A Causal GeoAI Framework for Capturing the Unearned Increment to Self-Finance Public Infrastructure (Yılmaz, Alkan & Teke, 2026)

Summary

"A causal GeoAI framework for capturing the unearned increment to self-finance public infrastructure," by Okan Yılmaz, Mehmet Alkan, and Alihan Teke, appeared in Sustainable Cities and Society, September 2026 issue. The paper opens from an explicitly Georgist premise, in the authors' own words: a sustainable urban economy relies on "a circular value mechanism where the spatial wealth generated by the city is recaptured to finance its own infrastructure" — but complex urban market dynamics make it difficult to isolate the causal impact of any one public investment, and "massive uncaptured rent fuels private speculation and undermines local government financing."

The Sequencing Paradox

The paper's central methodological contribution is naming and addressing what it calls a sequencing paradox in self-financing infrastructure design: the land-value increment a public investment will generate cannot be reliably quantified before the investment is made (uplift is normally measured after the fact, by comparing pre- and post-investment land values), yet the fiscal instruments meant to capture that increment — levies, tax increment financing, betterment charges — require their rates or boundaries to be calibrated in advance. This is a real gap in the wiki's existing self-financing coverage: the Henry George Theorem and land value capture both describe the theoretical logic (public investment raises land rent, which can then fund the investment), and mass appraisal methods documents how assessors estimate land value generally, but none of the wiki's existing pages address the specific ex ante prediction problem — sizing a fiscal instrument before the value it will capture yet exists.

The LOIS Framework

The authors propose a causal GeoAI (geospatial artificial intelligence / causal machine learning) framework, which per the paper's results is referred to as "LOIS," to estimate infrastructure-driven land-value uplift ex ante — before the investment occurs — using geospatial and causal-inference methods rather than waiting for post-hoc before/after comparison. The paper's key result: unidimensional revenue maximisation (designing a capture instrument purely to maximize the revenue collected) exacerbates spatial inequality, whereas the LOIS framework is designed to let municipalities convert public-investment planning into a genuinely self-financing instrument that simultaneously captures the unearned rent and advances spatial justice — treating revenue capture and distributional fairness as a joint design problem rather than a revenue-first, equity-second sequence.

Relation to the Georgist Case

This paper directly operationalizes the unearned increment concept — the term appears in the paper's own title — with a modern causal-ML methodology aimed at a real implementation bottleneck: without some way to predict uplift before committing to an investment and its financing instrument, self-financing infrastructure schemes either underprice the levy (leaving rent uncaptured) or overprice it (deterring the development that would have generated the value in the first place). The finding that naive revenue-maximizing calibration worsens spatial inequality is a useful, concrete caution for any jurisdiction designing a betterment levy or land-value-capture instrument from scratch.

Nuances and Limits

• Full text not independently verified beyond the abstract. This page is built from a search-indexed abstract summary, not a direct read of the Sustainable Cities and Society article, which was access-blocked (Elsevier) to this session. Specific model architecture, data requirements, and validation results for the LOIS framework are not covered here (B-claim).
• A methodological/technical proposal, not a deployed system. The paper proposes and evaluates a framework; this page does not claim any municipality has yet adopted it in practice.
• "LOIS" as a framework name is drawn from search-indexed summary text, not confirmed as an acronym expansion or verified against the paper's own terminology — flagged for re-verification if a future session obtains full-text access.

Bears On

• Concept: Henry George Theorem — supplies a practical answer to the theorem's implicit calibration problem: how to size a self-financing instrument before the value it captures exists.
• Concept: Mass Appraisal Methods — extends ex-post assessment techniques into ex-ante causal prediction for fiscal-instrument design.
• Objection: Land Value Can't Be Assessed Accurately — a modern causal-ML approach to a specific, hard version of the assessment problem: predicting value before it materializes.

See Also

• Unearned Increment
• Henry George Theorem
• Land Value Capture
• Mass Appraisal Methods
• Public Investment Capitalizes Into Land

Sources

1. Okan Yılmaz, Mehmet Alkan & Alihan Teke (2026), "A causal GeoAI framework for capturing the unearned increment to self-finance public infrastructure," Sustainable Cities and Society, September 2026 issue, DOI 10.1016/j.scs.2026.107595. doi.org — fetch blocked (Elsevier) to this session 2026-08-27; summary drawn from search-indexed abstract text, used for the "circular value mechanism" framing, the sequencing-paradox diagnosis, the LOIS framework name, and the unidimensional-revenue-maximisation-worsens-inequality finding (C-claim; not independently verified against the paper's own text, no verbatim quotation offered beyond the two short phrases quoted above from the indexed abstract).