Land Value in the Digital Realm

Adam Smith divides the economic forces into three areas for, the profit obtained from labour, that from capital (or stock), and that obtained from land. The value extracted from land is here the natural resources of the earth, the sunshine and rain that lands there (allowing for agriculture), the timber that can be cut down, or the ore to be mined. But there is another value that can be obtained from the land.

In the movie Up (2009) the protagonist Carl Fredricksen refuses to let his villa be demolished despite the large amount of development occurring around him. Since he owns the land that the house is built on, the construction companies can not develop it, even as it is surrounded by soon-to-be skyscrapers. Mr Fredricksen is offered “twice [as much as the] last offer”, showcasing the value of the land. But Mr Fredricksen has not exploited the land at all, he has simply continued living in his house as the area was developed. So where has this value come from?

It has come from the labour and capital of others, in its proximity to services provided by the rest of society. This is the observation made by Henry George, one of the great (but quite overlooked) economists of the 19th century. The value of land is then partly the natural resources in and on the land, but the value of the land is also a social construction originating from how much the rest of society wishes to make use of that land.

When American settler colonialism spread over the North American continent, the US government gave out the lands settled for free. This perhaps sounds like an amazing deal today, but it is easy to overlook the fact that the value of the land was perhaps zero, or at least close to it. It had no significant infrastructure present, was located far away from centres of industry and trade, and was surrounded by a hostile nomadic population. And yet today the value of land in the Las Vegas Strip, built in the close-to uninhabitable Mojave desert, is extremely high simply due to the amount of wealth and development in Las Vegas.

While land may have some innate value, the value of land is mostly a social construction created by the labour and investment of society.

Digital land is however practically infinite. Much like the vast expanse of the empty American continent, it is mostly worthless, with a few islands of “civilisation”. The most obvious example of this are domain names. google.com may be the most valuable domain name right now, but it does not obtain its value from any innate quality, but merely from the service that is running when one connects to that domain. gooåle.com does not have any value, despite being “close” to google.com.

Short domain names, like the (in)famous x.com or the more wholesome x.org are by some measures valuable due to their rarity¹, but many very short domain names are still very cheap. Something that makes a bigger difference is if the domain is a word in any language, or an acronym for a series of words (although it turns out many of these are also unused). That domain names that are words are more valuable makes clear that their value is of a social nature; had the names and words been different we would expect different domain names to be valuable. To summarise, the value of a domain name mostly originates from what is built on it, not any innate quality, just as the value of land originates mostly from where it is relative to other valuable land.

But this observation is not limited to domain names, but also to “digital platforms”. Facebook does not derive its value from its UI or its ability for targeted advertisement, but rather from its active user base. The conceptual location of Facebook derives its value from the fact that there are many other users there to interact with.

This is likely not too surprising for many readers, who may have even tried to ditch Facebook specifically (or thought about doing so) but given up due to the fact that “everyone else is on there”. In economics, this is called a network effect, where the more people use a service, the more useful it becomes. But this is not always beneficial to the consumer.

When a service has obtained enough market share, it becomes impossible for the user to change to a different provider, as the dominant player absorbs all new users with it’s huge network effects, causing an enormous collective action problem with thousands, if not millions, of individual actors. This inability to switch means that the dominant player can extract extra costs on to the consumer, either via directly raising prices or through indirectly exposing them to more and more advertisement. This is in effect a tax on the consumer, not connected to the costs of costs of providing a service, but instead a mere increase in the share of profit.

It may not be immediately clear, but much (if not all) of this applies to land too. As we have already established, land derives its value from the fact that other people are using it, or at least people nearby. But what we have not covered is the fact that controlling land that is in demand — land in cities especially — allows the owner to charge exorbitant prices to those who wish to make use of it. You can rent out a plot of land in the middle of Manhattan for much more than you could a similarly sized plot in the Libyan desert, for example. If there were two identical hotels in each of those places, the prices I could charge would both differ. There would of course be running costs that would be incorporated into the price, and the price in the desert may even be higher, but the share of profit I could take from the one located in New York is far greater. This is because my customers are not only paying for the services of the hotel — the food, the maintenance of the building, paying the staff — but also for the labour and investment of all the people surrounding the hotel, who you might want to visit, for which I do not have to pay anything at all. This is how digital services expose their quality as land, as a digital space where the labour of others is used to extract an artificially high profit without a proportionately high amount of investment or work.

No human being created land, and so no one has claim to it originally. In Das Kapital, Marx, as part of his reasoning behind primitive accumulation, describes how the origin of land and wealth ownership is where “conquest, enslavement, robbery, murder, briefly force, play the great part”. Henry George concurs, saying that

Historically, as ethically, private property in land is robbery. It nowhere springs from contract; it can nowhere be traced to perceptions of justice or expediency; it has everywhere had its birth in war and conquest, and in the selfish use which the cunning have made of superstition and law.

It is then morally unjust not to tax the increased value granted to land by the rest of society. But digital land is not the same, it was created by man, and there is no deeply ingrained story of murder and robbery within it, except perhaps wasted time. Is it then simply the case that the rent earned on the network effects of digital communication is the reward earned by those early movers who dared to risk their money and time? The answer is no.

Obviously the entrepreneurs starting these companies should receive a financial return on their investments (and it may be the case that they are dominant in their sector simply due to being the best) but they are not responsible for the returns owing to network effects. Companies should then be taxed for the value of these effects, since the users are those responsible for creating this value in the first place. Some major tech companies have repeatedly threatened to stop offering their services in the European Union after certain law proposals like the GDPR, but have chickened out of doing so. This is because their profit originates from those users, and their ability to not only use them to create the very value of their digital land, but at the same time to extract rent from our digital neighbours’ presence there.

This system of taxation is called a land value tax (LVT), a tax on the underlying value of the land, excluding any buildings or improvements to it. This has many benefits, such as the ethical merit of returning the stolen money to those who deserve it, improving economic growth by not hurting developers of land (like a property tax does), and allowing people to live where they want to live by giving workers a larger share of the value they produce. But one massive benefit is the ease of collection.

Taxes on wealth are famously difficult to collect, as money is easy to move around or deny possession of. But land is far more difficult. A plot of land with a given value should return a set amount of money, or else be seized by the state and sold. Who owns it doesn’t matter, and denying ownership doesn’t prevent losing it. Land can not be moved to a tax haven overseas or hidden away, it is not difficult to determine what tax jurisdiction it is present in. But digital land does not exist in any tax jurisdiction. It is a purely anational concept that transcends the nation state. So who should determine the tax, and to whom does it belong?

Who should collect the tax is comparatively easy, the country where users are present. A German-focused website should not pay any taxes to the Indian government, and an internationally used service (like Facebook) should pay taxes the each jurisdiction corresponding to its share of users. This has in practice already been proposed under the name of a Digital Services Tax (DST) although the name might make you think it is a tax on digital labour rather than digital land. DSTs are quickly becoming increasingly popular, especially in rich-world countries that lack a significant presence in these digital spaces such as India, the EU, The Commonwealth, and Israel. That these states have in practice independently discovered the teachings of Henry George almost 150 years later, albeit in cyberspace rather than the real world. If that fact does not convince you, I would recommend you read Progress and Poverty yourself and email me your thoughts on it.

This is not a perfect solution. An LVT is by comparison far more elegant, and is less liable to loopholes. One of the elegances of an LVT is the ease of collection and of its undeniable nature, but a DST is much harder to “pin down”.

Companies could deny user metrics, and the very fact that one needs to track users in the first place is problematic. It is not only concerning from an ethical viewpoint but for technical reasons as well; and it philosophically contradictory to the Laissez-Faire evolution of the digital realm that has facilitated its construction. It is also hard to apply to other forms of digital land that don’t have easy-to-split-up revenues, like cryptocurrencies or the value of NFTs.

The implementation is also perhaps more formulated as a response to digital rent-seeking as a whole, rather than a system to reduce rent-seeking in collective digital spheres specifically. I am thinking about these subjects continuously, so do reach out if you have any solutions or other ideas.