Let me give an example. Let's say there is currently a property in San Francisco. It's 2100 SQ/FT, it's current market value is $2.5 million, and the house is a

Let me give an example. Let's say there is currently a property in San Francisco. It's 2100 SQ/FT, it's current market value is $2.5 million, and the house is appraised by insurance for $500k. The most optimal result, the result that we call 100% LVT(although 101% is much worse than 99%, so we try to aim cautiously) - is if the new market price after a LVT is implemented is $500k. Maybe that means a $24000/year tax. Maybe it means more. We keep raising the tax until the sale price asymptotes at the value of the improvements. Let's say it ends up at $24000/year. $24000/year is the value of the positive externalities that is being captured by this property, and we are therefore taxing the entirety of it. It might help to also show you the lot next to it[0]. It's totally unimproved. It's going for $2 million. The goal for the land value tax is if the new sale price of that lot is $0. You'll note that if you try to do some math of current sale price to get some tax percent, you can't ever really get to zero. It would need to be an infinite percent tax. But that's because the tax is not based on the market price, but a set amount based on the lot. Because we know the lot next door is exactly the same size, and we know this lot is verrry very slightly closer to a train line, and we can see the general trend of the city is that lots closer to the train are worth more, so this lot based on some algorithm might be $24005/year, and we confirm that when someone buys it for $0. If someone bids thousands of dollars for it, next year we adjust the tax on it slightly up. [0]: https://www.zillow.com/homedetails/736-Hyde-St-San-Francisco-CA-94109/2065075537_zpid/ https://www.zillow.com/homedetails/736-Hyde-St-San-Francisco...