A Modern Analysis of the Effects of Site Value Taxation

Summary

Jan K. Brueckner's "A Modern Analysis of the Effects of Site Value Taxation" appeared in the National Tax Journal 39(1), pages 49–58, in March 1986 — the same National Tax Association journal that later carried Oates & Schwab's Pittsburgh study. Brueckner, an urban and public economist then at the University of Illinois at Urbana-Champaign (the paper first circulated as BEBR Faculty Working Paper No. 1045 of the university's College of Commerce and Business Administration, dated May 1984, under the reversed title "The Effects of Site Value Taxation: A Modern Analysis"), wrote the paper to fill a specific gap: formal theoretical analysis of "graded" or site-value taxation — a property tax that lowers the rate on improvements and raises it on land — was largely absent from the earlier literature on the subject, which had mostly been descriptive or advocacy-driven. Brueckner set out to model the effects of such a shift using the standard tools of urban land-use economics.

This paper is the theoretical anchor most later empirical work on split-rate taxation cites or presupposes. Where Oates & Schwab (1997) and Plassmann & Tideman (2000) measure whether split-rate taxation increased construction in Pennsylvania cities, Brueckner (1986) is the paper that explains why standard economic theory predicts it should.

The Core Argument

Brueckner builds a long-run market-equilibrium model of urban land development — in the tradition of the monocentric-city / urban land-use literature — in which a developer chooses how much capital (improvements) to put on a given site, and land value itself is allowed to adjust so that development yields zero economic profit in equilibrium. Within this framework he analyzes a revenue-neutral shift from a uniform property tax (which taxes land and improvements at the same rate) to a graded tax: the rate on improvements is lowered and the rate on land is raised to hold total revenue constant.

Four results stand out (this summary is drawn from a direct read of the full working-paper text, freely available via the Internet Archive — see Sources — not merely the abstract). Brueckner is explicit about which are new: "While the impact of gradation on the level of improvements was well understood in the earlier literature, the predictions relating to land values and the price of housing are entirely new."

1. Capital intensity rises. For plausible values of the relevant elasticities, lowering the tax rate on improvements while raising the rate on land increases the intensity of improvements per unit of land — i.e., it raises the capital-to-land ratio. This is the formal mechanism behind the empirical claim that split-rate taxation increases construction: a tax on improvements raises the effective cost of the capital a developer sinks into a structure, distorting the capital-land ratio downward relative to the untaxed optimum; cutting that tax rate removes part of the distortion, so more capital is built on the same land. Brueckner notes this direction of effect was already the consensus of the earlier (informal) literature — his contribution here is to derive it rigorously rather than to overturn it.
2. The effect on land value is a genuine open question — and the answer depends on how large the taxed area is. Raising the land tax rate mechanically reduces the after-tax return to landownership, which by itself would push land value down. But lowering the improvements tax rate raises the profitability of developing that land more intensively, which by itself pushes site value up (developers will bid more for land they can build on more profitably). Brueckner shows the net effect is not a wash, and — importantly — is different in his two cases:
3. Small taxed zone (housing price held fixed): land value rises, not merely stays flat. Brueckner's headline "surprising" result is that this net effect is not actually ambiguous once the algebra is done: "the mathematics unambiguously indicate that the impact of the lower improvements tax wins out and land value rises." The result holds provided the revenue-preserving cut in the improvements tax needed to offset a higher land tax is negative (∂τ/∂θ < 0), which Brueckner shows is "almost certainly negative" when evaluated at the starting point of a uniform tax (τ = θ). The condition turns on the elasticity of substitution between capital and land being large enough; a knife-edge case with zero substitution (σ = 0) leaves land value exactly unchanged, and positive substitution tips it to an increase. Land value keeps rising as the tax burden shifts further onto land, "reaching a maximum in the case of pure site value taxation."
4. City-wide taxed zone (housing price allowed to adjust): the result flips. Improvements per acre still rise, but the improvements-tax cut is no longer fully offset by a housing-price change, and Brueckner finds that "an increase in θ depresses the price of housing, raises the level of improvements, and, as long as housing demand is not highly elastic, depresses the value of land." So land value falls in the economy-wide case.

This is the key nuance the popular framing misses: whether a graded tax raises or lowers land value in Brueckner's model depends on whether it is applied to a small sub-market (land value rises) or the whole housing market (land value falls, alongside a fall in housing prices) — not merely on unspecified "parameter values." What is genuinely conditional is the sign of ∂τ/∂θ (driven by the elasticity of substitution); given ∂τ/∂θ < 0, the direction of the land-value effect is pinned down within each case. 3. Housing prices fall when the tax is city-wide. In the small-zone case housing price is (by assumption) unaffected. But when gradation is applied across the whole housing market, the improvements tax is shifted forward into housing prices, so cutting it lowers the price of housing: "an increase in θ depresses the price of housing." Brueckner flags this as an "entirely new" prediction the earlier literature had been "virtually silent" on — and it is why land value falls in the city-wide case (the housing-price decline drags site value down with it, unless housing demand is highly elastic). 4. Short-run windfalls have a surprising spatial incidence. Brueckner separately analyzes the incidence of the transition itself: when a graded tax is first imposed, existing developed parcels experience one-time windfall gains or losses. His counter-intuitive finding is that, "contrary to a common impression, the most intensively developed parcels suffer windfall losses in the form of higher taxes, while the least intensively developed parcels benefit from windfall gains" — i.e. the parcels with the lowest improvements-to-land ratio (underused land) are the transitional winners. These one-time effects are conceptually distinct from the paper's long-run, steady-state predictions.

Relation to the Georgist Case

Brueckner's paper is squarely in the tradition Georgists cite for the theoretical case that taxing land more and improvements less encourages building rather than idling land — but it is a mainstream urban-economics paper, not a Georgist tract, and its result is more nuanced than the popular framing. It does not claim that a pure site value tax (with zero tax on improvements) is costless or panacea-like; it formally derives conditions under which a partial, revenue-neutral shift toward land taxation raises capital intensity, and it treats the effect on land value as a genuine open theoretical question resolved only under specific assumptions, not as an automatic or unconditional gain. This is the "theoretical backbone" role assigned to it here: it supplies the mechanism — removing a tax-induced distortion on the capital-land ratio — that later empirical studies of Pittsburgh and Pennsylvania municipalities are interpreted as confirming.

Because the paper concerns a property tax base shift (structures vs. land within a jurisdiction's existing property tax), rather than a shift away from broader capital taxation (corporate income tax, capital gains tax, etc.), its bearing on the outcome that LVT can replace capital taxes without efficiency loss is best read narrowly: it is a rigorous, formal instance of the general principle that taxing a produced, elastically-supplied factor (capital/improvements) creates a real distortion that taxing a fixed factor (land) does not, in the specific context of urban development decisions. Brueckner himself makes the general connection only briefly, in his concluding normative aside: because land's supply elasticity is zero, optimal-tax theory implies the tax burden should fall on land alone, so "modern theory vindicates George's belief in the efficiency of site value taxation." The paper is therefore a strong primary citation for the split-rate/construction mechanism, and a supporting (rather than standalone) citation for the broader economy-wide capital-tax-substitution outcome, for which a general optimal-tax or capital-taxation source is a better anchor.

Nuances and Limits

• Partial vs. general equilibrium. The model is a long-run equilibrium of urban land development with zero-profit developers, not a full general-equilibrium simulation of an entire regional or national economy. A direct computational general equilibrium extension of the same question was carried out the following year by Joseph A. DiMasi (1987), "The Effects of Site Value Taxation in an Urban Area: A General Equilibrium Computational Approach," National Tax Journal 40(4), which parameterized a spatial model to a specific metropolitan area and estimated a welfare gain from the tax shift of up to roughly 6.6% of tax revenue raised under its most realistic parameterization — a result consistent in direction with Brueckner's theoretical prediction of a welfare-improving shift, obtained via a different (computational) method.
• A short-run companion model reaches a partly different emphasis. Chin W. Yang and Dwight B. Means, "A Welfare Analysis of the Site Value Taxation Model," Journal of Real Estate Finance and Economics 5 (1992), explicitly build a short-run version of Brueckner's long-run analysis (with land value held fixed rather than adjusting), finding that when land is relatively inexpensive, the graded tax system delivers a higher capital-improvement-per-unit-of-land ratio, higher initial housing output, and higher consumer surplus than the uniform tax it replaces — broadly reinforcing Brueckner's direction of effect while relaxing his long-run, adjustable-land-value assumption.
• The scope of the tax zone is the headline nuance, not a flaw. The paper's result on land value is conditional in a specific, now-verified way: whether land value rises or falls depends chiefly on whether the graded tax is imposed on only a small part of a housing market (housing price fixed → land value rises) or on the whole market (housing price adjusts down → land value falls, as long as housing demand is not highly elastic). Within each case the direction is pinned down once the revenue-preserving improvements-tax cut is negative (∂τ/∂θ < 0), a condition Brueckner argues is "almost certainly" met at the starting uniform-tax point and which turns on the elasticity of substitution between capital and land. Readers should not treat "land value rises" as an unconditional prediction of the model; the small-zone-vs-city-wide distinction is what flips the sign. (This is more precise than the common secondary paraphrase that the land-value effect is simply "ambiguous.")
• Single-city/theoretical model, not a policy evaluation. As with most monocentric-city-style urban models, the analysis characterizes comparative-static effects of a stylized policy change rather than evaluating a specific real-world jurisdiction's history; the empirical confirmation for a real transition comes from separate studies such as Oates & Schwab (1997) and Plassmann & Tideman (2000).
• Normative note in the paper itself. Brueckner explicitly declines to evaluate whether the shift is desirable, then observes that the answer "is implicit in the optimal tax literature" and "affirms Henry George's original claim": because land supply is perfectly inelastic, optimal-tax theory implies "it is optimal for society's tax burden to be carried by that commodity alone," so "modern theory vindicates George's belief in the efficiency of site value taxation." This is the one place the paper connects directly to the Georgist efficiency case.
• Primary-source access (resolved). The full working-paper text (BEBR No. 1045, identical in model and results to the published National Tax Journal article) is freely readable via the Internet Archive plain-text copy, and the figures, propositions, and quotations above were verified directly against it in this review. The University of Chicago Press journal version remains paywalled, but the working paper is not.

Bears On

• Outcome: Split-rate taxation increases urban construction — this is the direct theoretical mechanism (rising capital-land ratio under a graded tax) that the Pittsburgh and Pennsylvania empirical evidence is interpreted as confirming.
• Outcome: LVT can replace capital taxes without efficiency loss — a rigorous, narrower instance of the general claim that shifting tax burden off capital (here, improvements) and onto fixed land removes a real distortion, in the specific setting of urban property taxation rather than economy-wide capital taxation.
• Concept: Split-Rate Taxation — this paper is the primary theoretical source for why the split-rate design is expected to work.
• Concept: Deadweight Loss — the paper's mechanism is a specific application of the general result that taxes on elastically-supplied factors (capital/improvements) create deadweight loss while taxes on fixed factors (land) do not.
• Concept: Tax Capitalization — the paper's ambiguous/positive land-value result turns on how the two tax-rate changes are separately capitalized into site value.

See Also

• Split-Rate Taxation
• Split-rate taxation increases urban construction
• The Impact of Urban Land Taxation: The Pittsburgh Experience (Oates & Schwab, 1997)
• A Markov Chain Monte Carlo Analysis of the Effect of Two-Rate Property Taxes on Construction (Plassmann & Tideman, 2000)
• Assessing the Theory and Practice of Land Value Taxation (Dye & England)
• Land Value Tax
• Deadweight Loss

Sources

1. Jan K. Brueckner (1986), "A Modern Analysis of the Effects of Site Value Taxation," National Tax Journal 39(1):49–58. University of Chicago Press / DOI (paywalled) — the published journal version; used for the paper's title, journal, volume/issue/pages. The model, propositions, and quotations summarized on this page were verified against the identical full text of the working-paper version (Source 2).
2. Open-access full text (primary source used for this page): Jan K. Brueckner, "The Effects of Site Value Taxation: A Modern Analysis," BEBR Faculty Working Paper No. 1045, College of Commerce and Business Administration, University of Illinois at Urbana-Champaign, May 1984. Free plain-text scan at archive.org/details/modernanalysisof1045brue (full text confirmed accessible in this review via the item's _djvu.txt). Directly verified: the abstract; the long-run capital-intensity result; the small-zone land-value result ("the mathematics unambiguously indicate that the impact of the lower improvements tax wins out and land value rises"); the city-wide result ("an increase in θ depresses the price of housing... and... depresses the value of land"); the "entirely new" characterization of the land-value and housing-price predictions; the surprising windfall incidence ("the most intensively developed parcels suffer windfall losses... while the least intensively developed parcels benefit from windfall gains"); and the concluding optimal-tax/Henry George normative aside.
3. Joseph A. DiMasi (1987), "The Effects of Site Value Taxation in an Urban Area: A General Equilibrium Computational Approach," National Tax Journal 40(4). DOI — used for the general-equilibrium computational follow-up to Brueckner's model and its welfare-gain estimate.
4. Chin W. Yang & Dwight B. Means (1992), "A Welfare Analysis of the Site Value Taxation Model," Journal of Real Estate Finance and Economics 5:281–290. DOI — used for the short-run companion model and its findings on capital improvement per unit of land, housing output, and consumer surplus.
5. Richard J. Arnott & James G. MacKinnon (1977), "The Effects of the Property Tax: A General Equilibrium Simulation," Journal of Urban Economics 4(4):389–407 — an earlier general-equilibrium-simulation study in the same broad research tradition. Note: a direct full-text search of Brueckner (1986)/BEBR No. 1045 confirms that Brueckner does not cite Arnott & MacKinnon (his paper's related-work discussion runs through the site-value-tax literature — Carmean, Becker, Harriss, McCalmont, Archer, etc. — and he flags the absence of a tractable general-equilibrium treatment as an open question in his conclusion). The two papers are linked in the secondary literature, not in Brueckner's own text; this entry is retained only as background on the general-equilibrium lineage.
6. Richard F. Dye & Richard W. England (eds.), Assessing the Theory and Practice of Land Value Taxation, Lincoln Institute of Land Policy — wiki summary — a standard literature review of land value taxation theory and practice; cited here only for general context on how theoretical results of this kind sit within the wider LVT literature, not for any specific figure or quotation about Brueckner (1986).
7. Companion empirical evidence: Oates & Schwab (1997) — wiki summary · Plassmann & Tideman (2000) — wiki summary