Objection: Land value can't be assessed accurately

The Objection

A land value tax requires assessing land separately from the buildings on it. Critics argue this is impractical: in dense cities, land rarely sells unimproved, so there are few "pure land" sales to benchmark against. Inaccurate or arbitrary assessments would make the tax unfair and legally vulnerable.

The Response, with Evidence

The objection is serious but is generally treated as a solvable engineering problem, not a fundamental barrier.

1. It is already done at scale. Estonia (pure land tax since 1993), Denmark's grundskyld, and several Australian states assess land values separately as routine practice. Most US assessments already publish a land/improvement split. The US Federal Housing Finance Agency has since turned millions of appraisals into land prices and land shares for every county, ZIP code and census tract — a demonstration that land can be valued separately from structures at national scale.

2. Land is easier to assess than buildings, not harder. In Does Georgism Work? Part 3 (2022), Lars Doucet makes the key counter-intuitive point: land value is spatially smooth — neighbouring parcels have similar land values — whereas building value varies house-by-house. That smoothness means location-based methods (comparable vacant-lot sales, teardown sales, and the "residual" of sale price minus depreciated structure cost) can estimate land value with accuracy comparable to or better than whole-property assessment. Career assessor Ted Gwartney's practitioner account of estimating land values — the expert Doucet consulted — describes exactly these location-based methods as routine professional practice.

3. Modern methods exist — including where data is thin. Computer-assisted mass appraisal (CAMA), hedonic regression, and cooperative-game approaches such as the Shapley-value land/building separation (Humanities and Social Sciences Communications, 2020) formalise the split. And the "it only works with rich sales data" version of the objection is answered directly: Bencure et al. (2019) built an expert-weighted (AHP) land-valuation model for sub-urban Baybay City, Philippines, that explained ~67% of land-value variability and outperformed ordinary multiple regression — a developing-country case with sparse market data.

3a. Even dense cities generate pure-land sales. The premise that unimproved land "rarely sells" in built-up cores is overstated: Barr, Smith & Kulkarni constructed a 64-year (1950–2014) land-value index for Manhattan — one of the densest markets on earth — using vacant-land transactions alone.

3b. Land-only ratio studies are standard professional practice, not an ad hoc Georgist workaround. The IAAO's Standard on Ratio Studies — the assessing profession's own quality-assurance standard, read directly for the wiki rather than at second hand — explicitly instructs that "for a ratio study of vacant land, the numerator in the ratio should reflect only the appraised value of the land," and sets published accuracy-tolerance bands (coefficient of dispersion, 5.0–25.0) specifically for vacant land. Land-only assessment testing is not exotic; it is a named, standardized procedure.

3c. The attempt to formalize land valuation shows up worldwide, not just in a few rich, English-speaking jurisdictions. A synthesis of six single-country case studies — Croatia, Turkey, Lithuania, Spain, China, and Quebec — drawn from Doucet's own citation list shows regression, fuzzy-logic, multi-criteria, and machine-learning valuation methods being applied independently across unrelated legal and academic traditions. Individually most of these studies are thin (several only accessible at abstract or bibliographic level), so this is evidence of breadth of attempt, not confirmed accuracy in any one country — and the Lithuanian case in that set is itself a caution (see Limits, below).

3d. Direct empirical tests find CAMA land models work even in neighborhoods with no vacant-land sales, provided the wider model has some. Gloudemans's two Lincoln Institute working papers (2000, 2002) tested combined vacant-and-improved-sales CAMA models against three North American residential datasets — including Jefferson County, Colorado, where only 4.5% of sales were vacant lots — and found that adding vacant sales barely changed predictive accuracy for improved properties (Clareview's coefficient of dispersion rose only from 5.87 to 5.97; Jefferson County's actually fell from 5.75 to 5.55) while substantially improving the plausibility of the land/building split. Gloudemans's own conclusion: "Modern mass appraisal methods are capable of producing reasonable estimates of the value of land as if vacant even in neighborhoods with no or few vacant land sales, provided there are other neighborhoods in the model with adequate vacant land sales to provide reality checks."

4. Errors are bounded and contestable. Because land is immobile and visible, gross mis-assessment is easy to challenge; assessment quality improves with investment (the focus of the Center for Land Economics, co-founded by Doucet). Independent evidence bears this out: hedonic decomposition of over a million home sales separates land from structure at scale (Kuminoff & Pope), and Australian mass-appraisal appeal rates run well below the internationally accepted 2% benchmark (Hefferan & Boyd) — revealed taxpayer acceptance that value-based land assessment is treated as reliable.

Limits and Caveats

• The profession's own quality bands tolerate more dispersion for vacant land than for improved residential property, not less. The IAAO Standard on Ratio Studies sets an acceptable coefficient-of-dispersion range of 5.0–25.0 for vacant land against 5.0–10.0/15.0 for improved single-family residential — the opposite of what a naive reading of "land is spatially smooth, therefore easier" would predict. The more likely explanation is thinner vacant-land sales samples (fewer transactions mechanically widen dispersion measures), not that land is intrinsically harder to value than buildings — but the wiki should carry this as an open nuance rather than only citing evidence for the "land is easier" framing.
• At least one real mass-valuation system has badly mistracked market value. In the Lithuanian case inside the six-country synthesis, three defensible valuation methods for the same Vilnius parcel diverged by a factor of 3.4×, and the zone-based mass-valuation approach then used for Lithuania's actual tax base captured only about 30% of the site's market value — the study's own authors call this a fairness and revenue problem, not a solved one. This is genuine counter-evidence, not merely an implementation footnote, and should temper any claim that mass appraisal reliably converges on one right answer.
• Gloudemans's assessment-feasibility findings are single-author, Lincoln Institute-funded working papers, not peer-reviewed, and carry their own distributional caution. Both papers state on their cover page that their findings "are not subject to detailed review." And the 2000 paper's own tax-shift analysis found land's share of value was strongly correlated with smaller, older, lower-quality, lower-value homes — concluding that "implementation of a land value tax would be less advantageous to lower-value than higher-value residential properties" under a simple flat-rate design, a distributional finding distinct from the assessability question itself.

Net Assessment

Assessment difficulty raises the cost and care of implementing LVT well; it does not undermine the case. The objection is best read as an implementation agenda, not a refutation — though the Lithuanian and COD-tolerance evidence above shows the agenda is still open in places, not merely a matter of scaling up known-good methods.

See Also

• "LVT Enables Corrupt Assessors and Will Be Gamed" — the companion political-economy objection: not whether land value can be measured, but whether the measurer can be trusted
• Domesday Book (1086) — an 11th-century nationwide land-value survey showing large-scale land assessment is not a modern invention requiring modern tools.
• Johannesburg Site Value Rating — a documented site-value-only tax base run at municipal scale from 1918/1919, cited by Fred Harrison as evidence against the claim that isolating land value is impossible
• Ted Gwartney — career assessor whose British Columbia and Connecticut work is cited as practical counter-evidence to this objection
• Kuminoff & Pope — tract-level hedonic land/structure separation at million-transaction scale
• Kolbe et al.: Berlin land-value appraisal — the accuracy evidence, first-hand · Almy: OECD valuation survey — the cost evidence · Hefferan & Boyd: Australian mass appraisal — revealed taxpayer acceptance
• Bencure et al.: iLVM, Baybay City, Philippines — the developing-country / thin-data case · Barr, Smith & Kulkarni: What's Manhattan Worth? — vacant-land sales in the densest US market
• Gloudemans: LVT Assessment Feasibility — direct empirical test of CAMA land models in vacant-sale-poor neighborhoods, read in full
• IAAO Standards: Ratio Studies & Property Tax Policy — the professional standard defining COD/PRD/PRB and land-only ratio studies · Mass Appraisal International Cases — the six-country breadth synthesis, honestly graded
• Land Value Tax · Estonia
• Mass Appraisal Methods (CAMA, Hedonic Regression, Land/Building Separation)
• Narrative: The Tax You Can't Dodge — where under-assessment (not evasion) is the real leak

Sources

1. Lars Doucet (2022), "Does Georgism Work? Part 3: Can Unimproved Land Value Be Accurately Assessed?", Astral Codex Ten — wiki summary · original — used for the evidence that mass-appraisal methods do separate land from improvements accurately in practice.
2. Kumar et al. (2020), "Land and building separation based on Shapley values," Humanities and Social Sciences Communications (Nature). Article — used for a formal method of decomposing property value into land and building components.
3. Lincoln Institute of Land Policy — assessment research. wiki summary — used for the body of assessment research showing land-value estimation is a solved practical problem in many jurisdictions.
4. International Association of Assessing Officers (2013), Standard on Ratio Studies. wiki summary · PDF — used for the land-only ratio-study instruction and the vacant-land coefficient-of-dispersion tolerance range cited in Limits and Caveats.
5. Saulius Raslanas, Edmundas Kazimieras Zavadskas, Artūras Kaklauskas & Arūnas Remigijus Zabulėnas (2010), "Land Value Tax... Part II: ...the Case of Vilnius," International Journal of Strategic Property Management 14(2). wiki summary · PDF — used for the Lithuanian three-method valuation divergence cited in Limits and Caveats.
6. Robert J. Gloudemans (2000, 2002), Implementing a Land Value Tax in Urban Residential Communities and An Empirical Analysis of the Incidence of Location on Land and Building Values, Lincoln Institute of Land Policy Working Papers WP00RG1/WP02RG1 — wiki summary, both read in full — used for the direct empirical test of CAMA land-value estimation in neighborhoods without vacant-land sales, and for the working-paper/funding-provenance and distributional caveats in Limits and Caveats.