Land rent could fund a large share of government
Estimates of total land rent suggest it could fund a substantial fraction — by some accounts most — of government, though figures are sensitive to assumptions.
At a glance — Estimates of how much government land rent could fund vary widely with method and scope — from a major share of spending to near-total replacement of existing taxes — so the honest reading is a large but genuinely uncertain fraction. Evidence: Contested (estimates vary widely with method and scope) · 10 supporting sources · 0 challenging Strongest support: Larson (2015) — the BEA study puts contiguous-US land at roughly $23 trillion (2009), the wiki's measured benchmark. No structural counter-evidence is currently wired; see Limits.
The Claim
The total economic rent of land and natural resources is large enough to fund a substantial portion of government spending, potentially replacing many taxes on labour and capital.
The Evidence and the Debate
| View | Argument | Source |
|---|---|---|
| Optimistic | US land rent is "enough and to spare" once hidden rent categories and the ATCOR base-expansion effect are counted | Gaffney (2009) |
| Formal mechanism | Algebraic tax-capitalization model showing untaxing buildings raises the land-tax base by more than the abated revenue (timing, excess-burden, and synergy effects each add to the rise) | Gaffney (1998) |
| Advocate aggregation | US land/resource rent ≈ $3.5T/yr, over a third of GDP, via two back-of-envelope methods — an attributed advocate estimate, not an official figure | Jeffery J. Smith, Counting Bounty (2020), ch. 20 |
| Advocate extrapolation | ~$11T US land value from Federal Reserve Z.1 balance-sheet tables (partial-sector; also documents the Fed's famous negative-land artifact from the land-residual method) — attributed; Larson 2015 remains the wiki's benchmark | Scott Baker, progress.org (2016), from Fed Z.1 |
| Conservative | Currently measured land rent is a smaller share of GDP — enough for a major but not total replacement of existing taxes | mainstream national-accounts estimates |
What each side is counting is clarified by Foldvary's kinetic/potential rent decomposition: measured national-accounts rent is part of kinetic rent; the optimistic case claims potential rent (including tax-suppressed rent) far exceeds it.
The gap turns on two things: what counts as rent (Gaffney includes resource, spectrum, and under-assessed urban rent that standard accounts omit) and whether one credits the ATCOR effect — that abolishing other taxes raises land values, enlarging the LVT base.
Three of the older advocate data points that reach Lars Doucet's twelve-method aggregate-land-value survey through Smith's Counting Bounty are worth pinning to source. Steven Cord (How Much Revenue Would a Full Land Value Tax Yield?, American Journal of Economics and Sociology 44(3), 1985, pp. 279–293) capitalised 1981 Census-Bureau and Federal-Reserve assessed land values — adjusted upward for under-assessment, tax-exempt land, and unassessed federal holdings — at a conservative 14%, arriving at a full-land-rent yield of roughly $658 billion, which he put at 28% of 1981 national income (nearly two-thirds of all tax revenue then levied). Counting Bounty and Doucet render this as the ~24% "Cord" line; the primary paper's own headline is 28%. Richard Ebeling (There Is No Social Security Santa Claus, Future of Freedom Foundation, 21 Dec 2015) valued the federal government's land and mineral reserves at about $5.5 trillion (Smith extrapolates this to $6.6T for 2020) — a figure Doucet flags as method-free and carries only as a bullish outlier. The lone machine-learning estimate in the span is the PLACES Lab (Christoph Nolte's group at Boston University) high-resolution fair-market-value map of US private land (Nolte 2020, PNAS 117(47):29577–29583; data at placeslab.org), which despite its very different method produces a single-year total that tracks the cost-approach cluster.
The Evidence in Detail
The measured base first. Larson (2015) — the BEA working paper — puts contiguous-US land at roughly $23 trillion (2009) using hedonic methods; Albouy, Ehrlich & Shin (2018) supply the first transaction-based index for every US metro, finding urban land worth over twice GDP, with just five metros holding 48% of it; and Davis & Heathcote (2007) built the widely-used quarterly residential-land price series the field runs on. On the institutional side, the Mirrlees Review (2011) — the UK's most authoritative modern tax review — concludes land value should be taxed and proposes replacing business rates with an LVT. The advocacy wing's national worked example is Common Wealth Canada's 2023 report: its January 2023 version estimates ~$421B/year of newly collectible rent across all resource categories, $362.5B of it from a national LVT; the publisher's July 2023 revision replaced the land method (average land-price growth) with a 5.5% capitalization rate on StatCan land values, cutting the net-new LVT figure to $194B/year. An advocate estimate, carried as such alongside the table's other attributed aggregations — and the ~46% cut between the publisher's own versions is itself a caution about how method-sensitive these aggregates are. Its two methodological templates are also on this wiki: Prosper Australia's Total Resource Rents of Australia (2013), which tallies Australian rents at 23.6% of GDP and finds rent taxation could raise 87% of all-levels government revenue, and the University of Vermont's Valuing Common Assets study (2008), which costs Vermont's uncollected common-asset rent at ~$1.2B/year — roughly 43% of the state's in-state revenue. Both are advocacy/student-project estimates, carried with the same attribution caveats.
Burgess: LVT as Alternative to Taxation
Ronald Burgess's Public Revenue Without Taxation (1993) argues that land value taxation is not merely a better tax but the legitimate replacement for taxation altogether. Burgess traces the argument from the Physiocratic tradition through Henry George, arguing that public revenue should come from the community-created rental value of land rather than from coercive taxation of productive activity (Burgess 1993, Ch. 1–3). Burgess develops the ATCOR argument — that all taxation comes out of rent — and contends that replacing existing taxes with LVT would actually expand the revenue base by unburdening production (Burgess 1993, Ch. 4). (C-claim; theoretical)
Burgess also addresses the inflation dimension, arguing that failure to collect land rent forces governments to monetize debt, creating inflation that disproportionately harms wage-earners (Burgess 1993, Ch. 5). He connects unemployment to land speculation, arguing that high land prices price labour out of productive opportunities (Burgess 1993, Ch. 6). (C-claim; theoretical)
See: Public Revenue Without Taxation (Burgess)
Strength of Evidence
Contested. That land rent is large is well established; how large relative to government depends heavily on method. Honest framing: enough to be a major revenue source, with full replacement an open question.
An attributed high-end estimate. Foldvary's 'geo-rent' projection — that land rent could fund roughly half of US government revenue at all levels — is an advocate estimate, not a consensus figure, and sits at the optimistic end of the range above.
See Also
- Jeffery Johnson Smith — author of Counting Bounty (2020), whose ~$44T US land-value estimate is the highest figure in Land is a Big Deal's comparison table
- Barr, Smith & Kulkarni (2018): What's Manhattan Worth? — a transaction-based estimate ($1.47T, 2014) of Manhattan's developable land value alone
- Steven Cord — Georgist scholar whose Census/Federal Reserve-based land-rent estimates (cited in Harrison's The Power in the Land) are among the empirical inputs to the revenue-sufficiency debate, and who led the on-the-ground campaign behind Pennsylvania's split-rate tax adoptions
- Kuminoff & Pope — hedonic land/structure separation at scale, which also disciplines the residual method's over-attribution to land · Hudson: Where Did All the Land Go? — the understatement critique of official land accounts (attributed)
- Tideman, Kumhof, Hudson & Goodhart (2021) — the macro-model tax-shift case (stub, pending full read)
- Dwyer (2003), Taxable Capacity of Australian Land — the national calculation (stub, pending full read)
- World Bank, The Changing Wealth of Nations 2021 — wealth-accounting scale evidence
- ATCOR · Single Tax · Objection: LVT can't raise enough revenue
At the measurement layer, Kuminoff & Pope separate land from structure value hedonically across a million-plus transactions — establishing that the land component is large and separable, while cautioning that residual methods over-attribute boom-bust swings to land.
Sources
- Mason Gaffney (2009), "The Hidden Taxable Capacity of Land: Enough and to Spare" — used for the estimate that land's taxable capacity is far larger than conventional assessments imply (once ATCOR/EBCOR effects are counted). wiki summary · PDF
- Counterpoint framing: Objection — LVT can't raise enough revenue — used as the steelmanned opposing view on revenue sufficiency.
- Mason Gaffney (1998), "The Philosophy of Public Finance," Ch. 7 in Fred Harrison (ed.), The Losses of Nations (Othila Press, 1998) — used for the algebraic demonstration that a revenue-neutral shift from building taxes to land taxes raises the land-tax base by more than the abated revenue (C-claim; theoretical model). wiki summary
- Ronald Burgess, Public Revenue Without Taxation (London: Shepheard-Walwyn, 1993) — used for the ATCOR revenue argument and the case that LVT replaces rather than supplements taxation (C-claims). Book page