You can’t tax what doesn’t exist yet. Governments tax models, not reality — and homeowners get hit twice.
Origin: RattlesnakeTV review of Pool vs Kyla debate on LVT (land value tax) Links: Economics, Business Cycles, Inflation, The Equation of Exchange, The Weighting Problem, Market Efficiency and Human Limits
Governments increasingly tax unrealized gains — increases in the paper value of assets that haven’t been sold. This means taxing a model, not a transaction. The “value” being taxed is:
This is the same structural problem the vault identifies in GDP measurement, CPI, and “the price level” — treating a model number as though it’s reality. It isn’t. Reality happens at transactions, not assessments.
Homeowners are uniquely punished by a double taxation structure that applies to no other asset class:
The government assesses the value of your property annually and taxes you on it. This functions as though you’re paying “rent” on your own home — a fictional income stream that the government treats as real for tax purposes.
But you’re not generating rental income. You’re living in your house. The “imputed rent” is a statistical fiction — the same fiction that GDP accounting uses when it counts owner-occupied housing as economic output. No money changes hands. No transaction occurs. You’re taxed on what someone ESTIMATES your home MIGHT rent for.
When you actually sell, the government taxes the gain. NOW there’s a real transaction — cash changes hands, a gain is realized, and taxing it (whether you agree with it or not) is at least taxing something real.
The same asset is taxed twice through two different legal fictions:
No other asset class gets this treatment:
| Asset | Annual tax on assessed value? | Tax on sale? | Double taxed? |
|---|---|---|---|
| House/Land | YES (property tax) | YES (capital gains) | YES |
| Stocks | No | Yes (capital gains) | No |
| Gold | No | Yes (capital gains) | No |
| Business equity | No | Yes (when sold) | No |
| Bonds | No (interest is taxed, but not the bond’s market value) | Yes (if sold at gain) | No |
| Cryptocurrency | No | Yes (capital gains) | No |
If the government treated stocks the way it treats houses, you’d owe an annual tax on your portfolio’s paper value whether you sold or not. The outcry would be immediate. But homeowners accept it because property tax is normalized — “that’s just how it works.”
The double taxation is built on a measurement fiction that runs all the way up to national accounting.
How GDP counts housing: When calculating GDP, national accounts include imputed rent for owner-occupied housing. If you own your home, GDP counts the estimated rental value of your home as economic output — as though you’re paying rent to yourself and generating income.
This isn’t real economic activity. No money changes hands. It’s a statistical adjustment designed to make GDP comparable between countries with different homeownership rates (otherwise, a country where everyone rents would appear to have higher GDP than one where everyone owns, even if real living standards are identical).
How big is the fiction? According to the Richmond Fed (2025), imputed rent from owner-occupied housing accounts for approximately 12% of GDP. Twelve percent of the entire measured economy is a statistical fiction — homeowners “renting to themselves.” No transaction occurs. No money changes hands. One-eighth of GDP is made up.
And it distorts monetary policy: the Richmond Fed’s own research shows that imputed rent is a key driver of inflation persistence in CPI and PCE. After COVID, goods, energy, and food prices normalized relatively quickly. Housing inflation stayed elevated for years — because the imputed rent calculation lags behind real-world changes. The Fed is partially reacting to a fictional inflation signal when it sets interest rates. The fiction doesn’t just inflate GDP — it distorts the monetary policy that affects every borrower and saver in the economy.
The chain of fictions:
The homeowner pays taxes on the fictional annual “rent” their property generates (it doesn’t), AND on the real gain when they sell. The fiction and the reality are taxing the same underlying appreciation. And the Fed sets interest rates partly based on the same fiction, which affects the homeowner’s mortgage rate. The fiction compounds at every level.
The land value tax (Henry George, 1879) proposes taxing only the land value, not improvements. The theoretical argument: land value increases are created by the community (roads, businesses, population growth), not the landowner, so taxing them captures “economic rent” — unearned value.
Milton Friedman called it “the least bad tax.” Many economists agree it’s theoretically elegant.
Why it still fails in practice:
Same unrealized gains problem. LVT taxes the assessed land value annually. The assessment is an estimate. The owner hasn’t realized anything. The cash-flow problem is identical to property tax — you owe money on value you don’t have in your pocket.
Who assesses? The government — the same entity that benefits from higher assessments. The conflict of interest is structural.
Assessment is circular. Land value is estimated from comparable sales, which are influenced by existing land values, which are estimated from… comparable sales. There’s no independent anchor. This is the same circularity the Weighting Problem identifies in any composite index.
Bubble vulnerability. In 2006, land values were sky-high. By 2009, they’d crashed 30-50%. Anyone taxed on 2006 assessments paid taxes on phantom wealth that evaporated. The government keeps what it collected. The asymmetry is permanent.
The “didn’t earn it” argument cuts both ways. Georgists say you didn’t earn the land value increase, so taxing it isn’t confiscatory. But “didn’t earn it” also means “didn’t choose it” — the value increase was imposed on you by external forces. Taxing someone for an event they didn’t cause, on value they can’t access without selling, using an estimate they didn’t agree to, is coercion based on fiction.
The vault’s position, consistent across economics and epistemology: real economic activity happens at transactions. Everything between transactions is estimation.
| What’s real | What’s estimated |
|---|---|
| Sale price when property changes hands | Assessed value between sales |
| Cash actually received | “Imputed rent” |
| Realized capital gain at sale | Paper value of unsold assets |
| Fedwire transaction volume | GDP (includes imputations) |
| CPI inputs (actual prices paid) | CPI weighting (subjective) |
Building tax policy on the estimated column rather than the real column is the same error the vault identifies throughout economics:
The consistent principle: tax transactions, not assessments. When a property sells, tax the real gain — cash changed hands, the gain is realized, the amount is definite. Between sales, the “value” is a guess, and building tax obligations on guesses is building policy on fiction.
The RattlesnakeTV review that prompted this page also covered suffrage — specifically, whether people who don’t own land or pay net taxes should vote on tax policy.
The vault’s existing position from the SAVE Act page: restrict scope, not the vote. The problem isn’t who votes — it’s what they’re voting ON. If government scope is limited, there’s less to vote on, and low-information voters can’t do much damage.
The taxation debate demonstrates why scope matters: Kyla proposes a “genius” tax policy (LVT) without ever having owned land. She’s voting on policy that affects a class of people she doesn’t belong to, using a model she doesn’t fully understand, on an asset she’s never held. This isn’t an argument against her right to vote — it’s an argument against the scope of what votes control. If taxation policy were more limited in scope (less government spending → less need for revenue → fewer tax policy decisions), the damage from uninformed tax proposals would be structurally contained.