Assessing the Distributive Impact of a Revenue-Neutral Shift from a Uniform Property Tax to a Two-Rate Property Tax with a Uniform Credit

Summary

"Assessing the Distributive Impact of a Revenue-Neutral Shift from a Uniform Property Tax to a Two-Rate Property Tax with a Uniform Credit" is a peer-reviewed article by Richard W. England (Professor of Economics, University of New Hampshire, and a long-time Lincoln Institute of Land Policy-affiliated public-finance economist specializing in property taxation) and Min Qiang Zhao (at the time of the working paper, Department of Economics, Ohio State University), published in the National Tax Journal, Vol. 58, No. 2 (June 2005), pp. 247–260 — the same flagship peer-reviewed US public-finance journal that later carried Bowman & Bell's 2008 replication of this study. A working-paper version was also circulated by the Lincoln Institute of Land Policy (dated December 2004), which is the free, stable, publicly accessible copy this page cites; the published version carries DOI 10.17310/ntj.2005.2.05.

The paper studies a hypothetical revenue-neutral shift in Dover, New Hampshire from the city's existing uniform (single-rate) property tax to a two-rate ("split-rate") property tax — a lower rate on the value of improvements (buildings) than on land value, the classic Georgist "two-rate" reform — combined with a uniform tax credit applied per parcel, using assessment data on single-family residential properties. Because a two-rate tax that shifts weight from improvements toward land moves partway toward a pure land value tax, this paper is one of the earliest rigorous parcel-level empirical tests of who wins and loses under an LVT-style reform, and it is the paper that Bowman & Bell (2008) explicitly set out to replicate on a different city (Roanoke, VA) to test whether its result generalizes.

The Core Argument and Findings

• Design tested. England and Zhao model a revenue-neutral move from Dover's uniform property tax rate to a two-rate structure that taxes land value more heavily than improvement value, holding total municipal revenue fixed, and evaluate adding a uniform (flat, per-parcel) tax credit as a distributional offset.
• Sample. The empirical analysis uses parcel-level assessment data for single-family residential properties in Dover, NH, for the 2002 assessment year — 5,250 single-family homes (Table 4, "N = 5250"), with a mean assessed value of $198,170 and a mean building-to-land value ratio of 1.88. (Dover's full 2002 developed-parcel roll comprised 8,475 taxable parcels.)
• Headline finding: regressive by assessed value. Using assessed property value as the incidence proxy (a standard proxy for wealth in this literature), the authors find that shifting toward land taxation without a credit would increase the property tax burden on lower-value single-family homes relative to higher-value ones — i.e., the reform is regressive among residential owners within Dover. In their own words, "movement towards land value taxation would, in the absence of a credit, benefit primarily those with more expensive homes." Under a pure land value tax with no credit, roughly 74 percent of homeowners in the top group would get tax cuts (averaging about $138/year), while about 80 percent of the middle group and essentially all of the bottom group would face hikes — averaging +$311 for the middle group and a "wrenching" +$634 for the bottom group. The authors conclude such a shift "would, in the absence of a credit provision, have a highly regressive impact on homeowners." The mechanism is the parcel's value ratio (building value ÷ land value): properties whose ratio exceeds the citywide average benefit from taxing land more heavily, and because parcel assessment is positively correlated with the value ratio in Dover, higher-valued homes tend to carry relatively less land value as a share of assessed value and so gain from the shift.
• The proposed remedy: a uniform credit. England and Zhao's own response to this regressivity is not to abandon the two-rate design, but to pair it with a uniform (flat) tax credit per parcel, funded out of the same revenue-neutral envelope, which offsets the increased burden on lower-value properties and restores what they call "rough" progressivity. Their preferred "restrained" design — cutting the building tax rate to $10.98 per $1,000 of assessed value while introducing a $1,000 annual credit — would leave roughly 72 percent of the middle group and nearly 80 percent of the bottom group paying less. The authors caution the result is "imperfect, at best" (even under this design, about 4 percent of the bottom group would see tax hikes exceeding 10 percent of their prior bill, owing to unusually low value ratios). A more aggressive pure land value tax with a large ($2,000) credit was rejected as politically infeasible: financing it would require a land tax rate near $70 per $1,000, and over 72 percent of the bottom group would still face a tax hike (averaging ~$170/year). This is the origin of the "with a Uniform Credit" clause in the paper's own title — the credit is not a footnote but the paper's central proposed policy design.

Relation to the Georgist Case

This paper is the honest counterweight to the wiki's broader claim that land value taxation tends to be progressive. The outcome page A land value tax can be progressive rests on the premise that because land ownership and land-heavy housing wealth are concentrated among wealthier households, taxing land value disproportionately falls on the wealthy. England and Zhao's Dover result is a direct, peer-reviewed empirical counter-example to the unqualified version of that claim: in Dover's specific housing stock, it was lower-valued (and, the authors argue, likely lower-income) residential properties that carried relatively more land value as a share of total assessed value, so a land-favoring shift raised their relative burden rather than lowering it.

Crucially, England and Zhao do not treat this as a reason to reject the two-rate/LVT approach — their paper's proposed fix, a uniform credit funded from the same revenue, is structurally the same idea echoed decades later in Common Wealth Canada's modelling of a national LVT paired with a large flat refundable credit (CWC finds LVT-alone is regressive by income but that the credit reverses this for a large majority of households). Read together, the papers converge on the same design lesson: an unadorned shift toward land taxation is not automatically progressive, and its progressivity in practice depends on pairing it with a redistributive mechanism (a uniform credit, a citizen's dividend, or an equivalent rebate) rather than relying on land-concentration alone.

This paper should therefore be read alongside, not instead of, Bowman & Bell (2008), which applied the identical method to Roanoke, VA and found the opposite result (progressive by both assessed value and census-tract income/poverty). The pair of studies together demonstrate that the raw distributional incidence of a property-tax-to-land-tax shift is jurisdiction-dependent — it hinges on the local relationship between land value, improvement value, and income across the housing stock — rather than a fixed property of land value taxation in either direction.

Nuances and Limits

• Single city, single year, residential-only. The finding is drawn from one New Hampshire city's 2002 assessment roll for single-family residential parcels only; it does not model commercial, industrial, or agricultural land, does not cover multiple assessment years, and — as the existence of Bowman & Bell's replication itself demonstrates — does not generalize to other jurisdictions.
• Incidence proxy, not household income. Like the later Bowman & Bell study, England and Zhao measure incidence primarily via assessed property value as a proxy for household wealth/income, not directly observed household income; this can mask cases such as asset-rich, cash-poor households whose true income does not track their property value (see Objection: LVT hurts the "asset-rich, cash-poor").
• Statutory/first-round incidence only. The analysis assumes the assessed burden change falls on the current owner; it does not model longer-run capitalization effects, behavioral responses (e.g., changes in improvement investment under a lower improvement rate), or landlord-to-tenant pass-through for rental properties.
• The regressive finding is about the shift, not about a fully-phased-in LVT level. The paper studies the distributional effect of changing from one tax structure to another in a revenue-neutral way; it says less about the level of burden under a mature, long-established two-rate or full land value tax system where prices have already adjusted.
• The remedy is offered by the same authors, not an independent rebuttal. England and Zhao's own proposed uniform credit is a design fix presented within the same paper, not a separate line of evidence — a reader should not treat "regressive, but fixable with a credit" as two independently corroborating findings; it is one study's empirical result plus its authors' policy recommendation.

Bears On

• Outcome: A land value tax can be progressive — this paper is direct counter-evidence to the unqualified version of the claim; it belongs in that outcome's challenged_by, not its supported_by, since the paper's own headline finding for Dover is regressivity, not progressivity. [VERIFY: orchestrator to confirm wiring — this page's own supports_outcomes is left empty precisely because the source does not support the claim as stated.]
• Research: Bowman & Bell — Distributional Consequences of Converting the Property Tax to a Land Value Tax — the direct replication of this study's method on Roanoke, VA, which reaches the opposite conclusion; the two papers together establish jurisdiction-dependence.
• Research: Common Wealth Canada — Assessing the Distributional Impacts of a Land Value Tax — an independent, decades-later case reaching a structurally identical conclusion (LVT alone can be regressive; a flat credit/rebate restores progressivity), reinforcing this paper's proposed remedy.
• Objection: LVT hurts the "asset-rich, cash-poor" — this paper's assessed-value incidence proxy and its regressivity finding are directly relevant background for this objection.
• Concept: Land Value Tax — a concrete parcel-level empirical test of a two-rate (partial LVT) reform's distributional consequences.

See Also

• Bowman & Bell — Distributional Consequences of Converting the Property Tax to a Land Value Tax
• Assessing the Distributional Impacts of a Land Value Tax (Common Wealth Canada)
• A land value tax can be progressive
• Land Value Tax
• Objection: LVT hurts the "asset-rich, cash-poor"

Sources

1. Richard W. England & Min Qiang Zhao (2005), "Assessing the Distributive Impact of a Revenue-Neutral Shift from a Uniform Property Tax to a Two-Rate Property Tax with a Uniform Credit," National Tax Journal 58(2), pp. 247–260. DOI 10.17310/ntj.2005.2.05 (resolves to journals.uchicago.edu, which returned HTTP 403 to this session's automated fetch) — used for the exact citation, journal, volume/issue/pages, and year. The published version is paywalled, but the substantive findings below are verified against the free Lincoln Institute working paper (source 2), whose data, tables, and conclusions the published article carries forward.
2. Richard W. England & Min Qiang Zhao, "Assessing the Distributive Impact of a Revenue-Neutral Shift from a Uniform Property Tax to a Two-Rate Property Tax with a Uniform Credit" (working paper WP04RE2, December 2004), Lincoln Institute of Land Policy. Landing page: Lincoln Institute; full 24-page PDF directly fetched this session at go.lincolninst.edu/1000_england_complete_web.pdf. Used for the directly-verified confirmation of: authors and Zhao's affiliation (Department of Economics, Ohio State University); the Dover, NH study city and 2002 assessment year; the sample of 5,250 single-family homes (Table 4); the regressivity magnitudes under a no-credit shift (top group ~74% cuts, avg −$138; middle avg +$311; bottom avg +$634); and the "rough progressivity" restored by the preferred two-rate-plus-$1,000-credit design (≈72% of the middle group and ≈80% of the bottom group paying less).
3. John H. Bowman & Michael E. Bell (2008), "Distributional Consequences of Converting the Property Tax to a Land Value Tax: Replication and Extension of England and Zhao," National Tax Journal 61(4), Part 1, pp. 593–607. wiki summary — used for the contrasting Roanoke, VA result and for confirming the Dover regressivity finding via an independent secondary source that cites the original paper directly.
4. Common Wealth Canada, "Assessing the Distributional Impacts of a Land Value Tax," 2024. wiki summary — used for the parallel later finding that LVT-alone incidence can be regressive absent a rebate/credit, and that a flat credit restores progressivity.

Verification note (2026-07 sweep): the full 24-page Lincoln Institute working-paper PDF was directly fetched and read this pass, confirming the parcel count (5,250 single-family homes), Zhao's affiliation (Ohio State University), the Dover/2002 scope, and the regressivity and credit-offset magnitudes quoted above. The paywalled published National Tax Journal version (journals.uchicago.edu) still returns HTTP 403 to automated fetches, but the working paper is the same study and is the copy this page relies on.