Lars Doucet — Progress, poverty, Georgism, & why rent is too damn high

Lars Doucet, the developer of the game Defender’s Quest, is on the show because of a review of Henry George’s Progress and Poverty that won Scott Alexander’s Book Review Contest and has since grown into the book Land is a Big Deal. Doucet describes Georgism as the philosophy of the 19th century American economist Henry George, and says George’s thesis is essentially the title of his own book.

Doucet notes that Georgism is often reduced to its main policy prescription: a land value tax on the unimproved value of land, with no tax on buildings, infrastructure or anything else humans add. In his view the deeper insight is that land plays a hidden and underrated role in the economy, which real estate sayings such as “location, location, location” already hint at. He calls control over land the oldest struggle in human history, one that animals share too, and argues that over the last century Capitalism and Socialism have absorbed the debate by assuming you must be either pro-worker or pro-business. Georgism, he says, is genuinely both, and what it opposes is land speculation.

The title of George’s book names a paradox: why does poverty advance even when progress advances? Doucet points out that industrial technology in George’s era and computers in ours let us make more than ever, with enough wealth for everybody, yet inequality persists. He presents sharing the earth as the way out of that paradox. George answers the question in his book, and Doucet says he expands on it in his own.

Dwarkesh asks whether Doucet’s interest comes from dealing with platform rent seekers such as Steam or the iOS app store. Doucet says that is not what first drew him in, though he has become interested in rent seeking in what he calls land-like assets, which produce first-mover advantages in large platform economies. For him the essence of land speculation is a class of people who gatekeep a scarce resource everyone needs and cannot opt out of needing, so everyone must pay them rent. These people do not necessarily do anything; they got there first and say, in his words, “if you want to participate in the world, you need to pay me.”

The connection that actually led him to Georgism was repeated virtual housing crises in online multiplayer games. In Ultima Online, which he played at 19, houses were fairly cheap: a player could work long enough in the game to afford a house blueprint, but there was no land left to place it on. He assumed the server had simply filled up. Twenty years later he checked back and the crisis was still ongoing in a game that had run for about 25 years, with an entire black market for housing. He saw the same pattern in Final Fantasy 14.

The pattern returned in 2022 with the wave of crypto games, including Axie Infinity, Decentraland, the Sandbox and Yuga Labs’ Other Side, which held large land sales. Doucet was then an analyst at the video game consulting firm Novik, and he told his employers that virtual land speculation would reproduce the conditions of real-world housing crises. He says he was proven right and that the whole cycle has played out. What struck him most was that real-world problems were reproduced so precisely in a virtual world without anyone trying to do it.

Dwarkesh adds a joke he recalls from Tim Dillon on Joe Rogan’s podcast about Decentraland, where being Snoop Dogg’s neighbor in the Metaverse costs a couple of million dollars. Rogan asks whether Dillon can afford to live there, and Dillon answers that he will start another Metaverse and work hard.

Dwarkesh relays a Tyler Cowen blog post that compares taxing land with taxing unimproved labor or unimproved capital. On that logic, leisure is the labor equivalent of an unimproved parking lot in San Francisco, and capital kept out of the economy could be taxed at the rate of the capital gains on a T-Bill, so that profitable investments are not punished. He asks what Doucet makes of the comparison.

Doucet says the answer depends on the moral principle you start from. If the principle is efficiency, taxing everyone for wasted opportunity cost, it raises the question of who decides what counts as waste. The Georgist argument instead comes from a more fundamental idea of justice. A worker produces value through work, and a saver who accumulates capital provides something valuable too, since saving lets a factory exist or a shipping company get off the ground. Land is different.

But nobody made the Earth. The Earth pre-exists all of us.

Doucet says someone who provides land actually does the opposite: they unprovide it and then charge you for opening the gate. Taxing the unimproved value of land therefore targets unproductive rent seeking and non-produced assets, and it encourages people to produce labor and capital, which we want more of. He cites the aphorism that “if you want less of something, you should tax it.”

He grants that a more elaborate case might exist for taxing unrealized opportunity costs, but says it interests him less. His priority is to start by taxing things that nobody made and that people gatekeep, meaning monopolies and rent seeking, and only afterwards discuss raising more revenue. He criticizes the current mentality in which everything is a trade-off and we must accept the downsides of income, sales and capital taxes because the revenue has to come from somewhere. His claim is that it can come from a much better somewhere.

Dwarkesh then asks what Doucet would prefer if a land value tax still left a revenue shortfall: a tax on the opportunity cost of your time, or the existing income and capital gains taxes. Doucet says probably not the time-based option, because it has too many degrees of freedom and is highly subjective. Whether someone adds more value as a podcaster, a computer scientist or a startup founder may not be evident for a long time. He gives two examples: the inventor of FM radio, who threw himself out a window because the invention was not adopted in his lifetime, and Van Gogh, who thought he was wasting his life as an artist. Taxing such people on a judgment of their activity would also need a much larger bureaucracy.

Doucet’s proposed order is land value tax first, then severance taxes on natural resources and other taxes on monopolies and rents, including land-like assets that are not literally physical land. Only if more revenue is still absolutely needed would he turn to what he calls the bad taxes, such as capital, income and sales taxes.

Asked what a severance tax is, Doucet defines it as a tax on the extraction of natural resources. He points to Norway’s oil industry, which he calls massively successful and a key reason Norway avoided the resource curse. A Georgist purist would say it is essentially a land value tax of a different kind: land is not normally used up, but non-renewable resources can be.

His example is Nestle and the water it uses. In his account, the company is enclosing a pre-existing natural resource that used to belong to the people, then bottling and selling it. It should be able to realize the value of what it adds to the water, he says, but not profit merely from taking the resource away.

Dwarkesh raises a puzzle: even without a land tax, why would a San Francisco property owner not already be incentivized to build as much as the law allows and collect rents, instead of keeping a parking lot? Doucet answers that a land speculator is optimizing for rate of return. If land appreciates by 10% a year, you are incentivized to invest in vacant land or tear down a building, because a tear-down property’s building has negative value and the land is cheaper because there is garbage on it. In his account, the marginal dollar is better spent on more land than on building up. He mentions a diagram in the book, put together by someone for him, that illustrates the rate of return.