Just a couple of comments on the discussion.

First, I think using site (geographic position) rather than land makes the argument clearer. The argument is that site  tends to be in more inelastic supply than other aspects of production, and tends to be more socially dependent on its value. I accept that. But there are complications. For example, there are many alternative organizaitons of activities on a wide range of sites that could take place, so there are likely multiple equilibria, each with different site values for subcomponents of the system. Seeing that your particular site has site value is highly entrepreneurial, and can have positive spellover effects for an area. I think that's one reason why communities shy away from taxing site value.  For a broader prespective, in a decentralized system, no individual has an incentive to search for a optimum among the alternative equilibria in the system, so we have no way of knowing whether the system is in some sense at "best" optimum, whatever that may mean.

I don't separate out entrepreneurs and speculators; they are both buyers who have different ideas about how site value may increase, and how it can be increased. To incrase site value often requires innovative ideas--for example, making an site popular by providing services and cache to an area that benefit far more than his particular site.

It is also very hard to clearly distinguish site value from the structure value in practice, and thus as a practical matter, the Georgian proposals can be difficult to implement, and will cause problems. But in some cases the proprosals make sense as long as one recognizes the limitations.

The ideas about site value carry over to other activities--for example in "idea space" there are site values, which we let people stake out with patents and copyrights. Taxing "idea sites" more heavily than new ideas, which is developing new idea space might make sense, just as taxing land rather than property might make sense. We do this in a way--the limited patent length that we give achieves that end. But I think there are better ways to do it, just as I think that there are better ways to achieve George's ideas behind a land tax. For example, if the US had given all land out on 100 year leases, not in perpetuity, then when the lease comes due, (or long before it comes due) the government can renegoiate the lease, and get more money--in this system the government takes advantage of its lower social rate of discount and builds the difference into the ownership rights, and can use that tax revenue to reduce its other taxes on more elastic supply activities. As I have played with this idea, I have found that land with 100 years leases sells for just about the same as land in perpetuity, so from an individual's perspective, this change has few incentive effects, but over time it has major revenue effects. China has done this--it leases its land rather than sells it,  and over time, this will mean that China will have a much less intrusive method of getting revenue than the US.

David Colander