Harberger Tax (COST)

Definition

A Harberger tax, also called a Common Ownership Self-Assessed Tax (COST), requires owners to self-assess the value of an asset, pay a tax on that value, and stand ready to sell at the declared price to any buyer. This neatly solves the assessment problem (owners reveal true values) and the holdout problem (assets are continuously for sale), while taxing the value of holding the asset.

Relationship to Georgism

COST is a generalisation of the Georgist idea: it captures the holding value of assets — most powerfully land — and discourages unproductive hoarding. Applied to land, it functions like a self-assessed land value tax that also keeps sites liquid. It directly addresses the objection that land can't be assessed.

Origins and Reception

Named after economist Arnold Harberger, the mechanism was popularised by Eric Posner and Glen Weyl in Radical Markets (2018). As the wiki's book page details, COST is a deliberate hybrid of Henry George's land value tax and Harberger's 1962 self-assessment proposal — owners declare their property's value publicly, pay tax on the declaration, and must sell to any bidder at that price — and it departs from George by extending the tax to all property, capital included, not just land.

It has been influential in technology and crypto circles — Ethereum's Vitalik Buterin endorsed it as a tool for property and digital-asset reform, exploring applications to domain names and other scarce digital resources; that endorsement is part of why the modern Georgist revival has strong roots in the tech world. COST-style designs also appear among the candidate instruments for taxing platform and data rents.

Mechanism-Design Kin

The self-assessment idea recurs across the Georgist mechanism-design tradition:

• Sun Yat-sen proposed that Chinese landowners self-declare land values, with the state taxing those values and capturing the increment — an early-20th-century precursor of the same self-assessment logic.
• Tideman and Plassmann's land-assembly mechanisms use auctions and self-assessed values to defeat the holdout problem, a friction closely related to the one COST's continuously-for-sale rule dissolves.

Evidence

How much do we actually know about whether self-assessed taxation works? The honest answer is: the theory is well developed, the field record is thin, and the most COST-like feature — the standing obligation to sell to any bidder — has never been implemented at scale. The evidence should be graded accordingly.

Theory and mechanism design (strong). The self-assessment idea is old — Arnold Harberger proposed it for Latin American property taxation in the 1960s — and the modern formalisation is robust. E. Glen Weyl and Anthony Lee Zhang's "Depreciating Licenses" (American Economic Journal: Economic Policy, 2022) models the core trade-off directly: perpetual ownership gives owners strong incentives to invest in an asset but "impede[s] efficient reallocation of resources to higher-valued entrants," while a depreciating license "owner annually announces a valuation at which she commits to sell" and pays a fee proportional to it, producing "time-stationary investment incentives" while keeping the asset allocable.[3] The key result is that the optimal tax rate is strictly positive but less than full — i.e. partial common ownership, not the 100% turnover rate a naive reading of COST implies. Earlier peer-reviewed variants (e.g. Dieter Gstach's self-assessment mechanism in Metroeconomica, 2009, where the authority commits to buy some declared parcels to induce truthful reporting) reach compatible conclusions.[4] This is genuine mechanism-design theory, not just advocacy.

Field precedent (partial, one long-running case, and the performance record is weaker than the enforcement design suggests). The closest thing to a real-world test is not a modern COST pilot but Taiwan's self-assessment regime, rooted in Sun Yat-sen's "equalization of land rights." Under the Equalization of Land Rights Act, owners self-declare their land value, and — crucially — the state carries a purchase right against under-declaration: "If the land value declared by the owner is lower than 80 percent of the announced land values, the government reserves the right to purchase his land at the announced land value."[5] This is a live, decades-old instance of the COST enforcement logic (name your price, and the state may take it at that price). But it is a weak form: the purchase right is a one-directional floor against low-balling, anchored to an official "announced value," not the continuous, symmetric, must-sell-to-any-private-buyer market that COST envisions. It disciplines self-assessment; it does not turn land into a perpetually contestable auction.

A Lincoln Institute of Land Policy working paper tracing the regime's actual operation from the 1950s to the 1990s (Alven Lam & Steve Tsui, 1998) shows the "self-declared" element eroding over time in favor of a government-set benchmark: an owner-self-declaration-plus-purchase-right design (the 1964 reform's "if self-declared value was 20 percent lower than the publicly declared value, govt. can purchase the land") was progressively superseded from 1968 onward by a wholly publicly declared value — a government-assessed benchmark reassessed periodically — that became the tax base "applied to the whole country" by the 1977 reform.[6] Where the mechanism could be measured, its performance was well short of statutory ambition: comparing assessed-value increments against actual Land Value Increment Tax (LVIT) collections in Taipei City from 1979–1993, the paper finds LVIT captured roughly 32 percent of the assessed-value increment against progressive statutory rates that reached 40–100 percent, and — once the gap between assessed and true market value (assessed value was "approximately 50 percent of market values" in earlier years, rising to roughly 80 percent by 1996) is factored in — estimates the effective share of real market-value gains actually captured at under 20–24 percent.[6] The paper attributes the shortfall to assessment lag (annual or triennial reassessment cycles that speculators could sell inside, escaping the tax entirely) and weak local fiscal incentives to set assessments aggressively, not to owner dishonesty as such — but the net finding is that Taiwan's declared-value system, even backed by a nominal purchase right, chronically undervalued land relative to market price and captured only a fraction of the land-value gains the law aimed at. This is genuine, if discouraging, field evidence on the practical limits of a declared-value mechanism once it is administered as a periodic government benchmark rather than a continuously-contestable owner declaration — the gap COST's forced-sale-to-any-bidder design is meant to close, but which no implementation has tested at scale.

What is missing (the honest gap). There is no controlled field trial of a full Harberger tax on land, and — as the wiki's Radical Markets page notes — direct behavioral evidence on how accurately real owners self-assess under a genuine must-sell rule, and on how much the forced-sale "anxiety cost" of losing a home one values above its price would deter participation, remains thin. Buterin's own sympathetic treatment flags exactly this: the mechanism trades allocative efficiency against owners' security of possession, and where that trade lands for homes (as opposed to spectrum, domain names, or other fungible licenses) is an empirical question the literature has not settled.[7] Grade: strong theory, one partial institutional precedent whose measured performance was weak, essentially no field or large-sample experimental evidence for COST as such — most confident for fungible, low-attachment assets (licenses, digital resources), least tested for owner-occupied land.

An academic modeling paper on Sun Yat-sen's original scheme specifically — Emerson M. S. Niou & Guofu Tan, "An Analysis of Dr. Sun Yat-sen's Self-Assessment Scheme for Land Taxation," Public Choice 78(1), 1994, pp. 103–114 — exists in the peer-reviewed literature (confirmed via Springer's DOI record, IDEAS/RePEc, and both authors' faculty publication lists) but its full text and abstract were paywalled with no free mirror locatable in this session (Springer requires institutional login even for the abstract; SSRN, ResearchGate, and Academia.edu copies all returned access-denied); its content is not cited here for that reason — a candidate for the Hermes unblocked-web channel.

Role in Answering Objections

• Against the valuation strand of the Austrian critique, COST shows owners can be made to reveal values directly, so assessment need not depend on state estimation.
• Against the public-choice critique, self-assessment is cited as a discretion-removing design: it takes valuation out of assessors' and politicians' hands.

See Also

• Weyl & Zhang (2017): Depreciating Licenses — the academic paper formalizing the investment-vs-allocative-efficiency tradeoff that underpins the COST/Harberger-tax design
• Arnold Harberger — the Chicago economist whose 1962 self-assessment proposal is the namesake and origin of this mechanism
• Radical Markets — the broader Posner & Weyl program this mechanism comes from
• Posner & Weyl, Radical Markets (book page) — the full COST design, lineage, and departures from George
• Land Value Tax · Land Monopoly · Holdout Problem

Sources

1. Eric Posner & Glen Weyl (2018), Radical Markets, Princeton University Press — used for the COST design and its Georgist lineage (A-claims; Deep scan on the book page). Publisher
2. Vitalik Buterin (2018), "On Radical Markets" — used for the sympathetic-critique reception (C-claims). wiki summary · Essay
3. E. Glen Weyl & Anthony Lee Zhang, "Depreciating Licenses," American Economic Journal: Economic Policy 14(3), 2022. DOI 10.1257/pol.20200426 · working-paper text PDF (Stanford) — mechanism-design theory; used (verified against the working-paper text) for the investment-vs-reallocation trade-off, the depreciating-license design ("owner annually announces a valuation at which she commits to sell"), and the result that the optimal fee corresponds to partial rather than full common ownership. (B-claim; theory.)
4. Dieter Gstach, "A Property Taxation Mechanism With Self-Assessment," Metroeconomica 60(3), 2009, pp. 400–408. IDEAS/RePEc — peer-reviewed self-assessment variant in which the tax authority commits to purchase some declared parcels to induce truthful reporting; cited as compatible theoretical support. (B-claim; theory.)
5. The Equalization of Land Rights Act (Republic of China / Taiwan), Article 16, Laws & Regulations Database of the Republic of China. law.moj.gov.tw — used (verified quotation) for the self-declaration regime and the government's purchase right where the owner's declared land value falls below 80% of the announced value; the closest live field precedent for the COST enforcement mechanism, and its limits. (A-claim; primary legal text.)
6. Alven H.S. Lam & Steve Wei-cho Tsui, "Policies and Mechanisms on Land Value Capture: Taiwan Case Study," Lincoln Institute of Land Policy Working Paper WP98AL1, 1998. PDF (lincolninst.edu) — fetched and read in full; used for the historical evolution of Taiwan's declared-land-value system (owner self-declaration with a purchase-right floor, 1964 reform, superseded by a government-set "publicly declared value" nationwide by 1977) and the measured performance of the Land Value Increment Tax: Taipei City effective capture of the assessed-value increment (~32%, 1979–1993, Table 6, sourced in turn to Taiwan Ministry of Finance and Huang 1995) against statutory rates of 40–100%, and an estimated effective capture of well under 20–24% of true market-value gains once the assessed-to-market-value gap is factored in. Working paper, not peer-reviewed ("not subject to detailed review" per its own disclaimer); author is a Lincoln Institute fellow with a Harvard doctorate who served as academic dean of Taiwan's Land Reform Training Institute — established institutional authorship, not independent peer review. (B-claim; empirical, institutional working paper.)
7. Vitalik Buterin (2018), "On Radical Markets," and the wiki's Radical Markets page — used for the honest gap: the forced-sale security-of-possession trade-off and the thinness of behavioral evidence on real-world self-assessment accuracy, especially for owner-occupied homes. Essay (C-claim.)
8. Emerson M. S. Niou & Guofu Tan, "An Analysis of Dr. Sun Yat-sen's Self-Assessment Scheme for Land Taxation," Public Choice 78(1), 1994, pp. 103–114, doi:10.1007/BF01053368 — a peer-reviewed formal analysis of Sun's original self-assessment-plus-purchase-right scheme; bibliographic existence confirmed (Springer DOI record, IDEAS/RePEc listing, both authors' faculty publication pages at Duke and USC), but full text and abstract were inaccessible in this session — Springer requires institutional login even for the abstract, and no free mirror was found on SSRN, ResearchGate, Academia.edu, or the authors' own pages (each attempt returned 403 or "not available for download"). Not cited for any specific finding; flagged for the Hermes unblocked-web channel. (Paywalled; access not verified.)