High land rents suppress productivity

At a glance — Model evidence indicates high land costs in the most productive cities misallocate workers and lower national output, but that evidence runs mainly on land-use regulation; the specifically Georgist version — that privately captured land rent itself suppresses productivity — is thinner and rests largely on a single model. Evidence: Emerging (strong top-journal evidence for the closely-related land-use-restriction channel; the direct privately-captured-rent version is thinner and one flagship magnitude is contested) · 6 supporting sources · 0 challenging Strongest support: Duranton & Puga (2023) — a quantitative model in which relaxing planning restrictions in seven large US cities raises aggregate output about 8%. No structural counter-evidence is currently wired; see Limits.

The Claim

When land costs in the most productive cities are very high, workers and firms that would create the most value there are priced out and pushed to less productive locations. This misallocation lowers a nation's total factor productivity (TFP) and output.

The three strongest citations:

1. Duranton & Puga (2023), Econometrica — a quantitative model in which incumbent residents use planning regulation to cap city growth; relaxing regulation in seven large, highly-restricted US cities raises aggregate output by about 7.95%. (Peer-reviewed, top-five journal, uncontested.)
2. Herkenhoff, Ohanian & Prescott (2018), Journal of Monetary Economics — a general-equilibrium model of state land-use restrictions in which moving all states halfway to Texas's (low) restriction level raises US labor productivity by about 12.4%; the authors conclude tightening in California and New York misallocates labor and capital nationwide. (Nobel-laureate coauthor; no Georgist affiliation.)
3. Hsieh & Moretti (2019), AEJ: Macroeconomics — the most-cited statement of the "housing-as-misallocation" mechanism, estimating that housing-supply constraints in a few superstar cities substantially lowered US growth 1964–2009. Note: its headline magnitude is the subject of an active, unresolved replication dispute (Greaney 2026) — cite the mechanism, not the number.

Honest limits: all three strongest sources concern land-use regulation (zoning/supply restriction), a mechanism closely related to but distinct from the Georgist claim that privately captured land rent itself suppresses productivity — and their remedy is deregulation, not a land tax. The evidence that runs directly on privately-captured rent is thinner: one supplementary model of a land value tax (Fiorentino & Moogan 2025), and an IMF paper often cited here (Bakker 2023) that on inspection is about TFP mis-measurement, not misallocation (see Counter-Evidence).

The Mechanism

Productive cities offer high wages and returns because of agglomeration — but if being there becomes very expensive, the cost of access rises until it offsets the advantage for marginal entrants. The economy ends up with too few people and firms in its most productive places, and aggregate output is lower than a less-constrained allocation would produce.

Two distinct stories can raise that access cost, and the wiki keeps them separate:

• Regulation-induced scarcity. Zoning, height limits, and permitting restrict how much can be built on valuable land, so extra demand shows up as higher prices rather than more housing. This is the channel modeled by Hsieh & Moretti, Duranton & Puga, and Herkenhoff, Ohanian & Prescott. Its natural remedy is building more (land-use liberalization).
• Privately captured location rent. Even without binding regulation, the agglomeration surplus can be absorbed as land rent by owners and speculators, raising the cost of being there. This is the specifically Georgist version, and the remedy it points to is taxing land value — see Fiorentino & Moogan (2025) and taxing land and rents increases productivity. A much earlier, informal statement of the same channel — overpriced land redirecting capital into slow-turnover, land-substituting investment rather than production — is Mason Gaffney's 1990s taxonomy of "land-saving," "land-enhancing," "land-linking," "land-capturing," and "rent-leading" capital; see Gaffney: How Land Markets Misallocate Capital. It is a conceptual precursor, not empirical confirmation — cited here at the weakest evidentiary tier, below the structural models. A second, still earlier theoretical precursor makes a related but distinct point: Gaffney's 1961 formal model of "time-indivisibility" shows that because land ownership requires financing a perpetual claim rather than renting services incrementally, and interest rates differ by wealth, land tends to gravitate toward low-interest, financially strong holders rather than the credit-constrained users who would work it most productively — see Gaffney: The Unwieldy Time-Dimension of Space. Also a conceptual precursor rather than empirical confirmation, cited at the same weakest tier. A third theoretical precursor, earlier than WP041/WP042 and more fully worked out mathematically, derives the mechanism from the opposite direction: Gaffney's 1976 "Toward Full Employment with Limited Land and Capital" derives, with calculus, that raising land rent (like raising interest rates) shortens the optimal capital-recovery cycle and lowers labor's share of an investment's return — i.e., that overpriced land structurally redirects capital toward slower-turnover, less labor-intensive forms. This is cited here in prose only, not added to this page's supported_by list — it is theoretical framework illustrated with assumed parameters, not empirical evidence, and the wiki keeps theory out of evidence-tier citation arrays.

The two channels overlap — regulation raises the scarcity rent that owners capture — but the empirical evidence is far stronger for the first than for the second, and a land value tax is not the same policy lever as upzoning. The strongest honest reading is that high land costs in productive places demonstrably misallocate activity; how much of that is fixable by taxation rather than deregulation is not settled by the papers below.

The Evidence in Detail

Three structural modeling papers, all by mainstream economists with no Georgist affiliation, converge on the same direction. Duranton & Puga (2023) build a quantitative model in which incumbent residents use planning regulation to limit city growth; relaxing regulation in seven large restricted US cities raises aggregate output by about 7.95% (sensitivity range roughly 5.7–8.2%). Herkenhoff, Ohanian & Prescott (2018) reach a parallel conclusion from a general-equilibrium model of state land-use restrictions: California and New York's tightening misallocates labor and capital nationwide, and deregulation would raise US TFP and output growth. Hsieh & Moretti (2019) supply the most-cited version of the argument via a labor-mobility channel. All are model-based estimates — their magnitudes carry the usual calibration caveats — but the direction is consistent across three independently specified models (partial-equilibrium, endogenous-regulation, and dynamic general-equilibrium). That convergence is the real strength of the case; the specific percentages are not interchangeable and, in the Hsieh & Moretti case, are actively disputed (see below).

On the specifically-rent version of the claim, Fiorentino & Moogan (2025) model a land value tax easing agglomeration misallocation and improving both efficiency and equity — the closest work to the Georgist mechanism, but a single unrefereed model of modest weight, not an empirical confirmation.

The rent-capture premise of that channel — that the agglomeration surplus is absorbed as land rent rather than left to accrue to the productive workers who generate it — does have strong causal evidence. Hornbeck & Moretti (2018/2024) use four alternative instruments for local productivity shocks and find that for renters the resulting earnings gains are "largely offset by increased cost of living," while "for homeowners, the benefits are substantial" — productivity growth capitalizing into land, an explicit revival of Ricardo (their words). This bears on the specifically-rent channel (private location rent, not regulation) that the three headline papers do not isolate. It is double-edged, and cited honestly here: once worker mobility is counted, Hornbeck & Moretti's national conclusion is that "the impacts on landowners are largely a transfer from one location to another," incidence "falls mainly on workers," and aggregate purchasing power still rises (~0.5–0.6%/yr). So the paper documents that high land rents capture the productivity surplus locally — the first link in the suppression mechanism — without establishing that they lower aggregate output. It strengthens the thin rent-channel premise; it does not, by itself, prove the suppression conclusion.

Counter-Evidence

This claim's evidence base has real weaknesses, and honest use requires stating them:

• The flagship magnitude is contested. Greaney (2026), a formal comment in the same journal as Hsieh & Moretti, reports that their counterfactual does not reproduce a positive output gain as specified, traces this to coding errors and an arbitrary dependence on population units, and — in a corrected model — finds an effect "two orders of magnitude smaller than what they report" (Greaney 2026, abstract). Hsieh has replied and the dispute is live and unresolved. Bryan Caplan's earlier (2021) correction, by contrast, argued the true effect was larger. The direction survives; the number does not.
• Other credible estimates are much smaller, and reverse causality is possible. In the published discussion of Herkenhoff–Ohanian–Prescott, Tonetti (2018) computes deregulation gains "much smaller than the 0.8 percentage points per year or doubling of growth as found in Hsieh and Moretti," and raises a reverse-causality concern: strict regulation may be a purchased amenity of high-productivity workers, so "California may be a high (measured) TFP state because it has strict land-use regulations that attract high productivity workers" — in which case deregulation could lower measured TFP. Duranton & Puga also note Desmet & Rossi-Hansberg (2013), who find reducing cross-city frictions moves city populations a lot but has only minor welfare effects.
• The most-cited "direct rent" anchor is actually a measurement paper. Bakker (2023, IMF) is frequently cited here as showing that land rents lower TFP. Its abstract argues something different — that "standard growth decompositions have underestimated TFP growth by overestimating the contribution of capital, failing to account for the substantial part of capital income directed to urban land rents." That is a claim about TFP mis-measurement (land rents misattributed to capital), not about land rents suppressing real productivity; Bakker in fact concludes true TFP in high-land-rent Singapore is higher than measured. It bears on the capital-share debate more than on this claim, and is not independent confirmation of it.
• Regulation, not taxation. The strong evidence identifies restricted land supply as the binding friction. Whether a land value tax (rather than upzoning) would relieve the misallocation is untested by any of these papers — and where land value has been captured without relaxing supply, prices did not fall (see Land capture didn't make housing cheap).

Strength of Evidence

Emerging. Multiple independent, top-journal structural models agree on the direction — high land costs in the most productive cities misallocate labor and capital and lower aggregate output — which is a genuine and non-trivial result. But the case falls short of "Moderate" on this claim as stated: the strongest evidence concerns land-use regulation, a mechanism distinct from the specifically-Georgist privately-captured-rent version; the direct-rent evidence is a single supplementary model plus an IMF paper that turns out to be about measurement; and the single most-cited magnitude is the subject of an active, unresolved replication dispute. The direction is robust; the size of the effect, and how much of it a land tax (rather than deregulation) could recover, remain open.

See Also

• Albouy: What Are Cities Worth? — evidence that productivity, not amenities, drives most cross-city land-rent variation (the premise this claim builds on, not the suppression conclusion)
• Taxing land and rents increases productivity — the mirror-image benefit claim, with the least-harmful vs raises-productivity gap kept explicit
• The capital-share rise is really land — where Bakker's measurement argument (land rents hidden inside measured capital income) actually bears
• Land capture didn't make housing cheap — why capturing land value without relaxing supply need not lower costs
• Gaffney: How Land Markets Misallocate Capital — the theoretical precursor to the privately-captured-rent channel
• Gaffney: The Unwieldy Time-Dimension of Space — a second, earlier (1961) theoretical precursor: financial power, not productivity, allocates land under time-indivisibility
• Gaffney (1976): Toward Full Employment with Limited Land and Capital — a third theoretical precursor, cited in prose only: a formal derivation of how land rent shortens optimal capital-turnover cycles and lowers labor's share of returns
• Economic Rent · Land Value Tax · Agglomeration Economies

Sources

1. Gilles Duranton & Diego Puga (2023), "Urban Growth and Its Aggregate Implications," Econometrica 91(6): 2219–2259. DOI — used for the ~7.95% aggregate-output gain from relaxing regulation in seven restricted US cities. wiki summary
2. Kyle F. Herkenhoff, Lee E. Ohanian & Edward C. Prescott (2018), "Tarnishing the Golden and Empire States," Journal of Monetary Economics 93: 89–109. DOI — used for the general-equilibrium finding that state land-use restrictions misallocate labor and capital nationally (halfway-to-Texas raises labor productivity ~12.4%). wiki summary
3. Chang-Tai Hsieh & Enrico Moretti (2019), "Housing Constraints and Spatial Misallocation," AEJ: Macroeconomics 11(2): 1–39. AEA — used for the labor-misallocation mechanism; magnitude treated as contested per Greaney (2026). wiki summary
4. Brian Greaney (2026), "Housing Constraints and Spatial Misallocation: Comment," AEJ: Macroeconomics 18(2): 409–428. AEA — used for the counter-evidence that the corrected effect is "two orders of magnitude smaller." documented via wiki summary
5. Christopher Tonetti (2018), "Comment on 'Tarnishing the golden and empire states'," Journal of Monetary Economics 93: 110–113. Author copy — used for the smaller-estimate and regulation-as-amenity reverse-causality caveats. documented via wiki summary
6. Fiorentino & Moogan (2025), "LVT and Urban Agglomeration Dynamics," SSRN — used for the (supplementary) model of a land value tax easing agglomeration misallocation. wiki summary
7. Bas Bakker (2023), "Unveiling the Hidden Impact of Urban Land Rents on Total Factor Productivity," IMF Working Paper 2023/170. IDEAS/RePEc — cited for the caveat that urban land rents cause standard growth decompositions to underestimate TFP growth (a measurement point, not misallocation evidence). wiki summary
8. Richard Hornbeck & Enrico Moretti (2018/2024), "Estimating Who Benefits from Productivity Growth," NBER WP 24661 / Review of Economics and Statistics 106(3). DOI — used, with the double-edged national-incidence caveat, for four-IV causal evidence that local productivity growth capitalizes into land rent (the specifically-rent channel's premise). wiki summary
9. Mason Gaffney, "How Wayward Land Markets Lead to Misallocating Capital," working paper WP042, undated (circa 1993) — wiki summary — used as a weakest-tier theoretical/historical precursor to the privately-captured-rent channel (not empirical evidence; an informal 1990s taxonomy, not a modern structural model).
10. Mason Gaffney (1961), "The Unwieldy Time-Dimension of Space," American Journal of Economics and Sociology 20(5): 465–481 — wiki summary — used as a second, earlier weakest-tier theoretical precursor: a formal model of how unequal financing costs allocate land to low-productivity holders (not empirical evidence; a 1961 theoretical article, not a modern structural model).
11. Mason Gaffney (1976), "Toward Full Employment with Limited Land and Capital," in Arthur D. Lynn Jr. (ed.), Property Taxes, Land Use and Public Policy (University of Wisconsin Press), pp. 99–166 — wiki summary — used as a third theoretical precursor, cited in prose only (not in supported_by): a formal derivation of how land rent shortens optimal capital-turnover cycles (not empirical evidence).