But something about them feels off to me: They feel like a combination of a tax on the actual value of land and the value of degree of connection to other valuable nodes.
Like, land in Manhattan is probably not much more productive than land elsewhere on the coast of New Jersey. It’s mostly valuable because of the connection to other productive parts of land very near to it. But there was no principled reason why New York is where it is, I think it probably could’ve been 50 kilometers further north or south and that wouldn’t change much about its productivity.
I think this is because there are marginally increasing returns to scale to agglomeration of humans, but which clusters someone bets on when they buy some land in a developing city is probably often just part of a coordination/anti-coordination game.
This seems connected to Metcalfe’s Law, and can probably be generalized to graphs (or even manifolds?), where some centrality metric could tell us the the connectedness value of some node (land) and thereby its value.
One example would be domain names: scarce, somewhat amenable to coordination (amazon.com is much more valuable because of some investment, and semantically “nearby” domains are also more valuable for it (e.g. amazon.co.uk or aws.amazon.com)). Value of unimproved domain names is determined by the number of “semantic connections” to other domain names, and specifically the value of those domain names.
I really like land value taxes.
But something about them feels off to me: They feel like a combination of a tax on the actual value of land and the value of degree of connection to other valuable nodes.
Like, land in Manhattan is probably not much more productive than land elsewhere on the coast of New Jersey. It’s mostly valuable because of the connection to other productive parts of land very near to it. But there was no principled reason why New York is where it is, I think it probably could’ve been 50 kilometers further north or south and that wouldn’t change much about its productivity.
I think this is because there are marginally increasing returns to scale to agglomeration of humans, but which clusters someone bets on when they buy some land in a developing city is probably often just part of a coordination/anti-coordination game.
This seems connected to Metcalfe’s Law, and can probably be generalized to graphs (or even manifolds?), where some centrality metric could tell us the the connectedness value of some node (land) and thereby its value.
One example would be domain names: scarce, somewhat amenable to coordination (amazon.com is much more valuable because of some investment, and semantically “nearby” domains are also more valuable for it (e.g. amazon.co.uk or aws.amazon.com)). Value of unimproved domain names is determined by the number of “semantic connections” to other domain names, and specifically the value of those domain names.
Or, in meme form (not my meme, source):