Effects of Split-Rate Taxation on Tax Base

Summary

"Effects of Split-Rate Taxation on Tax Base" is a peer-reviewed article by Zhou Yang (Robert Morris University) and Zackary B. Hawley (Texas Christian University), published in Public Finance Review 50(6): 651–679 (November 2022), DOI 10.1177/10911421221129956. It began as Lincoln Institute of Land Policy Working Paper WP21ZY1, "Split-Rate Taxation: Impacts on Tax Base" (January 2021), available via the Lincoln Institute, and updates the panel Yang first built for her 2014 Lincoln Institute working paper on Pennsylvania's two-rate property tax. Yang's research focus is land value taxation, tax competition, and state/local tax policy; Hawley's is urban economics, housing markets, and property taxation. The paper carries weight as the most direct empirical test of a specific, practically important question left open by earlier Pennsylvania split-rate studies: does shifting the tax burden toward land actually raise the property tax base — the revenue-relevant outcome for a city council deciding whether to adopt the policy — rather than merely encouraging construction activity (the outcome measured by Plassmann & Tideman and Oates & Schwab).

The Core Argument / Findings

Using panel data on all Pennsylvania municipalities' aggregate property market values (Pennsylvania State Tax Equalization Board, 1990–2018) and a separate land-value panel derived from Davis et al.'s (2020) Federal Housing Finance Agency appraisal-based land-value estimates (2012–2018), Yang and Hawley estimate the effect of a jurisdiction's land-to-structure tax rate ratio on (1) aggregate property market values, (2) market values by property type, and (3) land values specifically, using two-stage least squares with municipality and year fixed effects.

• Aggregate market values (Type B — empirical): Using a difference-in-differences specification (a policy dummy for adopting split-rate taxation) on the sample of municipalities in counties that ever had a split-rate jurisdiction, the estimated impact of switching from conventional to split-rate taxation is significantly positive, raising aggregate market values by roughly 21.5%. However, the authors caution that further changes to the tax rate ratio after initial adoption show a much smaller, often statistically insignificant, effect — the tax-base gain appears concentrated in the adoption event itself rather than scaling continuously with how aggressively land is taxed.
• Effects by property type (Type B): The market-value gain is not uniform. Using the dummy specification, residential properties rise by about 11.9% and commercial properties by about 19.9% — commercial properties benefit the most of the three types. Industrial property effects are mixed and often statistically insignificant or negative depending on specification.
• Land values (Type B): Using the FHFA-derived land-value panel (2012–2018, only 7 Pennsylvania split-rate municipalities with land-value data available), the estimated effect of a higher land-to-structure tax ratio on land values is negative but small — no more than roughly a 2% reduction in land values for a municipality taxing land at twice the millage of buildings, scaling to a maximum of roughly 8–12% at ratios of 5:1 or 7:1. The authors are explicit that "it does not appear that land values would drastically fall."
• Overall tax-base conclusion: Because building/market values rise more than land values fall, the net effect on the aggregate tax base is positive — the paper's central policy-relevant finding for municipalities considering adoption.

Relation to the Georgist Case

This paper is complementary, supporting evidence for the split-rate policy case, but it measures a different outcome than the wiki's construction-permit evidence base. Plassmann & Tideman (2000) and Oates & Schwab (1997) measure building permits and construction activity directly; Yang & Hawley measure aggregate assessed market value, a related but distinct variable that is also affected by construction, renovation, and price appreciation together. Rising market value after adoption is consistent with — and helps corroborate — the claim that split-rate taxation stimulates building, since new construction is the most direct channel by which a jurisdiction's aggregate market value could rise following a tax change that specifically favors improvements. But the paper does not itself report permit counts or units built, so it should be read as reinforcing the same directional story from a tax-revenue angle, not as a second independent measurement of construction volume. It is also notable for showing the "land values fall" concern raised by critics of split-rate taxation is empirically small under Pennsylvania's historical range of tax ratios — relevant to the split-rate taxation concept page's claim that shifting the burden toward land does not require land values to collapse.

Nuances and Limits

• Not a direct construction/permit measure. Unlike the wiki's other split-rate sources, this paper's outcome variables are assessed market values and land values, not building permits or units constructed. Readers should not treat the 21.5% market-value estimate as equivalent to a 21.5% increase in construction.
• Land-value sample is very small. The land-value regressions rely on only 7 Pennsylvania municipalities with usable FHFA-derived land-value data (2012–2018), and no municipality adopted split-rate taxation during that window — the land-value estimates are identified from a handful of municipalities that changed (mostly reduced) their tax ratios, which the authors flag as a limitation.
• Results are sample-dependent. The authors report results differ substantially across the "full sample," "split-rate counties," and "split-rate municipalities only" specifications, and explicitly note the full-sample regressions likely suffer from omitted variable bias (a counterintuitive sign on the land tax rate coefficient); their preferred estimates come from the split-rate-counties sample.
• Diminishing/uncertain marginal effect. The strong ~21.5% effect comes from the initial-adoption (dummy) specification; models using the continuous tax-rate-ratio or tax-rate-difference variables — which capture what happens as a municipality pushes the ratio further after adopting — show much weaker, sometimes null, effects. This complicates any claim that "taxing land more heavily" produces proportionally larger gains without limit.
• External validity. As with all the wiki's split-rate evidence, the empirical base is Pennsylvania municipalities — by far the largest U.S. population of split-rate adopters — so generalizing to jurisdictions with different land markets, valuation institutions, or tax administration is not established by this paper.

Bears On

• Outcome: Split-rate taxation increases urban construction — provides tax-base-level corroborating evidence (rising aggregate market values, concentrated in the adoption event) consistent with the construction-stimulus story documented by Plassmann & Tideman and Oates & Schwab, though it measures market value rather than construction directly.
• Concept: Split-Rate Taxation — direct empirical test of the concept's practical fiscal consequences (tax base and land-value effects) in its main real-world setting.
• Concept: Land Value Tax — bears on the practical question of whether shifting taxation toward land erodes the tax base via falling land values; finds the erosion is small.

See Also

• Split-Rate Taxation
• Split-rate taxation increases urban construction
• A Markov Chain Monte Carlo Analysis of the Effect of Two-Rate Property Taxes on Construction (Plassmann & Tideman, 2000)
• Land Value Tax
• Harrisburg, Pennsylvania

Sources

1. Zhou Yang & Zackary B. Hawley (2022), "Effects of Split-Rate Taxation on Tax Base," Public Finance Review 50(6): 651–679. Publisher (SAGE/DOI) — used for the published citation, abstract, and confirmation this is the peer-reviewed version of the working paper.
2. Zhou Yang & Zackary Hawley (2021), "Split-Rate Taxation: Impacts on Tax Base," Lincoln Institute of Land Policy Working Paper WP21ZY1. Lincoln Institute PDF — used directly (full text read) for methodology, data sources, all reported coefficients/percentages, caveats, and the conclusion; this is the working-paper precursor of source 1, and its full text was read in this session.
3. Florenz Plassmann & Nicolaus Tideman (2000), "A Markov Chain Monte Carlo Analysis of the Effect of Two-Rate Property Taxes on Construction," Journal of Urban Economics 47(2). Wiki summary — used for comparison/contrast on construction vs. tax-base outcome measures.
4. Wallace Oates & Robert Schwab (1997), "The Impact of Urban Land Taxation: The Pittsburgh Experience," National Tax Journal 50(1). Wiki summary — used for comparison/contrast.
5. Morris A. Davis, William D. Larson, Stephen D. Oliner & Jessica Shui (2020), "The Price of Residential Land for Counties, ZIP Codes, and Census Tracts in the United States," FHFA Working Paper 19-01 — used (via citation in source 2) for the provenance of the land-value data underlying the paper's land-value estimates.