A land value tax for schools and local government

Early versions of this design put every tax into one national pool and paid everything out of it, including schools and police. It was simpler and it was wrong. Local services are local: the right amount of policing in Wheeling is not the right amount in Phoenix, and a federal pool that funds both will get both wrong. So the design carves out a local leg, funded by the one tax base that is local by definition.

The tax

2.5 percent a year on the assessed value of land, not buildings. Land is worth about 1.2 times GDP in the Federal Reserve's accounts¹. When a land tax is announced, land prices fall by roughly the capitalized value of the tax, so the model nets 8 percent off the base. A homestead exemption worth about $200 billion a year removes most owner-occupied lots below a local threshold. What is left raises about $630 billion in 2027 and $1.75 trillion by 2045, growing with the economy. Banks collect it in escrow alongside the mortgage, as they already do for property tax, and distribute it: 75 percent to the school district, 25 percent to the local government.

Taxing land rather than buildings is the oldest idea in public finance still waiting to be used. It cannot be avoided (the lot is where it is), it does not punish building (the tax is the same whether the lot holds a parking space or an apartment block), and it recovers for the public the value the public created with roads, transit, schools and neighbors. Speculators who hold land empty waiting for the city to grow around it pay the same as the builder next door, which is why they are on the losers' list in piece 12.

The local essentials grant

The land tax does not cover local essentials on day one. The model prices the essentials of local government at about $560 billion in 2027: police, fire and EMS at $650 per resident, roads $300, courts and permits $250, parks and libraries $200, administration $150, public health $120. The local quarter of the land tax covers $160 billion of that in 2027. The pool fills the gap with a local essentials grant, $403 billion in the first year, that shrinks on two tracks: the land tax grows with the economy, and local costs fall as shared AI back offices replace paperwork (the model assumes 25 percent cheaper by 2037). The grant reaches $99 billion in 2040, $16 billion in 2045, and switches off around 2047. Schools get their three quarters from the start; the remaining K-12 gap after the land tax is an open question we flag in the last piece.

What it does to the universal income

Moving the land tax out of the national pool, and paying the local grant from it, cost the payment about $2,600 per adult in the early years and pushed the $52,000 year from 2040 to 2042. Put the schools' share to zero in the model and the grant balloons, because the local quarter is all that offsets it. The design accepts the cost. A universal income that quietly defunds the local school to pay itself would not survive its first budget season, and should not.

Phasing and the homestead

Land taxes are state and local instruments, so this leg is enacted state by state, phased over five to ten years, with the rate rising as the existing property tax on buildings falls. The homestead exemption is what makes it politically survivable: a family in a $300,000 house on a $60,000 lot pays about $1,500 a year at the full rate, before the exemption, and little or nothing after it. The burden lands on commercial land, large holdings, and empty lots in growing places, which is where the unearned gains are.