make $0 every year from the land alone, meaning the land value would be $0 as well.

Importantly, these rents are based on the most productive possible use of the land. So it doesn’t matter what you’re actually using the land for, the LVT would be the same. This makes sense: you’re welcome to misuse your land, but the community should be compensated for this loss in economic output. A LVT would also encourage efficient land usage and development. At the moment, a landowner can just hold onto a plot of land and do nothing with it, paying very little in taxes and probably selling it for a profit years later. This is why you sometimes see empty plots even in growing cities. But under a full LVT, the only way a landowner doesn’t lose money from owning land is if they put it to productive use.

Conclusion

It’s important for me to emphasize that land rents are not an insignificant amount of money. Estimates put land rents at over 13% of national income (~$2.8 trillion dollars). In a sparsely populated country like the US, this is largely driven by urban land rents, which are up to 27% of income in big cities. That is a tremendous amount of money we can collect without distortion to fund progressive policy from infrastructure to UBI. Collecting land rents should be a key part of any left-wing agenda.

¹ Note how I proved last time that the revenue-maximizing flat tax rate is 1/(1+e), where e is the elasticity. In simple terms, the elasticity of land is zero, and 1/(1+0) = 1, a 100% taxation rate.

² Folks may see this is just an infinite geometric series, the formula for which is well known. With land rent L and discount rate r, total land value here would be L/(1-r).