The Value of Residential Land and Structures during the Great Housing Boom and Bust (Kuminoff & Pope)

Summary

"The Value of Residential Land and Structures during the Great Housing Boom and Bust" is a 2013 paper by Nicolai V. Kuminoff (Arizona State University) and Jaren C. Pope (Brigham Young University), published in Land Economics vol. 89, no. 1 (February 2013), pp. 1–29 (DOI 10.3368/le.89.1.1; the journal version is paywalled at UW Press, but author copies and the 2011 Lincoln Institute working-paper version are open — see Sources).[1] Using assessor data on more than a million residential sales in ten metropolitan areas between 1998 and 2009, the authors apply a refined hedonic estimator to separate the market value of land from the market value of structures at the Census-tract level, tracking how each moved through the great housing boom and bust.[1]

It is the source behind the "Kuminoff & Pope (2013) hedonic-regression method" that Lars Doucet's Does Georgism Work? Part 1 flags as one of the twelve estimation methods for aggregate US land value. This page de-references that citation. The paper is evidentiary, not advocacy — the authors do not argue for a land value tax, but their method and results bear directly on how large and how separable land value is.

What It Establishes

• Land and structures can be separated with micro data, at fine spatial resolution. The hedonic estimator uses Census-tract fixed effects (to capture localized, access-based amenity value) plus tract × living-area interactions (to capture unobserved structure quality), yielding annual land- and structure-value estimates per tract. These spatially explicit estimates are "typically an order of magnitude larger than estimates based on the conventional hedonic approach," and broadly consistent, before the boom and after the bust, with the replacement-cost estimates of Davis & Palumbo (2008).[1]
• The boom-bust was large and land-heavy in the aggregate. Residential property values in major metros "more than doubled between 1998 and 2006 and then declined by approximately 40% between 2006 and the end of 2009" (Case-Shiller). Land is the more volatile component across metros: consistent with prior work, land values swing more where housing supply is less elastic.[1]
• But the replacement-cost method over-attributes the swing to land. This is the paper's central, and most Georgism-relevant, correction. Prior replacement-cost studies pin structures to construction cost, so all boom-bust volatility falls on the land residual. Kuminoff and Pope's hedonic estimates instead show that "the market value of structures exceeded their replacement cost during the height of the boom" — by "up to 100% for San Francisco" — with the largest markups in the most affluent neighborhoods.[1] So during the boom, part of what the replacement-cost method records as land appreciation was really structure markups above build cost.
• A counterintuitive within-metro pattern. "Lower value land at the urban fringes of metropolitan areas was the most volatile during the boom-bust" — the opposite of the cross-metro pattern. Tracts with high initial (1998) land shares saw smaller increases in land share during the boom; the negative correlation held in every metro (from −0.393 in Pittsburgh to −0.754 in San Jose).[1]

Relation to the Georgist Case

The paper cuts two ways, and both are useful to a careful Georgist argument.

Supporting the magnitude premise. Like Davis & Heathcote (2007) and Albouy, Ehrlich & Shin (2018), Kuminoff and Pope supply a mainstream, peer-reviewed measurement showing that land is a large, distinctly measurable share of residential property value — the empirical premise beneath land rent could fund a large share of government. Their demonstration that land can be estimated separately at the Census-tract level from ordinary sales data is also direct evidence for the assessment literature: it works against the objection that land value cannot be assessed separately from improvements, and against the informational objection to split-rate taxation, which the paper explicitly motivates ("Successful implementation of a split-rate tax requires accurate estimates for each component of value").[1]

Disciplining the strong version of the "land absorbs the swing" claim. Georgist commentary sometimes reads the replacement-cost land series (e.g. the Fed Flow-of-Funds residual) as showing land bearing essentially the entire housing cycle. Kuminoff and Pope show that inference is partly a method artifact: pinning structures to construction cost mechanically dumps structure markups onto land. This is the same mechanical mis-split that Michael Hudson's Where Did All the Land Go? criticizes from the opposite direction (Hudson argues the residual method makes land disappear in downturns; Kuminoff and Pope find it overstates land's amplitude in booms). The shared lesson — that residual/replacement-cost accounting mis-attributes value between land and structures — is one Albouy, Ehrlich & Shin reach independently.

On the revenue-volatility objection. A standard worry about LVT is that land-based revenue is more volatile than property-tax revenue. The paper's closing implication runs the other way: because the replacement-cost approach "may overstate the value of land during a boom-bust cycle," the authors suggest "moving from a property tax to a land tax may actually help to stabilize revenue streams for some municipalities."[1]

Limits

• Residential land only. Ten metros, owner-occupied-style residential sales; no commercial, industrial, agricultural, or rural land. Aggregate national land totals are not attempted.[1]
• Not a policy paper. The authors make no recommendation on land value taxation; the LVT implications are brief closing remarks, and the split-rate discussion is about measurement, not desirability.[1]
• Hedonic-method assumptions. The results depend on the fixed-effects specification and on assessor micro-data quality; the authors' own refinements move estimates by "an order of magnitude," a reminder that hedonic land-value estimates are sensitive to method. The structure-markup finding is inferred from the hedonic-vs-replacement-cost wedge rather than observed directly, and the paper offers two candidate explanations (imperfect competition; Tobin's-q construction dynamics) without adjudicating between them.[1]
• Stock, not rent. Like the other US land-value studies, the paper reports values, not an annual land-rent flow; converting to rent needs an outside capitalization-rate assumption.

Bears On

• Problem: Land rent could fund a large share of government — a mainstream, peer-reviewed confirmation that land is a large, separately measurable share of property value (magnitude premise), with the caveat that it studies residential land only and reports stock, not rent.
• Benefit: Split-rate taxation increases construction — supplies a fine-grained method for the land/structure separation a split-rate tax requires, while cautioning that a replacement-cost basis would over-tax structure markups in high-amenity neighborhoods.
• Objection: Land value can't be assessed accurately — demonstrates tract-level hedonic separation of land from structures on ordinary sales data.
• Concept: Economic Rent · Land Value Tax

See Also

• Does Georgism Work? (Doucet) — Part 1 cites this paper's hedonic method among its land-value estimators
• Michael Hudson — Where Did All the Land Go? — the parallel critique of residual-method land accounting
• Davis & Heathcote — Price and Quantity of Residential Land — the replacement-cost tradition this paper compares against
• Albouy, Ehrlich & Shin — Metropolitan Land Values — transaction-based alternative, same critique of the residual method
• Most of the rise in the capital share is land

Sources

1. Nicolai V. Kuminoff & Jaren C. Pope, "The Value of Residential Land and Structures during the Great Housing Boom and Bust," Land Economics, vol. 89, no. 1 (February 2013), pp. 1–29. DOI 10.3368/le.89.1.1 (paywalled, UW Press). Open author copy fetched and read in full (2026-07-10): nickkuminoff.github.io/webpage/KP_Land_forthcoming.pdf ("Draft with revisions for Land Economics: April, 2012"); mirror jarenpope.weebly.com/.../2013_kuminoff_pope_land.pdf. Used for: the >1-million-sale / ten-metro / 1998–2009 hedonic design; the finding that structures' market value exceeded replacement cost during the boom (up to 100% in San Francisco); the urban-fringe volatility result; the within-metro land-share regressions (−0.393 to −0.754); and the closing LVT implications (replacement cost may overstate land; land tax may stabilize municipal revenue).
2. Lincoln Institute of Land Policy, working-paper landing page for the same study (2011). lincolninst.edu/publications/working-papers/value-residential-land-structures-during-great-housing-boom-bust — the funded working-paper version; used to confirm provenance and open availability.
3. Lars Doucet (2021), Does Georgism Work? Part 1: Is Land Really a Big Deal?, Astral Codex Ten — flags "Kuminoff & Pope (2013) hedonic-regression method" among the land-value estimation methods; this page de-references that citation. See wiki page.