The latest reveal from Extropic:
My prior comment stands only with more emphasis than ever. Guys like Musk and Verdon must at some level realize they need a lot more guys like themselves but they apparently have less interest in cultivating them than in being them.
The gap between where they are and understanding why Militia.Money would cultivate guys like them is why I spend most of my time working on bootstrapping Hume’s Guillotine with my meager resources. There may be hope for them if they get just a bit more ruthless about their own stated principles. As I stated in Andrew Edwards’s latest Warhorse interview, I’m doing my best to avoid going to war against them in order to produce more men like them.
I had a conversation with GPT-o4-mini regarding the basis of Militia.Money’s demurrage monetary system revenue, not including how the revenue is spent to protect property rights which it would find objectionable since it would privatize the delivery of all social goods in corporations and solve TFR by being a sex-realist.
Bear in mind that there isn’t even one prominent economist in the last century that has studied the likely consequences of replacing the 16th Amendment with a tax on net assets. Friedman’s passing mention of LVT’s advantages notwithstanding:
Has any government assessed property values by bidding on property?
ChatGPT said:
Governments do sometimes use competitive bidding to establish the value of particular interests in land—but only for specific classes (e.g. state land or tax liens), not as a universal method for assessing every private parcel.
- State‐land leases by open tender (Singapore)
The Singapore Land Authority regularly offers state land and buildings for lease via sealed‐bid tenders. Prospective tenants submit the rent-or-premium they’re willing to pay, and the highest bid wins the lease; that bid then effectively becomes the “market” rent or premium for that site sla.gov.sgsla.gov.sg.
- Land sales and lease premiums (Hong Kong)
Under Hong Kong’s leasehold system, almost all government land is granted by public auction or open tender. Bidders compete on the land premium (a lump-sum payment); the winning bid both conveys the lease and signals the government’s valuation of that parcel landsd.gov.hkLincoln Institute of Land Policy.
- Tax‐lien auctions (United States)
Many U.S. jurisdictions auction property-tax liens on delinquent parcels. Investors bid either the lowest interest rate they’ll accept (“bid-down” method) or the highest cash premium over the lien amount. The winning bid determines who holds the lien—and reflects the market’s valuation of that tax debt Wikipedia.
Note: in each case, the bidding applies to public interests (state leases or tax liens), not to routine annual assessments of privately held property on the tax rolls. No government has ever adopted a system whereby all private parcels are routinely valued by open bidding.
You said:
singapore’s “state-land lease” offers state land “AND BUILDINGS” for lease. why then is it called state-land lease?
ChatGPT said:
In Singapore law, “State land” is a legal term covering not just the bare soil but all State-owned property affixed to it—i.e. the land and any buildings or other permanent improvements on that land. When the SLA grants a lease of State land, it’s automatically a lease of the entire “bundle of rights” in that parcel: the soil, the strata beneath it, and the structures erected by or for the State.
You’ll see this reflected in two key places:
- SLA’s own mandate: “SLA oversees the management of 11,000 hectares of state land and 2,700 state properties,” where “state properties” means buildings, facilities and other improvements that sit on State land Wikipedia.
- Lease‐renewal policy: Under common law the land and buildings revert to the lessor (the State) at lease expiry, so the SLA is entitled to charge both a land premium and a building premium—even though they sometimes waive the latter to encourage reinvestment in the buildings sla.gov.sg.
Put simply, even though the program is called a “State-land lease,” it really means “State-land and buildings lease,” because the buildings are legally treated as part of the landholding.
—
If you’d like more detail on how premiums are calculated, or on the distinction between Temporary Occupation Licences (short-term rights) vs. long-term leases, just let me know!
You said:
What is the land area of singapore?
ChatGPT said:
Singapore is one of the world’s smallest countries by land area. As of the latest surveys, its total land area is approximately 735.7 km² (about 284 sq mi) Wikipedia. Over the decades, extensive land-reclamation projects have expanded its size from roughly 580 km² in the 1960s to today’s figure.
You said:
735.7 km²-11,000 hectares?
ChatGPT said:
Let’s put both on the same footing:
- Total land area of Singapore
735.6 km² (73 560 hectares) UNFCCC
- State land under SLA management
11 000 hectares (110 km²) Ministry of Law
Remainder of Singapore’s land
- In hectares:
73 560 ha – 11 000 ha = 62 560 ha
- In square kilometres:
735.6 km² – 110 km² = 625.6 km²
So about 15 % of Singapore’s area is State land under SLA, leaving roughly 625.6 km² (62 560 ha) in other tenures.
You said:
how much of the land value is in each?
ChatGPT said:
There isn’t a single published figure that splits the aggregate market value of all Singapore land into “State-land” versus “other tenures.” However, we can at least get a rough sense of how they compare:
- By area
- State land (managed by SLA): 11 000 ha (110 km²) → ~15 % of Singapore’s total 73 560 ha.
- Other tenures (private, HDB, etc.): 62 560 ha → ~85 % of the total.
- If land value were uniform per hectare
You’d simply apportion 15 % of total land‐value to state land, and 85 % to the rest.
- Why “uniform” is a poor assumption
- State land is disproportionately prime. Major reclamation zones, the CBD, Changi, Jurong Lake District, Kallang, Punggol, Changi Airport land—all originate as State land. Their per‐hectare values are easily 3–10× the national average.
- Private tenures include lower‐value parcels. Smaller urban sites, older residential estates, outlying industrial zones, agricultural land, military areas, etc., often trade at much lower rates.
- A back-of-the-envelope estimate
- Suppose the national average land price is S$3 million/ha (a conservative middle‐of‐the‐road URA figure).
- → Total land value ≈ S$3 M × 73 560 ha ≈ S$220 billion.
- State‐land share by area (15 %) → S$33 billion.
- Other tenures (85 %) → S$187 billion.
- But if prime State land trades at S$9 M/ha (3× national avg) while the other 85 % averages S$2.5 M/ha, you get:
- State land: 11 000 ha × S$9 M/ha ≈ S$99 billion (≈ 45 % of total).
- Other tenures: 62 560 ha × S$2.5 M/ha ≈ S$156 billion (≈ 55 % of total).
How you could get a more precise split
- Download URA’s quarterly “Estimated Land Value” dataset (by planning area), from Page Not Found
- Overlay the SLA’s state-land boundaries (available via OneMap) to sum “State‐land hectares × land‐value/ha” for each planning area.
- Do the same for the complement (“other tenures”) and you’ll have an accurate, up-to-date dollar split.
Without doing that GIS + URA-data exercise, any “exact” number is necessarily an approximation. But the key takeaway is:
- By area: State land ≈ 15 % of territory
- By value: State land likely commands 30 – 50 % of total market land value, because it’s concentrated in the most sought-after districts.
You said:
when did the government’s policy of auctioning leases start?
ChatGPT said:
The idea of putting State-land leases up for competitive tender in Singapore goes right back to the late 19th century under British colonial administration:
- First reclamation auctions (c. 1886):
After reclaiming the strip that became Cecil Street, the colonial Public Works Department began leasing (essentially by tender) the newly created land to merchants in 1886. Those early leases were awarded to the highest bidders, making it the de facto start of auction‐based land grants in Singapore BiblioAsia.
- Statutory framework in 1921:
The State Lands Act of 1920 (in force from 28 November 1921) then codified how “statutory grants and leases” of State land were to be regulated. Although it didn’t invent the auction method, it provided a clear legal basis for issuing 99-year leases by tender or auction across the colony AGC Singapore.
So, while the legal foundation appeared in 1921, the practice of auctioning leases on newly reclaimed parcels began as early as 1886.
You said:
what do you mean ‘newly created land’?
ChatGPT said:
By “newly created land” I meant parcels that didn’t naturally exist as dry ground but were formed by land reclamation—that is, by filling in parts of the harbor or riverbank with earth and sand.
In the 1880s the colonial government dredged and infilled the north shore of the Singapore River (where today you have Cecil Street, Robinson Road, etc.), literally “making” new land out of the old foreshore. Those reclaimed strips had to be leased out, so the administration put them up for competitive tender—the very first auction-style land grants in Singapore.
You said:
don’t you find it somewhat ironic that one of the main reasons economists claim land value tax suffers no dead weight loss is that the market is perfectly inelastic, and that one of the few places where land value tax is de facto in effect is land that was created based on demand for it?
ChatGPT said:
It is an interesting tension, but I wouldn’t call it a fatal irony—more a reminder of the distinction between “short-run” inelasticity and “long-run” supply responsiveness.
1. Why land is “perfectly inelastic” in most LVT models
- Fixed geography. Classical land-value–tax theory assumes that, in the short run, you cannot create more land: every hectare that exists today already exists, so a tax on its unimproved value doesn’t distort location or intensity decisions.
- No substitution. If you tax land value, owners can’t “shift” that tax onto suppliers of anything—there is no “input” you can replace land with—so there’s (in theory) no dead-weight loss.
2. Reclamation as “long-run” supply expansion
- Reclaiming land is costly and time-consuming. Turning sea into city isn’t like minting more coins: you need massive dredging, environmental approvals, engineering, and years of work. That keeps supply quite inelastic in any practical policymaking timeframe.
- Still depletable. Even if you can reclaim, there’s a finite envelope of shallow coastal zones; it’s not an open-ended ability to create endless new land. So while supply isn’t mathematically “fixed,” it remains highly inelastic relative to most taxed goods.
3. Singapore’s “de facto LVT” and reclaimed land
- In effect, when Singapore auctions leases on reclaimed parcels, the winning “bid” captures much of the incremental scarcity value that reclamation projects generate. That’s very much in the spirit of a land-value tax: capturing unearned gains for the public.
- But because reclamation itself is driven by demand (and managed by the state), you could argue that supply does respond to price signals—just on a multi-year, multi-billion-dollar scale rather than a day-to-day market.
4. Does that undercut the “no dead-weight loss” claim?
- Not in practice. Dead-weight loss arises when taxes change behavior at the margin. As long as the annual supply of land that can change hands remains essentially fixed, taxing its value doesn’t discourage productive use of existing plots.
- Long-run vs. short-run elasticity. Yes, in the long run you can bring new land stock online, but policymakers factor that in via explicit reclamation planning and pricing. It doesn’t invalidate the LVT principle so much as shift the elasticity-analysis window from months or years into decades.
Bottom line:
Land-value tax relies on practical inelasticity of land supply within the policy horizon. Reclamation shows that supply can shift—but only at enormous cost and lead time—so the “perfect inelasticity” assumption remains a very good approximation for most tax-policy purposes.
You said:
what is the history of attempting to factor out the portion of asset value that is due to the private capture of positive network externalities of civilization other than land ownership?
ChatGPT said:
Economists and policymakers have long wrestled with how to separate—and sometimes recapture—the portion of an asset’s value that arises not from the owner’s efforts or improvements, but from positive externalities created by the community, infrastructure, institutions, or network effects. Here’s a rough chronology of major intellectual and policy efforts beyond the classic Georgist focus on land:
1. Early Recognition of External Economies (late 19th–early 20th c.)
- Alfred Marshall (1890) introduced the distinction between internal economies (firm-specific cost savings) and external economies (cost or demand benefits accruing to all firms in a locality or industry), laying the groundwork for thinking about “social” value embedded in location or clustering—what we’d now call network or agglomeration externalities Wikipedia.
- Arthur Pigou (1920) built on Marshall in The Economics of Welfare, arguing that positive externalities (e.g. knowledge spillovers, infrastructure) could justify subsidies—essentially a mirror image of Pigouvian taxes on negative externalities Wikipedia.
2. “Unearned Increment” and Betterment Levies (early 20th c.)
- John Stuart Mill and later John Rae recognized that part of urban land’s rising value came from public investments (roads, sewage, schools), coining the idea of the “unearned increment.”
- London’s 1909 People’s Budget introduced betterment duties on land gains following new infrastructure (e.g. trams, parks). These levies were an early attempt to tax the “community-created” portion of property value .
- Argentina and Australia in the 1920s–30s experimented with “value capture”—special assessments on parcels that appreciated due to public works.
3. Mid-Century Public Capital and Growth Accounting (1950s–1970s)
- Richard Musgrave and James Meade formalized public good theory, showing how public capital (roads, ports, education) raises private returns. They advocated mechanisms to ensure users (or beneficiaries) share in the cost, foreshadowing modern infrastructure levies.
- Aschauer (1989) and Jorgenson & Griliches quantified the social rate of return on public investment, effectively estimating how much of aggregate capital value stems from publicly provided infrastructure rather than private enterprise Investopedia.
- Tiebout (1956) and Hirschman (1958) explored how local public goods and social spillovers shape urban growth and thus property values.
4. Network Externalities and Intangible Capital (1980s–2000s)
- Katz & Shapiro (1985) and Arthur (1990) formalized network externalities—where the value of an asset (e.g. software, telephone service) increases with the number of users. This spawned debates on how to tax or regulate tech monopolies that privately capture widespread social gains Wikipediaifo Institute.
- OECD’s Oslo Manual (1992, 2005) and researchers like Corrado & Hulten developed frameworks for intangible capital accounting, splitting corporate value into physical assets, human capital, and “structural capital” (including organizational networks, brand, and institutional environment). This was a systematic attempt to measure and potentially tax the non‐land externalities embedded in firm valuations.
5. Modern Value-Capture Mechanisms (1990s–present)
- Tax Increment Financing (TIF) in the U.S. (since the 1950s) lets municipalities borrow against future gains in property‐tax revenue generated by new development—effectively capturing part of the uplift from public improvements.
- Special Assessment Districts or Community Improvement Districts (e.g. in Denver, Atlanta) impose extra levies on businesses or property owners near new transit lines or stadiums, aligning costs and benefits of the public investment.
- Infrastructure-based land auctions (e.g. Stockholm’s congestion-pricing lease auctions, Seoul’s development rights auctions) directly allocate and price the value created by new roads, subways, or zoning changes.
Why no universal “externality capture” tax beyond land?
- Measurement challenges. Unlike unimproved land, externalities from networks and institutions are diffuse, dynamic, and hard to parcel out per asset.
- Political economy. Businesses and industries wield influence to limit any levy on their intangible gains.
- Administrative complexity. Assessing and auditing “knowledge spillover value” or “infrastructure‐driven premium” at scale is daunting compared to mapping land parcels.
In sum, from Marshall’s external economies through Pigou’s welfare economics, 20th-century public-capital research, network-externalities theory, and modern TIFs or betterment levies, there’s a long lineage of trying to isolate—and sometimes tax—the value private owners capture from collective, civilization-wide benefits. However, practical barriers mean that, beyond land, few jurisdictions have implemented broad‐based mechanisms to “factor out” and recapture externality‐driven value at anything approaching scale.
You said:
Since MPT’s calculation of asset value depends on the so-called “Risk Free Interest Rate” – in particular as it manifests in the net present value calculation used by financial institutions not directly involved in asset management – doesn’t it make sense to use something like the aggregate 30-year government debt rate (not the current market but rather the effective average rate across all 30-year debt instruments the government must service) as the tax rate on the current high bid in escrow for each property right recognized by the government?
ChatGPT said:
You’re essentially proposing to turn each property’s “high‐bid” value into a kind of capitalized stream—taxed at the government’s own long‐term borrowing rate—much like how Modern Portfolio Theory (or any DCF model) uses a risk-free discount rate to value future cash flows. It’s a neat idea, and in fact it mirrors the capitalization rate that real-estate investors already use:
Value of an asset ≃ Net cash flow ÷ cap rate
—and the cap-rate is often benchmarked to a long-term treasury yield plus a risk premium.
Why tying your “land-tax rate” to the 30-year yield makes sense
- Consistency with finance theory. In theory, the risk-free rate is what you’d use to discount a perfectly safe cash flow. If you treat land-rental—or in your auction model the “imputed rent” from a lease bid—as a safe stream, then taxing at the same rate (or at least pegging the mill rate to it) aligns the government’s revenue claim with the asset’s financial valuation.
- Automatic stabilization. When interest rates rise, asset values fall (all else equal), so a tax rate indexed to long-term yields would rise in tandem with the government’s borrowing costs and fall when rates drop—keeping revenue roughly level against market swings.
- Transparency. Using a public, observable benchmark (the average yield on outstanding 30-yr debt) is more transparent than the patchwork of millage rates and assessor “equalization factors” we see today.
But there are some important caveats
| Issue | Why it matters |
|–––|–––|
| Risk-free ≠ risk-adjusted | Most property values embed location risk, development risk, or simply liquidity risk. A 30-yr U.S. Treasury yield (say ~4 %) is far below the cap-rates investors demand on even the safest office towers (often 5 –6 %). If you tax purely at 4 %, you’ll systematically under-tax higher-risk parcels and over-tax the safest. |
| Rate volatility | Treasury yields can swing by 1 –2 % in a year. If your tax rate jumped from 4 % to 6 % between assessments, that’s a 50 % hike in someone’s property tax bill—potentially bankrupting leveraged owners. Governments typically smooth mill rates to avoid that volatility. |
| Administrative lag | The “current high bid in escrow” might only update when parcels change hands or when leases roll over—perhaps every 30 or 99 years in a leasehold system. Mismatches between when you re-auction or re-bid and when you reset the tax rate could leave long gaps where your yield-indexed rate no longer reflects market reality. |
| Political feasibility | Taxpayers bridle at any link between government debt costs and their bills. Even neutral-rate indexing can become a lightning rod in political debates over “taxing treasury yields.” |
A middle path: a weighted cap-rate surcharge
Most real-estate tax reform proposals that look to finance theory instead start with a long-term Treasury yield and then add a small, fixed “risk surcharge”:
Land-tax rate = Treasury 30-yr yield + α (average risk premium)
Where α might be 1 – 2 % depending on property type. That approach captures your spirit—tying the base rate to the government’s own cost of capital—while still recognizing that private property isn’t as safe or liquid as U.S. debt.
Bottom line
- Yes, anchoring a land-tax rate to the government’s 30-year borrowing cost is conceptually elegant and would align tax policy with modern financial valuation.
- But you’d need to build in risk adjustments, volatility buffers, and administrative timing rules to avoid wild swings and mismatches.
If you’re designing a Georgist-inspired “self-assessment with penalty” system, starting with the 30-year yield as your base cap-rate and then layering on a modest risk-premium surcharge is probably the most practical way to marry MPT’s risk-free rate with the real-world heterogeneity of property.
You said:
Regarding volatility, I did go out of my way to say "(not the current market but rather the effective average rate across all 30-year debt instruments the government must service) ". Regarding risk-free vs risk-adjusted, risk management is an essential aspect of asset management and is hence a function of ownership. Administrative lag doesn’t apply if anyone can bid on any property at any time and the only time a transaction occurs is when the owner accepts the high bid in escrow. As for “taxing treasury yields” if the tax is principled, as in the governments obligation to, in essence, store the property right under legal protection, it seems we’re simply talking about the basis for a sound demurrage currency.
ChatGPT said:
You’re right—framing the levy as a demurrage on the “property‐token” that the government must safeguard makes the Treasury’s own funding cost the natural benchmark. Let me pull the threads together:
1. Demurrage as the time‐value fee for “storing” property rights
If we view a property title like a currency token, the State acts as the “vault” or “ledger keeper” and must backstop legal, administrative, and enforcement functions in perpetuity. Charging a demurrage rate:
Demurrage fee=rgov×property value\text{Demurrage fee} = r_{\text{gov}} \times \text{property value}Demurrage fee=rgov×property value
where rgovr_{\text{gov}}rgov is the government’s effective average cost of 30-yr debt, simply means the owner pays exactly what it costs the Treasury to carry a matching liability. Any excess bid value over the owner’s own cost of capital becomes public revenue.
2. Why use the average yield, not spot rates
- Smooths out spikes.
By weighting across all outstanding 30-year bonds—say,ravg=∑i(interesti)∑i(principali) r_{\text{avg}} = \frac{\sum_i \bigl(\text{interest}_i\bigr)}{\sum_i \bigl(\text{principal}_i\bigr)}ravg=∑i(principali)∑i(interesti)—you mute one-off market gyrations and ensure the demurrage rate only moves as the debt portfolio itself evolves.
- Aligns incentives.
When the Treasury refinances at higher rates, the demurrage levy rises automatically—giving the government revenue to offset its higher service costs. If yields fall, owners pay less.
3. Owner‐borne risk versus Treasury’s “risk‐free” cost
You’re correct that risk‐management is part of ownership. By decoupling the demurrage fee from an individual parcel’s risk profile, you:
- Keep the tax neutral—it doesn’t distort location or leverage decisions, since everyone faces the same base rate.
- Incentivize owners to hedge or leverage according to their own return‐requirements. If their cost of capital is above ravgr_{\text{avg}}ravg, they’ll only hold parcels that earn them a spread.
Any project‐specific risk premium belongs in their financing structure (debt vs. equity), not in the public demurrage.
4. Continuous bidding removes administrative lag
Because anyone may bid at any time, and the owner only sells when they accept the bid, the system:
- Treats the property as a perpetual option rather than a 1-time reappraisal event.
- Ensures the current market price is always discoverable, so the demurrage base (the high bid in escrow) is up-to-date by construction.
5. Putting it all together: A “Demurrage‐Indexed Land Rights” regime
- Each parcel has a standing high‐bid PtP_tPt in escrow.
- Monthly or quarterly, the owner paysΔtax=Pt × ravg \Delta \text{tax} = P_t ;\times; r_{\text{avg}}Δtax=Pt×ravgwhich funds the Treasury’s debt service.
- If a new bidder offers Pt′>PtP_{t’} > P_tPt′>Pt, that becomes the new base—and the former owner realizes the capital gain.
- If an owner lets the bid expire without matching it, the State acquires the right and can re‐auction, pocketing the spread.
Next steps
- Data: You’d need the Treasury’s annual interest outlay and average debt stock to compute ravgr_{\text{avg}}ravg.
- Simulation: We could model historical 30-yr yields vs. property‐value indexes to see how stable your demurrage levy would have been.
- Legal design: Drafting the precise “high‐bid escrow” mechanism and ensuring due‐process protections.
Does that capture your idea? Would you like to dive into a historical yield‐based simulation or explore how the escrow mechanics could work in practice?
