An Introduction to Two-Rate Taxation of Land and Buildings

Summary

"An Introduction to Two-Rate Taxation of Land and Buildings" is a survey paper by Jeffrey P. Cohen (then assistant professor of economics, Barney School of Business, University of Hartford) and Cletus C. Coughlin (deputy director of research and vice president, Federal Reserve Bank of St. Louis), published in the Federal Reserve Bank of St. Louis Review, May/June 2005, 87(3): 359–374. The full text was fetched and read (the original stlouisfed.org PDF URL has gone dead and now redirects to a FRASER browse page; the paper was recovered via a mirror on scispace.com, its abstract independently confirmed against IDEAS/RePEc). The article opens by quoting Nobel laureates Milton Friedman ("the least bad tax is the property tax on the unimproved value of land") and William Vickrey ("[the property tax is] a combination of one of the worst taxes... and one of the best taxes—the tax on land or site value") as intellectual authority behind renewed interest in split-rate taxation.

The Core Argument / Findings

The paper is a literature review and policy primer, not a new empirical study, structured in four parts: an elementary economic rationale for two-rate taxation, a history of its use in the United States (chiefly Pennsylvania), a summary of existing quantitative estimates of its effects, and a discussion of the practical and political obstacles to adoption.

The theoretical case. Cohen and Coughlin explain the standard efficiency argument: because land supply is fixed, a tax on land value does not reduce the quantity supplied and so creates no deadweight loss, unlike taxes on buildings or income, which discourage construction and work. Shifting from uniform property taxation toward taxing land more heavily than structures should therefore reduce the tax system's overall distortion and could "generate additional economic activity."

The Pittsburgh experience. The paper's central empirical case study is Pittsburgh, which sharply increased its land-tax-to-building-tax differential in 1979–80. Citing Oates and Schwab (1997), the authors report Pittsburgh saw a 70.4% increase in average annual building-permit value between 1960–79 and 1980–89 — the largest of 15 comparable Rust Belt cities, 13 of which saw building permits decline over the same period (only Columbus, +15.4%, also grew). Oates and Schwab's regressions found a positive, significant post-1979 coefficient for Pittsburgh in their baseline dummy-variable specification — the one city in the 15-city sample to show one on that specification, though a second (Buffalo) also turns significant once a time trend is added — concentrated in nonresidential and office permits rather than residential. Cohen and Coughlin flag that Oates and Schwab themselves called this "surprising" against standard theory (only the land rate rose; the building rate didn't fall) and that Pittsburgh was simultaneously undergoing a major post-steel revitalization effort ("Renaissance II"), muddying causal attribution.

Other estimates. Aside from Pittsburgh, "only one comprehensive study... has attempted to identify the actual effects of two-rate taxation" — most other work estimates only the hypothetical, short-run redistribution of tax burden. The paper summarizes Bowman and Bell (2004), who used parcel-level data from three Virginia jurisdictions to compute how liabilities would shift under a pure land tax, finding owners with high land-to-improvement ratios (vacant/underbuilt lots) would see liabilities rise while intensively developed parcels would see liabilities fall.

Practical and political obstacles. Administratively, the authors report Edwin Mills's (1998) argument that separating "raw land value" from improved-property value is difficult — land is rarely observed unimproved once developable, and hedonic pricing is hard to extend to commercial parcels. On revenue adequacy, they cite Mills's estimate that even a 100% tax on land rent might raise at most roughly 1% of property value (p. 370) — insufficient to replace existing revenue — while Netzer and McGuire, citing Pittsburgh, are more optimistic. The paper also cites theory (Lee 2003; Wildasin and Wilson 1998) arguing uniform taxation can be more efficient than pure land taxation when land is partly absentee-owned or landowners value risk-pooling across jurisdictions — a real complication to the textbook "land tax is always efficient" story. Politically, a shift to two-rate taxation creates clear winners (underbuilt, high-land-value owners) and losers (intensively improved, low-land-value owners), and the losers have strong incentive to organize opposition.

Relation to the Georgist Case

This is a mainstream (non-Georgist-advocacy) Federal Reserve publication making the efficiency case for taxing land more heavily than buildings, useful precisely because it is not written from a Georgist standpoint — the authors treat two-rate taxation as one policy option among several, note real theoretical counterarguments to "land taxation is always efficient," and flag the empirical evidence as thin. It directly supports the wiki's split-rate taxation increases construction benefit page, for which the Pittsburgh/Oates-Schwab finding is already a listed source, by providing an independent secondary summary confirming how that finding has been read by later economists (suggestive but confounded by Pittsburgh's concurrent redevelopment push). It is also a useful primer for split-rate taxation and Pennsylvania, tying the concept to its main real-world US laboratory.

Nuances and Limits

The paper is a survey, not new research, and its headline empirical claim — the Pittsburgh building-permit surge — carries the same caveats Oates and Schwab themselves gave: no quasi-experimental design isolates the tax change from Pittsburgh's concurrent "Renaissance II" redevelopment campaign, the effect concentrates in commercial/office rather than residential permits, and Oates and Schwab's own regressions show this result is specification-dependent: in their baseline dummy-variable model (Model 1), Pittsburgh is the one city among the 15 in the comparison sample with a statistically significant positive post-1979 break in building permits, but adding a time trend (Model 2) also yields a significant positive break for Buffalo — smaller than Pittsburgh's, whose coefficient remains the largest under either specification. The authors are candid that comprehensive real-world tests of two-rate taxation's effects are scarce — most literature they summarize models hypothetical short-run burden shifts rather than measured construction responses. The paper does not simply repeat the textbook "land tax is neutral" claim; it presents genuine efficiency counterarguments (absentee ownership, risk diversification) that qualify when uniform taxation might beat a pure land tax — a nuance the wiki's own pages should reflect rather than flatten.

Resolved 2026-07-18: the Mills (1998) revenue-adequacy claim was re-verified directly against a full PDF of the paper (also independently recovered via the Wayback Machine's archive of the true stlouisfed.org PDF, since the live URL is dead; see Sources). Cohen and Coughlin's own wording is precisely "[a]nnual real estate taxes are 1.5 to 2.0 percent of the market value of taxable property and site values are estimated to be, at most, 10 percent of property values. If site and structure rents are capitalized at the same rate, gross-of-tax site rents are, at most, 1 percent of property values" (p. 370) — and they conclude a 100% land rent tax "would not generate sufficient revenue to replace the revenue from existing real estate taxes." This confirms the wiki's summary/paraphrase accurately reflects Cohen and Coughlin's own text, not a distorted summary. Two independent corroborations support the substance beyond Cohen and Coughlin's own paraphrase. First, Dick Netzer's own contemporaneous summary of the 1998 Lincoln Institute conference where Mills's paper was first presented corroborates the substance: "Some, like Mills, held that [land value taxation] could not even replace the conventional American property tax on land and buildings, much less a substantial portion of other state and local taxes as well" (Netzer, "Land Value Taxation," Land Lines, Lincoln Institute of Land Policy, March 1998). Second, as an independent sanity check on the magnitude, Mason Gaffney's contemporaneous rebuttal of a separate 1998 Mills piece ("Response to Edwin Mills on Land Value Taxation," 2002) records Mills asserting elsewhere that land rents are "5-7% of GDP" and that "land value is 15% of real estate value" — multiplying those two figures (roughly 15% x 6-7%) independently lands close to the same ~1% of property value magnitude Cohen and Coughlin report, suggesting the figure is consistent with Mills's other published estimates rather than a one-off transcription error. The underlying Mills chapter itself — "The Economic Consequences of a Land Tax," in Dick Netzer, ed., Land Value Taxation: Can It and Will It Work Today? (Lincoln Institute of Land Policy, 1998), pp. 31–48 — was not found free online this session (not on archive.org or Google Books; a print/paywalled Lincoln Institute title), so the specific "1 percent" figure remains sourced to Cohen and Coughlin's paraphrase rather than to Mills's own words. The claim should still be attributed to Mills via Cohen and Coughlin's secondary summary, not treated as independently confirmed against Mills's primary derivation. See sources/verification-backlog.md for the retry row.

Bears On

• Benefit: Split-rate taxation increases urban construction — this paper is an independent secondary summary and corroboration of the Pittsburgh/Oates-Schwab finding already cited on that page, and adds the cross-city building-permit comparison table (Pittsburgh +70.4% vs. 13 of 15 comparison cities declining).
• Concept: Split-Rate Taxation — supplies the standard efficiency rationale, US implementation history, and the leading practical/political objections (assessment difficulty, revenue adequacy, absentee-owner and risk-diversification counterarguments) that the concept page should represent.
• Place: Pennsylvania — the paper's central real-world case study is Pittsburgh's 1979–80 tax reform, directly relevant to Pennsylvania's role as the wiki's split-rate "natural laboratory."

See Also

• Split-Rate Taxation
• Split-rate taxation increases urban construction
• Pennsylvania
• Land Value Tax
• Henry George

Sources

1. Jeffrey P. Cohen & Cletus C. Coughlin (2005), "An Introduction to Two-Rate Taxation of Land and Buildings," Federal Reserve Bank of St. Louis Review, 87(3): 359–374. IDEAS/RePEc abstract · original PDF via Wayback Machine — used for the full text and re-verified 2026-07-18 for the exact Mills (1998) revenue-adequacy wording (the live stlouisfed.org PDF link is dead, redirecting to a FRASER browse page; the previously-cited scispace.com mirror now also 403s).
2. Wallace E. Oates & Robert M. Schwab (1997), "The Impact of Urban Land Taxation: The Pittsburgh Experience," National Tax Journal, 50(1): 1–21. PDF via cooperative-individualism.org — read directly in full 2026-08-12 (previously cited only via Cohen and Coughlin's secondary summary); used for the exact building-permit regression results in the paper's Table 4: their baseline Model 1 finds a significant positive post-1979 coefficient for Pittsburgh alone among the 15 comparison cities, while their Model 2 (adding a time trend) also yields a significant positive coefficient for Buffalo, smaller than Pittsburgh's. wiki summary
3. Fed in Print listing — used to corroborate the direct PDF URL and bibliographic details. fedinprint.org/item/fedlrv/25220
4. Edwin S. Mills (1998), "The Economic Consequences of a Land Tax," in Dick Netzer, ed., Land Value Taxation: Can It and Will It Work Today?, Lincoln Institute of Land Policy, pp. 31–48 — the primary source behind Cohen and Coughlin's "at most 1 percent" revenue-adequacy figure; not independently found free online this session (not on archive.org or Google Books), so cited here only via Cohen and Coughlin's own quoted paraphrase (p. 370), not read directly. See sources/verification-backlog.md.
5. Dick Netzer (1998), "Land Value Taxation," Land Lines, Lincoln Institute of Land Policy, March 1998 — the conference organizer's own contemporaneous summary of the 1998 Lincoln Institute conference (where Mills's paper was first presented), independently corroborating that Mills was skeptical of land value taxation's revenue adequacy. PDF — read in full.
6. Mason Gaffney (2002), "Response to Edwin Mills on Land Value Taxation" (reprinted from a Land-Theory online discussion, April 29, 2002). cooperative-individualism.org — used only as an independent sanity check on the order of magnitude of Mills's separately-published land-rent/land-value figures (5-7% of GDP; land value 15% of real estate value), not as a source for this page's substantive claims.