Is the Property Tax a Benefit Tax? The Case of Rental Housing

Summary

"Is the Property Tax a Benefit Tax? The Case of Rental Housing" is a 1994 article by Robert Carroll and John Yinger, published in the National Tax Journal, vol. 47, no. 2 (June 1994), pp. 295–316 (DOI: 10.1086/NTJ41789069). Yinger, then at Syracuse University's Maxwell School, is one of the property-tax-incidence literature's most active empirical contributors and a co-author of the standard reference work on capitalization (Yinger, Bloom, Börsch-Supan & Ladd, Property Taxes and House Values, 1988), which is itself the benchmark study cited in Palmon & Smith (1998). This paper matters specifically for the gap the other sources supporting this outcome do not close: Mieszkowski (1972) and Zodrow (2001) are general-equilibrium theory, not direct measurement, and Palmon & Smith (1998) measures capitalization into owner-occupied home sale prices, not rents paid by tenants. Carroll & Yinger instead take the incidence question directly into the rental market: using data on 147 towns and cities in the Boston metropolitan area in 1980, they estimate how rents actually respond to differences in local property-tax rates paid by landlords, holding housing and neighborhood characteristics constant. Published in the field's leading U.S. policy-oriented tax journal, this is a mainstream, peer-reviewed empirical test of tenant-side incidence — not a Georgist or advocacy publication — and it is one of relatively few studies of its kind to examine actual market rents rather than owner-occupied values or theoretical models.

The Core Argument and Findings

The paper is framed as a test of the "benefit view" of the property tax (associated with Hamilton's benefit-tax view): the hypothesis that a property tax funding local public services functions as a payment for those services rather than as a distorting tax, so that a tenant who is fully "mobile" (able to move costlessly between jurisdictions) should, in equilibrium, pay higher rent only insofar as the additional tax buys better local services they value. Carroll & Yinger set up the most favorable possible case for that hypothesis — fully mobile tenants — and test it against actual Boston-area rental data.

Using a hedonic rent equation with controls for housing unit characteristics, neighborhood quality, and local public-service levels, they estimate the direct effect of a local property-tax-rate increase on gross rent, isolated from any offsetting change in services the tax revenue buys. Their central empirical result: a $1.00 increase in property taxes is associated with a rent increase of only about $0.15, on average — implying landlords absorb roughly 85 percent of a property-tax increase directly, with only a small share showing up as higher rent. The authors report this pattern holds broadly across their sample, with the landlord's share of the tax differential remaining high (above roughly three-quarters) in most of the towns and cities studied, rather than being an artifact of a few outlying jurisdictions.

Carroll & Yinger then ask whether the indirect channel — tenants valuing the additional public services the tax revenue funds — closes the gap, which is the condition the benefit-tax hypothesis actually requires (full compensation via services, not via rent). They find it does not: even accounting for tenants' estimated willingness to pay for the resulting service improvements, the indirect, service-side benefit covers only about $0.77 of each $1.00 of additional property tax, on average, in their Boston-area data. Combining the direct (rent) and indirect (service-value) channels, the authors conclude that renters do not bear the full burden of a property-tax increase, and the property tax on rental housing is not, in practice, a benefit tax — landlords bear a substantial net burden even after crediting tenants with the value of services financed by the higher tax.

Relation to the Georgist Case

This paper supplies exactly the kind of evidence the outcome Landlords cannot pass a land value tax on to tenants had been missing: a direct empirical estimate of tax incidence in an actual rental market, rather than a theoretical model (Mieszkowski, Zodrow) or a study of owner-occupied sales (Palmon & Smith). The finding — that the great majority of a property-tax increase is absorbed by landlords rather than passed through as higher rent, and that this holds even in the "most favorable" case of mobile tenants where economic theory would predict the most shifting — is squarely supportive of the claim that landowners, not tenants, bear the incidence of a tax that falls on a fixed local asset.

Two scope qualifications matter for how strongly this paper should be read as evidence for LVT specifically, and both should be stated plainly. First, Carroll & Yinger study the conventional U.S. property tax — land and structures/improvements assessed and taxed together — not a land-value-only tax. Because a general property tax also taxes the building, which is not fixed in supply, standard incidence theory (see Mieszkowski) predicts more of it, not less, should be shiftable than a pure LVT would be; the low measured pass-through here is therefore, if anything, a conservative (harder) test that a land-only tax would be expected to pass even more decisively. Second, this is a single metropolitan area (Boston) at a single point in time (1980) — a real limitation on how far the specific $0.15-on-the-dollar and 85-percent figures generalize to other cities, tax structures, or eras; the paper should be cited for its directional finding and its rejection of full/near-full shifting, not as establishing a universal pass-through rate.

Nuances and Limits

• General property tax, not LVT. As above, the tax studied is the ordinary property tax on land plus improvements. The paper does not isolate a land-only component, so it cannot by itself distinguish "landlords absorb the tax because land is fixed in supply" from "landlords absorb the tax because of some other rigidity in this particular rental market" — though the fixed-land-supply mechanism is the one general-equilibrium theory (Mieszkowski, Zodrow) predicts should dominate for the land share specifically.
• Single metro area, single year. The 147-town Boston-area, 1980 cross-section is a real strength for internal identification (many jurisdictions with varying tax rates, comparable regional housing market) but a limit on external validity; rental markets with different vacancy rates, rent regulation, or supply elasticities elsewhere could show different pass-through.
• Benefit-tax framing, not a pure incidence study. Because the paper is designed to test the benefit view, its "indirect shifting" estimate (services covering $0.77 of $1.00) is itself a modeled willingness-to-pay figure, not a directly observed price — a further layer of estimation uncertainty beyond the direct rent-pass-through regression.
• Consistent with, but does not on its own prove, the LVT-specific claim. The paper is strong, direct evidence that rental markets in practice do not fully shift a property-tax increase onto tenants; it does not purport to model a hypothetical land-value-only tax, so a reader should describe it as strongly supportive analogous evidence rather than as a direct test of LVT incidence.
• Contrast worth noting honestly. Not all rental-market incidence evidence agrees on magnitude. Lyndsey Rolheiser's study of commercial office rents in Massachusetts (MIT dissertation, 2017/2019) finds substantially higher pass-through — on the order of 80–90% of a property-tax change absorbed into commercial rents — and Max Löffler & Sebastian Siegloch's study of German municipal property-tax (Grundsteuer) reforms reports pass-through to residential rents that grows toward full pass-through within a few years, with the share landing on renters estimated at roughly one-fifth of the welfare loss in their later synthesis. [VERIFY: this session could not directly retrieve searchable full text of the Löffler & Siegloch working paper(s) — the PDF was not machine-readable via this session's tools — so the pass-through percentages above are drawn from converging secondary/search-engine summaries rather than the primary text; a future editor with full-text access should confirm the exact figures, time horizon, and whether the German Grundsteuer is assessed on land value, on land-plus-improvements, or on a unit-value basis, since that distinction matters for how directly it bears on LVT incidence specifically.] These studies differ from Carroll & Yinger in tax base (Germany's Grundsteuer and Massachusetts commercial property taxes are not land-value-only), sector (commercial office space vs. general rental housing), country, and era, so they should not be read as directly overturning Carroll & Yinger's Boston finding — but they are a reminder that residential-rental incidence is not a fully settled empirical question across all contexts, and that pass-through appears to vary with local housing-supply elasticity and market structure. A balanced treatment of this outcome should acknowledge this variation rather than presenting Carroll & Yinger's low pass-through as the only empirical finding in the literature.

Bears On

• Outcome: Landlords cannot pass a land value tax on to tenants — the paper's central contribution: direct empirical evidence, from an actual rental market, that landlords absorb the large majority of a property-tax increase rather than passing it to tenants, filling the gap left by theory-only and owner-occupied-sales evidence.
• Research: Mieszkowski (1972), the "new view" of property tax incidence — supplies the general-equilibrium theoretical prediction (fixed-supply land borne by owners) that this paper's rental-market data are broadly consistent with.
• Research: Zodrow (2001), "A Room with Three Views" — this paper is an empirical test bearing on the benefit view Zodrow surveys as one of the three rival incidence traditions.
• Research: Hamilton's benefit-tax view — Carroll & Yinger directly test, and substantially reject, the benefit-tax hypothesis for rental housing in their Boston data, even under the most favorable (fully mobile tenant) assumptions.
• Research: Palmon & Smith (1998), property tax capitalization — the natural complement: Palmon & Smith measure capitalization into owner-occupied sale prices, while this paper measures the analogous incidence question directly in the rental market.
• Concept: Tax Capitalization — the low rent pass-through is the rental-market expression of the same capitalization mechanism documented for home sales elsewhere in this literature.

See Also

• Löffler & Siegloch — German property-tax pass-through
• Landlords cannot pass a land value tax on to tenants
• Mieszkowski (1972), the "new view" of property tax incidence
• Zodrow (2001), "A Room with Three Views"
• Palmon & Smith (1998), property tax capitalization
• Hamilton's benefit-tax view
• Tax Capitalization

Sources

1. Robert Carroll & John Yinger (1994), "Is the Property Tax a Benefit Tax? The Case of Rental Housing," National Tax Journal 47(2): 295–316. DOI: 10.1086/NTJ41789069 (paywalled via University of Chicago Press) — primary source for the paper's methodology (147 Boston-area towns/cities, 1980), the $0.15-per-$1.00 direct rent pass-through estimate, the ~85% landlord burden share, and the $0.77-per-$1.00 indirect service-value estimate; this session's sandboxed web access could not directly fetch the full paywalled text, so the findings above are corroborated across multiple independent, mutually consistent secondary summaries (John Yinger's own course notes, cited below, and several academic literature reviews) rather than quoted directly from the original. A future editor with full-text access should verify exact wording and add page-level citations.
2. John M. Yinger, "PAI735/ECN635 – Notes on the Incidence of the Property Tax," Syracuse University (Maxwell School) course materials. Yinger's site — used as a secondary summary, by one of the paper's own authors, of the Boston-area data, the $0.15/$1.00 finding, and the paper's relation to the benefit-tax and traditional views.
3. John Yinger, Howard S. Bloom, Axel Börsch-Supan & Helen F. Ladd (1988), Property Taxes and House Values: The Theory and Estimation of Intrajurisdictional Property Tax Capitalization, Academic Press — the companion capitalization study by an overlapping author team, also cited in Palmon & Smith (1998), used here for context on Yinger's broader body of incidence work.
4. Athiphat Muthitacharoen & George R. Zodrow (2012), "Revisiting the Excise Tax Effects of the Property Tax," National Tax Journal — used as a secondary cross-check that situates Carroll & Yinger's low forward-shifting estimate among other quantitative incidence studies.
5. Lyndsey Rolheiser, "Commercial Property Tax Incidence: Evidence from the Rental Market," MIT PhD dissertation (2017); summarized in MIT Center for Real Estate, "Can Landlords Really Pass on Higher Property Taxes to Tenants?" MIT CRE — used for the contrasting commercial-office-rent pass-through estimate (80–90%) noted in "Nuances and Limits."
6. Max Löffler & Sebastian Siegloch, "Welfare Effects of Property Taxation," CESifo Working Paper No. 8952 / IZA Discussion Paper No. 14195 (2021, subsequently revised). IZA · ZEW — used for the contrasting German Grundsteuer pass-through-to-rents finding noted in "Nuances and Limits." [CITATION NEEDED: this session could not extract clean, verified text from the PDF versions of this working paper — the file was not reliably machine-readable via this session's tools — so the specific pass-through percentage, time horizon, and renter/landowner/firm-owner welfare-loss split are drawn from converging search-engine summaries rather than a directly read primary text. A future editor should verify the exact figures and the Grundsteuer's assessment base directly against the paper.]

[CITATION NEEDED: a directly fetched, page-level-cited copy of Carroll & Yinger (1994) — this session's egress policy returned 403 responses for journals.uchicago.edu (the paywalled primary source), so all quantitative findings above are corroborated across multiple independent secondary sources (Yinger's own course notes; academic literature reviews citing the paper) that consistently report the same $0.15/$1.00 and $0.77/$1.00 figures and the 147-town Boston sample, giving reasonable confidence in the substance, but a future editor with full-text access should verify exact wording, page numbers, and any additional robustness results against the original.]