Guided by the Beauty of Our Taxes

The Theory and Philosophy of the Land Value Tax

The Land Value Tax is a graceful idea with an aesthetically pleasing philosophy and glorious theoretical properties. It feels like a slick solution to a physics problem that looked mind-numbingly tedious at first sight. Today we will leave the real world with its political economy and measurement problems, and ascend to the platonic plane of game theory, where humans are loss-function minimizers, and I will show you what's really possible.

The LVT

A Land Value Tax is a tax on land ownership, based on the "undeveloped value” of the land. "Undeveloped" means that the tax does not depend on the buildings and infrastructure developed on the land, in the way that property taxes usually do. Instead, it is a tax on the land at the same rate regardless of whether it holds a potato garden or a skyscraper, and regardless of whether it is being rented out, used by the owner, or left untouched.

It is like a constant 'rent' a landowner must pay the government. The tax rate per square foot only depends on the location: prime areas beget higher rates. It is a tax on land value. Ideally, the tax rate is set to be just slightly less than the market rental rate for an empty plot of land of the same size.

That's it, that's the beautiful tax. But before we see why this tax is beautiful, we must first see why most taxes are ugly.

I. All taxes are taxes on society

Taxes are a weird thing in statecraft. They are necessary: the state needs an income source to fund its activities. So you tax people for certain activities, like making an income or importing tea or having a beard. The problem is that taxes also "distort the economy". This means that people will change their decisions because of the existence of taxes, in a manner that results in worse outcomes for everybody.

Here is an example. Alia wants to get her hair cut, and is willing to spend up to ₹90 for it: her value of getting her hair cut is ₹90, and any haircut that costs less than this is a good deal to her. Bav will cut hair for anything above ₹70. They might agree to a price of ₹80; Alia gets her hair cut, Bav makes some money, and both of them gain value from the exchange ("they each gain a surplus of ₹10").

Now suppose there was a Haircut Tax of ₹25, charged to the provider. Alas! Now there is no price that works for both of them. Bav must charge at least ₹70 + ₹25 = ₹95 to make giving a haircut worth his while. But Alia will pay no more than ₹90. The exchange cannot happen! And they are both worse off than in the case with no tax: Alia does not get the hairstyle she wants, and Bav does not make any income. This is a distortion: some exchange that would have occurred has been prohibited. Value has been obliterated (further, the government loses out even more on income tax as Bav's income falls — so it's even more worse off than if it had charged no Haircut Tax).

The reduction of general welfare from the distortions due to a tax has a fitting name: it's called the deadweight loss of taxation. It evocatively conveys how devastating it is to discourage surplus-creating trades.

This is only the most overt kind of example; there are other more subtle yet no less impactful ways through which a tax can make a society worse off¹. In the end, the true cost of a tax on society ("cost of public funds") — including everything from deadweight losses to the cost of collection to costs of avoidance strategies — turns out to be much more than the government's actual revenue from the tax. In the fantastic In Service of the Republic, Kelkar and Shah estimate that for each ₹1 the Indian government draws in taxes, society loses more than ₹3 of value. Only ₹1 of this is lost as actual tax, and ₹2 is due to distortions, like the deadweight loss. And tax revenues amount to over 10% of GDP! Imagine the scale of the distortion, the loss of value!

We cannot do away with taxes, of course. But the scale of this value destruction tells us that we should really try to be very clever while designing taxes, because removing the economic distortions of taxes can restore to a society value worth a significant chunk of its GDP.

We do try to be a little clever about designing taxes. We try to target richer people since their wellbeing will be affected less by it, so we have bracketed income tax and luxury goods tax. We tax sin goods because a distortion in the sin good market might actually be good, as people might smoke fewer cigarettes. But there is a lot more weird clever stuff that is possible. There is a field of economics called optimal tax theory all about trying to design taxes that maximize social welfare, and the optimal tax theorists have some interesting takes.

First, there are taxes without deadweight loss. One trivial but implausible example is an unconditional lump sum tax: every individual is taxed the same fixed amount. This does not change anybody's choices, because they should still do what they were going to do anyway; the amount they will lose at the end of the year is set in stone, and unconditional on their actions. No deadweight loss! This is obviously a terrible idea in practice. But it is useful to appreciate the fact that getting rid of deadweight loss is not a theoretical impossibility.

There is another tax with theoretically no deadweight loss. This one, however, may not be a terrible idea in practice. It is the Land Value Tax.

II. Land rates and dead weights

Usually, when a tax is levied, fewer suppliers are able to profitably supply at any given price. For example, when there are no taxes, the set of suppliers who can supply a haircut at ₹90 is the barbers for whom it costs less than ₹90 to provide a haircut (including wages, materials, rent, and everything that goes into running a barbershop). If a tax of ₹25 is levied on each haircut, the set of suppliers who can profitably supply a haircut at ₹90 is actually the barbers for whom it costs less than ₹65 to produce a haircut, which is a smaller set. This is the phenomenon that raises prices and reduces supply when a tax comes into force.

But this changes when it comes to the land rental market, where the crucial difference is the fact that the total amount of land is fixed.

Suppose all land is owned by ethereal landlords, who rent their holdings out unto people and firms for housing and factories. Suddenly, LVT. What happens?

As long as the land value tax is less than the rental income of the asset, it's still profitable for a landlord to supply the property to the rental market. And it definitely is less, because LVT was defined to be slightly less than the rental rate of an empty plot, and so it would also definitely be less than the rental rate of a plot with a developed asset. A landlord wouldn't decide to not rent out their property, since then they would lose out on rental income and be strictly worse off. At best they can sell the land to another landlord, who would then rent it out. This is qualitatively different from the case of the barber, who could get forced out of producing haircuts. Land supply is 'inelastic' with respect to net rental income: all the land is always supplied. Only now, part of the rental income is extracted from the landlord as tax.

So there is no deadweight loss. In theory, nobody changes their decisions because of this tax. All trades that would have happened before the tax regime still happen. The only difference is that some of the landlords' rental income is going to the government. And it’s not stupid, like the lump sum tax: here, the tax targets landowners, who tend to be economic elites. We also know that they have the ability to pay the tax, because they are charged an amount less than what they can make from rent.

This alone is a remarkable fact. If you are an aspiring little state that faces no political barriers to passing tax laws, this might be a step you can take to increase the welfare in your country by a few percentage points: terminate some taxes with large deadweight losses, and make up the difference in tax revenue through this tax of no deadweight loss². That's it. All that dead weight is infused with new economic life and makes its way back into society.

Economists love the LVT for this reason. It’s sometimes referred to as ‘the perfect tax’, one that could replace all other taxes, and has been championed by the likes of Paul Samuelson, Milton Friedman, Paul Krugman, and Joseph Stiglitz (and these are just the Nobel Prize winners listed as supporters of the tax on the LVT Wikipedia page).

But wait, there’s more. That it causes no adverse distortions is among the less impressive effects of the tax.

III. The modern landed gentry

The place where LVT does its best work is in cities.

The value of a property can be split into two parts: the value of the building or development constructed on the land, and the value of the land itself. Depending on the building and the location of the land, the proportions that these two parts contribute can vary wildly.

In cities — especially rich, dense ones — a large part of a property's value really comes from the land and its location. In some areas, it is almost all the value of the property. Consider:

If you build a shack in the desert, nearly 100% of the property's value will come from the shack, because the land is worthless. But if you build a shack in San Francisco, nearly all of the property's value will come from the land.

See this map by the American Enterprise Institute, that shows the fraction of a property's value that comes from the land across the US:

In some of the USA's major cities, the land itself accounts for 70% of the value of a property on average. Seventy percent!

This is intuitively appealing. It checks out with personal experience. In big cities, if you search long enough, you'll come across flats that are in different leagues in terms of aesthetics and convenience going for comparable rents. That's because the quality of the house is only a minor part of the rent: most of it is just the location.

Cities are great economic machines of love and labour that chug along and do the good work of keeping civilizations alive. The story has been the same for a thousand years: you want to make money, you go to the city. It's because that's where you find people, opportunities, jobs, connections, and ideas. This is the premium: you are paying extra for the access to people, opportunities, jobs, connections, and ideas.

And so the city grows. As it does, demand for real estate raises rental prices. All the companies want to set up shop in the city, since that's where they'll find employers; all the people want to live in the city, since that's where they'll find jobs. The rent goes up. If you happen to own a few properties in good locations, then you've found yourself an excellent income stream. Be in the right cities at the right times, and you now have generational wealth.

Is this a problem?

There is usually nothing wrong in someone who owns something being able to charge high rates to rent it out. If the price of something is high, then it must be either that demand for it is high or that supply is low, and so the market is handsomely rewarding suppliers who are able to provide this rare or coveted good to society. Economists see prices as signals; a high price is like a bounty. It signals to individuals that the society is hungry for the good, and so a few of them will try to create more to supply it, seeking the reward. By doing this they satiate the market's hunger, and also bring prices down.

This is effectively the main working principle of prices in society³. This system has served us shockingly well: every prosperous society today owes its success in large part to the magic of the price mechanism.

But this idea is crucial: it tells us that the price mechanism only works its magic under certain conditions. The problem with the land rental market is that it's one of the places where it falls short. When land prices and rents shoot up, the price is attempting to signal to the supply-side of the market to create more land.

And that is the problem. The amount of land is fixed⁴. For land, the price ceases to be a useful 'signal' to the supply-side. Almost everyone in the city who owns spare land and cares about money at all is already renting it out. A higher price is not drawing new land into the supply market. And so the price mechanism takes on a grotesque form: it just keeps rewarding landowners without actually inducing further supply. Further, the value that is added due to the land is not by the landowner's merit in any real sense, but by the merit of the people around who are bringing the ideas and networks to the city.

Now there is a sense in which this is fair: the sense that through arrangements of history and law we have come to agree upon a certain notion of land property rights, by which a landowner should be able to do as they please with their land. In this view, we must accept that they should be able to make high returns from renting it out, if that's what the market allows them.

But also, there is a sense in which this is stupid: the sense that, look. The whole point of us getting together to do this whole 'law' and 'property rights' thing in the first place is for the benefit of society at large. But here, all that these rights are doing is allowing elite landowners to extract higher and higher rents, with no payoff to the rest of society⁵. The state is investing public money to enforce these rights! Perhaps this worked well once upon a time back when land only accrued value from what was farmed on it. Villages were tiny, and walking across one would take all of ten minutes. No need to try to be close to any place. Not that there was any place worth trying to be close to anyway; it was all just rice farms. For like ten thousand years. 

And suddenly, this wasn’t true anymore. It can take hours to get from place to place within some cities. There are skyscrapers and shopping centers and highways and train stations and other places you want to live close to. Living in the middle of nowhere sucks. The fraction of a property’s value that came from the location went from being approximately zero to much greater than half in many cities. Yet land rights remain mostly unchanged from their traditional versions from ten thousand years ago. Many historical arrangements have turned out to be worth axing to improve modern society. Maybe our idea of who should get the benefits from land capital is one of them?

This is the philosophical basis for the tax. It concedes that while our conception of land property rights is a good one, having served humans well for millennia, it is ultimately flawed when pushed to the limits to which we have pushed it, with our modern cities.

We noted that the rent that a landlord receives can be seen as coming in two parts: rent from the land's location value and the rent from the development. Under this frame, the LVT extracts the first part of the rent from the landlord, leaving them with only the development rent — the part that they actually contributed to. The part that is ‘location rent’ is a transfer that comes under what economists call ‘economic rent’ — a payment that someone is able to extract to give you access to a good that cost them nothing to produce. And economic rent is the worst thing ever.

In a perfect world we might want to assign to the surrounding networks and firms who raised the location value the land value part of the rent.

Some places do this by having the government own all land and rent it out. The LVT does this, but in a sneaky way, by:

1. pretending to just be a normal tax on land that you own, but is

2. actually equivalent to the land being leased to the landowner by the state, allowing the

3. state to spend the revenue for the welfare of the people who are the ones actually bringing the value, whilst

4. inducing no deadweight losses

IV. The benevolent ghost in the price machine

It would be nice, though, if we could devise a system that not only extracts undeserved profits to spend on societal welfare, but also incentivizes landowners to give society what it wants. In other words: fix the price mechanism on the supply-side. We want to ensure that if the town needs housing, landowners are building housing, and if the town needs sketchy laundromats, they are building sketchy laundromats. For this, we'd like that they be compensated exactly for the value that they bring.

The LVT solves this too.

Since a landowner can now only keep the part of the rent that comes from their development, they are strongly incentivized to build the buildings that society needs. Earlier, the rent they gained from land value would often be quite large. Large enough that perhaps spending time and effort to increase the development rent a bit might just not be worth bothering with.

But now, all that they keep is the development rent. They must create and rent out the buildings whose development rents are the highest in that location. The buildings that society wants. The incentives are perfectly aligned.

And they are punished for failure. If they want to develop something that is not valuable, or leave a prime plot vacant, they can, but they will constantly be charged a high LVT. If they want to build a winter palace that they visit once a year in the middle of a city, the city sends its blessings, followed by a large tax bill. The LVT is a penalty to the landowner for wasting a prime location that society could do so much with, and a compensation to said society. A landowner who cannot or does not know how to create enough value from the land and doesn't care to keep it will sell it to someone else. Someone else who will use it to develop something that creates more value.

This exactly fixes the broken rental price mechanism — no more or less. It restores meaning to the rewards to suppliers. Seen in this lens, the LVT is almost essential reading for free market societies. It doesn't just not distort the market as most taxes do, it actively un-distorts it, aligning supplier incentives more cleanly with what creates welfare. Wielded correctly, it can have not zero, but negative deadweight loss. That is, it can actually boost economic activity. The tax! It's hard to express how remarkable an idea this is. It is a tax — a thing that takes money away from people — and it could boost economic activity. Even if you threw away the proceeds.

In a sense, the land market is actually imperfect in the absence of the tax; the introduction of the LVT actually brings it closer to the spirit of the goal of free markets.

In theory, this fixes many, many things. Land development would become a fiercely competitive market, as substandard planning and quality becomes much more expensive. Central prime areas would start to be run by individuals and companies that master the art of efficiently developing and managing property (this is already kind of true, but to nowhere near the extent that it could be). The implementation of the LVT alongside fixing zoning laws has a good chance of fixing housing markets in many cities where it is broken. Landowners become finely tuned to the exact demands of society, and build houses, schools, malls, hospitals, and parks exactly as necessary.

This kind of magic is what the price mechanism has always done. And when it is truly set free in the land rental market, it's what it can do.

And so the obvious question.

V. If your tax is so clever, why is it still unimplemented?

It is time to descend from the platonic plane of game theory back into the mortal realm. Measurement problems return, and policies are not independent of politics. People have emotions about land, and state agents are self-interested. I'm sorry, but it is dangerous to spend too much time in the platonic plane. I hope you enjoyed your visit.

The following is no longer about economic theory, just my speculations the political economy.

The first and easiest answer is that we don't see LVT-type taxes much because incentives for its creation are not aligned. The LVT takes away landowners them the right to extract location rent. This immediately hurts all landowners in a city, and effectively ensures that a proposal to implement an LVT in a city with powerful and connected landowners dead on arrival. See this post on MR/Bloomberg for more discussion.

But blaming 'rich landlords' for all the bad things is the copout answer. It might be a part of the explanation, but there is probably more. If you were the tax czar and faced no political opposition, are there reasons why you wouldn’t want an LVT?

In a dense city packed with buildings and skyscrapers, how exactly does the government determine what the LVT rate should be? The LVT rate is supposed to be the market rental rate for an empty plot of land. So one might naturally say that the government should observe the market rental rate for empty plots, but… where are you going to find empty plots in a dense city? You'll find a couple, but probably not many, and not enough to accurately determine what the market rental rate is in each area. Wait, first of all, who's going around renting out empty plots of land in the city? Maybe it’s someone trying to manipulate the LVT rate?

One way out of this mess is to have a government-contracted analyst determine the market rate, and then undercharge, like by imposing a tax that's 80% of this estimated rate. The hope is that with such a large buffer you never need to worry about overcharging, and letting the landlord keep 20% of the land rent surplus doesn't seem like too bad a deal. I like this solution. But now you have government analysts with arbitrary power who will effectively get to decide on whom they want to impose taxes. This rarely goes down well, and this never goes down well in countries without razor sharp rule of law. Will it work out? Hard to say. More discussion here.

The LVT would have many surprising consequences. It is a bizarre economic invention, and not one whose real-life equilibrium effects we have good intuitions around.

For example, (more theory): under a perfectly implemented LVT regime⁶, the sale value of all empty land plots approaches zero. This can be proved with theory, but the broad idea is that it's because a land plot does not create an expected net cash flow for the owner. 

This is already bizarre, and probably breaks many things around which the world is set up to work. With good results, sometimes. The LVT basically puts an end to speculative land investments, since such exercises become expensive and useless⁷. But there will probably be domains in which the LVT causes bad outcomes too⁸.

Land also stops being an unreasonably good investment. It has been a truth universally acknowledged that a person newly in possession of a fortune must be in want of real estate. A lot of things are good investments, like large index stocks, but throughout histories and across cultures, land has meant power. It is a marker of status, a possible source of passive income, a foolproof means to propagate your excess to your future generations. Is this because land is magical and emotional and humans are deeply psychologically attached to the idea of owning territory? Or is this because land ownership accrues to its owner unearned land rents, and so our ancestral cultures learnt to love owning land?

Under LVT, land would become a commodity like everything else, and we would no longer have a landed gentry class. A landed gentry class is bad for a society in a way that a stock-marketed gentry class is not, because the former hoards a finite physical resource.

The implementation of the LVT can be, if your polity so desires it, a one-time wealth tax, since it takes away all the land value that all the landowners hold as future tax revenue. That makes such a sudden implementation practically impossible, but it is an entertaining idea. It's also necessarily one-time, since you can only once move from being a non-LVT regime to an LVT-regime.

The LVT is at the very least very interesting, and it's unfortunate that it's not found its place in common discourse, but I think it's slowly starting to catch on. I'm just happy to have come across a beautiful solution to a mechanism design problem.

1. A new tax creates new industries: industries of people enabling companies to get around the tax, and industries of lawyers that specialize in that tax, and industries that provide services to these industries, and so on. This is human labour that could have ex ante been directed at something more productive, that is instead being directed towards ‘administration’ of society. Note that I’m not calling tax lawyers unproductive; ex post having decided on implementing the tax, the tax lawyers do in fact play a crucial productive role in lubricating the functioning of society. But their removal from other productive industries is a cost, and one that should be considered.

2. One of the taxes you replace will need to be property taxes, since they cannot exist alongside the LVT. But that's fine, the LVT more than makes up for it in revenue.

3. Hayek, F. A. (1945). The Use of Knowledge in Society. The American Economic Review, 35, 519-530.

4. Kind of. If you do literally create more land — by reclamation — then there is an argument that you really should be entitled to all the rents.

5. Technically, the price signal still works on the demand-side — it allocates the good to the highest value user.

6. Under some assumptions on frictions, liquidity, etc.

7. The development value isn’t getting any higher with time if one is not building on their land. One would just bleed land value taxes until they are forced to sell.

8. Example: https://daviddfriedman.blogspot.com/2023/01/a-problem-with-georgism.html

---

I’m very grateful to Sidd and Maitreyi for proofreading and for their patient feedback. May you live in an LVT regime.