Evidence on the Distributional Effects of a Land Value Tax on Residential Households

Summary

"Evidence on the Distributional Effects of a Land Value Tax on Residential Households" is a 2010 empirical study by Elizabeth Plummer, an accounting professor at the Neeley School of Business, Texas Christian University, published in the National Tax Journal (Vol. 63, No. 1, pp. 63–92, March 2010) — the flagship peer-reviewed journal of the National Tax Association, and the same venue that has carried the discipline's major site-value-taxation debates (see, e.g., this wiki's Brueckner (1986) and Zodrow (2001) pages). Using parcel-level appraisal data from Tarrant County, Texas (1997–2006) — a large, urbanized county that includes Fort Worth — Plummer simulates what would happen if the county replaced its uniform property tax with a revenue-neutral land value tax (LVT), and asks who would win and who would lose. Because it is a direct empirical simulation using real assessment records rather than a theoretical model or a stylized microsimulation, it is one of the more concrete pieces of U.S. evidence this wiki has on LVT incidence, and it complements the theory-driven case made elsewhere.

A related, more detailed June 2009 Lincoln Institute of Land Policy working paper by the same author, "Evidence on the Distributional Effects and Administrative Feasibility of a Land Value Tax: Who Wins, Who Loses, and Can It Happen?", covers the same Tarrant County data and adds a survey of Texas property appraisers on administrative feasibility; the published 2010 National Tax Journal article is the peer-reviewed, refined version of this work and is the version cited here. Both full texts have now been read directly for this page, and the scope differences are as follows: the 2009 working paper (Lincoln Institute product code WP09EP1) contains a cross-property-class burden analysis (its Section 3) and the appraiser survey (its Section 6 — a September 2008 mail survey of the chief appraisers of all 254 Texas appraisal districts, yielding 132 usable responses, a 52.0 percent response rate); the 2010 journal article drops the survey material entirely and instead adds a tax-capitalization analysis (its Section VI) that the working paper does not contain. The core Tarrant County distributional analyses — property-wealth groups, Suits Indices, horizontal equity, and the census-tract income evidence — appear in both versions.

The Core Findings

Plummer models a revenue-neutral shift from Tarrant County's existing uniform (improvement-plus-land) property tax to a tax on land value only, holding total local revenue constant, and compares the resulting tax burden across property classes and across residential properties of different value.

• Burden shifts away from single-family homes and onto other property classes. The central finding is that an LVT would shift the tax burden away from single-family residential properties and onto other classes of property, because homes typically carry a lower land-value share of total assessed value than many other sites. The 2009 working paper's cross-class analysis of 2006 values spells out which classes: the aggregate tax burden would rise for commercial properties (+33.7 percent), utilities (+86.6 percent), residential inventory (+167.8 percent), and — most dramatically — vacant lots and tracts (+355 percent), while falling for single-family residential (–22.1 percent) and multi-family residential (–39.2 percent) properties; industrial properties would be roughly unchanged (–4.9 percent in aggregate, with a median change of –10.7 percent but a mean of +19.2 percent). For the more recent years in the sample, the study finds the average tax liability for single-family properties would fall by roughly 30 percent, and — notably — this decline held regardless of household income, i.e., it was not concentrated only among lower- or higher-income homeowners.
• Within residential properties, only "slightly more progressive." Using Suits Indices (a standard tax-progressivity measure, analogous to a Gini coefficient applied to tax burden against a base such as income or value), Plummer finds that within the residential class, an LVT would be slightly more progressive than the existing uniform property tax. This is a modest, incremental result, not a finding that LVT is strongly or dramatically progressive among homeowners.
• Horizontal-equity problems concentrate at the low end. The study also finds that horizontal-equity problems — i.e., similarly-situated properties ending up with dissimilar tax changes — would be greatest for the lowest-valued residential properties (in the paper's analysis, properties valued at less than $40,000) relative to other residential properties. This is a caution about winners and losers within the "homes benefit" story: not all homeowners gain equally, and the properties most likely to see erratic or inconsistent treatment are cheaper homes, not expensive ones.
• Capitalization effects. Section VI of the paper estimates how the change would affect property values themselves through tax capitalization — the standard prediction that shifting tax burden onto land-value-heavy parcels lowers their after-tax value while shifting it away from others raises theirs. Under a no-reassessment scenario, the median single-family property value would rise roughly 5–12 percent depending on year and wealth group (10.7–12 percent for the highest-valued group in the early years versus 5–7 percent for the lowest-valued group, converging to about 11–12 percent for all groups in 2002 and later), while owners of the lowest-valued properties would remain the most likely to see their property lose value (36 percent of that group in 1997, falling to 15 percent by 2006). If land values are instead reassessed when the LVT is adopted, Plummer estimates that "property values would increase for over 97 percent of single-family residential properties, and the median increase would be approximately 23.5 percent," with land values falling a median of about 47 percent and building values rising about 38 percent (all under the paper's assumptions of a 4 percent discount rate and a 25-year capitalization horizon; at a 7 percent rate the median total-value gain drops to roughly 16 percent).

Relation to the Georgist Case

Plummer's study offers real-world, parcel-level American evidence that a revenue-neutral LVT would not simply penalize ordinary homeowners — indeed the opposite, since it finds the residential class as a whole would see lower average tax liability under an LVT, with the burden shifting to commercial land, utilities, and — most sharply — vacant lots and tracts. That result speaks to a common objection that LVT would fall hard on homeowners (see Objection: LVT hurts the "asset-rich, cash-poor") and is broadly consistent with the Georgist expectation that land-value-heavy, high-value commercial sites bear more of the burden than typical single-family lots.

However, the paper's contribution to the progressivity case specifically — the outcome this page is filed under, Land value tax can be progressive — is narrower and more qualified than that outcome page's headline theoretical argument. The outcome page's main support (via Schwerhoff, Edenhofer & Fleurbaey (2022, IMF)) is a wealth-concentration argument: because land ownership is concentrated among the wealthy, taxing land should fall disproportionately on high-wealth households. Plummer's paper does not directly test that cross-household wealth-concentration mechanism; instead, it finds (a) a cross-class shift (residential losing share, non-residential gaining it) and (b) only a slight improvement in progressivity within the residential class by Suits Index, alongside a new equity concern — horizontal-equity problems concentrated among the lowest-valued homes, which cuts against, rather than for, a simple "LVT protects the less wealthy" story. This page therefore supports the progressivity outcome only in a qualified sense: real assessment data show LVT is not regressive relative to the existing property tax within the residential class, and burden shifts off homes generally — but the "slightly more progressive" magnitude is modest, geographically specific (one Texas county), and comes with a specific caveat about low-value-home horizontal equity that a fuller wiki treatment of the progressivity claim should carry alongside the wealth-concentration theory.

Nuances and Limits

• Single county, one U.S. state, one tax regime. The data cover Tarrant County, Texas, 1997–2006 — a single jurisdiction under Texas's specific property-tax and appraisal rules. Results depend on Tarrant County's actual land-to-improvement value ratios across property classes in that period and may not generalize to counties, states, or countries with different land/improvement ratios, zoning patterns, or assessment practices.
• "Slightly more progressive" is a modest claim, not a strong one. The paper's Suits Index values (its Table 4) are all very close to zero: –0.0125 in 1997, rising fairly steadily to a peak of 0.0295 in 2005 and 0.0226 in 2006, against a benchmark of exactly zero for a no-exemption uniform property tax on the paper's property-value income proxy. Plummer's own reading is that "the LVT would be slightly more progressive than the property tax in years subsequent to 2000" — in 1997–2000 the index was actually slightly negative (very mildly regressive). A genuine but modest improvement, not a demonstration that LVT is broadly redistributive within the homeowner population.
• New equity problem identified, not just resolved. The finding that horizontal-equity problems concentrate among the lowest-valued properties is itself a caution meriting attention alongside the progressivity result — it suggests LVT's within-class fairness effects are uneven rather than uniformly positive for less-well-off homeowners.
• A revenue-neutral simulation, not an implemented policy. Like most LVT distributional studies, this is a simulation of what an LVT would do to a real assessment roll, not an assessment of an LVT that was actually implemented and observed over time.
• Full text now directly verified. An earlier draft of this page was built from the article's indexed abstract and secondary summaries; the complete National Tax Journal text (pp. 63–92) has since been fetched via the Internet Archive's snapshot of the NTA's former open article archive and read in full, and the findings above — the exact Suits Index values, the cross-class shifts, the horizontal-equity result, and the Section VI capitalization estimates — are now cited to the primary text. One clarification from the full read: the paper reports no household income-decile breakdowns. Its income evidence instead uses 2000 U.S. Census data for 313 Tarrant County census tracts (per capita income, median family income, poverty rate, and median property value as a lifetime-income proxy), finding median percentage tax-liability changes of –25.2 to –28.2 percent for the highest-income tracts, –22.5 to –27.7 percent for the middle group, and –20.3 to –23.3 percent for the lowest-income tracts, with Spearman correlations showing "no evidence that Median%ΔTL is related to household income level" — i.e., at the census-tract level the change would be, in Plummer's words, "neither progressive nor regressive" for Tarrant County (in contrast to Bowman and Bell's progressive pattern for Roanoke, VA).

Bears On

• Outcome: Land value tax can be progressive — provides real parcel-level U.S. evidence that a residential LVT is not regressive and shifts burden toward non-residential land, though the "slightly more progressive" finding is modest and should be read alongside the wealth-concentration theory, not as an independent strong confirmation of it.
• Objection: LVT hurts the "asset-rich, cash-poor" — the finding that residential tax liability falls roughly 30% on average, regardless of household income, is directly relevant evidence against the strongest form of this objection, though the horizontal-equity caveat for low-valued homes is a genuine complication worth noting on that page.
• Objection: LVT would hurt farmers and rural landowners — Plummer's cross-class shift finding (burden moving toward commercial land, utilities, and vacant lots — classes with high land-value shares — while industrial parcels stay roughly unchanged) is consistent with the existing case that low-land-value-share agricultural land bears little added burden, though this study does not itself analyze farmland specifically.
• Objection: Land value can't be assessed accurately — the paper's use of real land/improvement splits from a working U.S. appraisal district is a practical existence-proof that land value is already separately assessed for large numbers of parcels in at least one major American county.
• Concept: Land Value Tax

See Also

• Bowman & Bell (2008) — the replication/extension of England & Zhao on the same question
• Land value tax can be progressive
• Assessing the Distributional Impacts of a Land Value Tax (Common Wealth Canada)
• Schwerhoff, Edenhofer & Fleurbaey — Equity and Efficiency Effects of Land Value Taxation
• Objection: LVT hurts the "asset-rich, cash-poor"
• Land Value Tax
• Brueckner — A Modern Analysis of the Effects of Site Value Taxation

Sources

1. Elizabeth Plummer (2010), "Evidence on the Distributional Effects of a Land Value Tax on Residential Households," National Tax Journal 63(1): 63–92. DOI: 10.17310/ntj.2010.1.03; full text via the Internet Archive's snapshot of the NTA open article archive — full text fetched and read directly (verified 2026-07-10); used for the paper's title, venue, year, page range, and all core findings: burden shift away from single-family residential properties (taxes fall for 85.5 percent of homeowners over 1997–2006; median change –28.6 percent; ~30 percent average decrease in the more recent years regardless of household income), the Table 4 Suits Index values (–0.0125 in 1997 to 0.0226 in 2006, peak 0.0295 in 2005), the horizontal-equity concentration among properties under $40,000, the census-tract income analysis (Tables 5–6), and the Section VI capitalization estimates.
2. Elizabeth Plummer (2009), "Evidence on the Distributional Effects and Administrative Feasibility of a Land Value Tax: Who Wins, Who Loses, and Can It Happen?", Lincoln Institute of Land Policy Working Paper WP09EP1. Lincoln Institute landing page; full text PDF — full text fetched and read directly (verified 2026-07-10); confirms the same Tarrant County, TX (1997–2006) dataset as the 2010 journal article, and used for the cross-property-class shift percentages (Section 3, 2006 values) and the chief-appraiser survey details (Section 6: 254 districts surveyed, 132 usable responses, 52.0 percent response rate).
3. Economic Possibility Library (economicpossibility.org), source page for Plummer (2010). economicpossibility.org — used as a cross-check quoting the "slightly more progressive than a property tax" and horizontal-equity findings directly from the source.
4. National Tax Association / University of Chicago Press, National Tax Journal journal listing for Vol. 63, No. 1. journals.uchicago.edu — used to confirm journal identity and indexing.