Single Tax for the Living World

Actual Modern Georgist Experiments went looking for what actually happens when a Georgist land value tax leaves the seminar room and meets a real city — Pittsburgh, Harrisburg, Estonia, Taiwan, the whole uneven record. The conclusion there was that plain land value taxation, taxing the site itself rather than what's built on it, reliably does what it says on the label wherever it's tried seriously: it discourages sitting on empty land, it encourages building, and it survives exactly as long as three unglamorous things hold — current assessments, a public that understands what it's paying, and a tax that isn't asked to solve problems outside its actual design. What that essay didn't get into is a newer, more ambitious variant, one that hasn't been tried anywhere yet outside of a computer: a land value tax whose rate bends not just around how a parcel is used, but around what it's doing to the ecosystem around it.

The idea: tax the land, then tax the damage

A team led by Dakota Walker at the University of Vermont published a proof-of-concept study in 2024 proposing exactly this — an ecological-impact-weighted land value tax, or ELVT. The starting point is ordinary land value taxation, exactly as described above: assess the land alone, separate from whatever's built on it, and tax that. What ELVT adds is a second layer of assessment sitting on top of the land-value one, scoring each parcel on two further axes. The first is ecological potential — roughly, how valuable this specific patch of ground would be to the surrounding ecosystem if it were healthy habitat, which in the model is estimated as a function of proximity to intact, conserved land. The second is ecological value — how much of that potential the parcel is actually delivering right now, given its current state.

The tax rate then moves inversely with the gap between those two numbers. A parking lot sitting right next to an intact forest has high ecological potential and low ecological value — a big gap, and a correspondingly higher tax rate than a conventional land value tax alone would charge. A parking lot out in the middle of an already-degraded industrial zone, far from anything ecologically significant, has low potential to begin with, so there's little gap to tax — it pays close to the ordinary land value tax rate. And a landowner who actually restores degraded land — converts a lawn into native habitat, say — closes their own gap and sees their tax rate fall, calibrated by the municipality to stay revenue-neutral against a conventional LVT overall. It's the same core Georgist logic as ordinary land value taxation — you're taxed on the value the community and the surrounding environment create around your parcel, not on what you personally built — just extended from "the market rewards you for proximity to a transit line" to "the ecosystem rewards you for proximity to intact nature, and the tax code should notice."

What the simulation actually found

Because nobody has run an ELVT in a real city yet, Walker's team tested it the only way currently available: a spatially explicit agent-based model, a simulated 100-by-100 grid of land parcels populated with homeowner, developer, and speculator agents making boundedly rational decisions to buy, sell, develop, and — under ELVT specifically — restore land, run for twenty simulated years across hundreds of Monte Carlo repetitions. It's worth being direct about what that is and isn't: a simulation of a hypothetical region, not a study of Portland or Pittsburgh, calibrated to plausible rule-of-thumb parameter ranges rather than to any specific city's actual land market. Read the numbers that way — as evidence about a mechanism, not a forecast for any particular place — and they're still striking.

Relative to a conventional status-quo property tax, both an ordinary land value tax and the ecological version pushed development toward denser, more capital-intensive housing and away from sprawl, in every land-use scenario the researchers tested. Total housing supply rose substantially under both tax interventions, average housing prices fell — in the paper's baseline scenario, 27.8% under LVT and 30.8% under ELVT — and the share of simulated homeowners who couldn't find housing dropped from an average of 64% under the status quo to under 30% under either tax. Land speculation, meanwhile, got substantially less profitable: speculators' average tax burden rose by 512% under LVT and 580% under ELVT relative to the status quo, and their average wealth change flipped from a modest gain to a loss under both. None of that is a surprise to anyone who read the Pittsburgh and Harrisburg case studies — it's the same mechanism, land value taxation punishing speculative land-hoarding and rewarding actual construction, showing up again in simulation.

The part that's genuinely new is what happened to ecological value specifically. Under plain LVT, the region still lost ecological value over the run — less than under the status quo, because denser development sprawls less, but real loss all the same. Under ELVT, that loss dropped to "almost neutral levels," and the amount of ecological value concentrated within the urban core rose noticeably above both alternatives. In other words: a plain land value tax makes sprawl less bad by making density more attractive. The ecological version does that and additionally makes staying ecologically intact pay for itself, which a plain LVT structurally cannot do, because a plain LVT has no mechanism that even notices whether a parcel is a parking lot or a pollinator meadow, only whether it's developed.

Where the paper is honest about its own limits

What earns this study a place in the same tradition as the Pittsburgh and Altoona case studies is that its authors don't oversell it, and I'd rather carry their own caveats forward than pretend the mechanism arrived fully validated.

First, restoration uptake barely moved even when the researchers cranked up the subsidy for it. Widening the gap between what degraded and restored land pay — which should, in theory, make restoration more attractive — produced only a slight increase in the number of parcels actually restored in the simulation. The likely culprit: as ELVT succeeded at densifying the urban core, it also shrank the number of homeowners who had enough land per unit to restore anything in the first place. A policy that works by making density cheap can end up working against its own secondary goal of encouraging on-site stewardship, simply because there's less "site" left per household once density does its job.

Second, and this is the sharper one: both tax interventions caused land values to appreciate faster, driven by all the extra development they induced. The paper is upfront that faster appreciation, even under a tax designed to claw much of that appreciation back into public revenue, can still displace existing residents — and flags this as a particular risk for ELVT given the well-documented pattern of green gentrification, where making a neighborhood more ecologically pleasant is also, reliably, how you make it unaffordable for the people who already lived there. The authors don't resolve this. They name it, correctly, as an open question their model can't answer, since it didn't build in the demand-side dynamics — rental markets, buyer bargaining power, commercial interest — that would be needed to actually test it.

Third, and most simply: this is a conceptual model with limited empirical calibration, not a policy recommendation. The authors say so explicitly in their own discussion section — implementation isn't recommended anywhere until the model is calibrated against a specific locality's real social, economic, and ecological data.

Running it back through the three conditions

The Georgism essay's closing framework asked three questions of every real-world land value tax it looked at: does assessment stay current, does the public understand what it's paying, and is the tax being asked to do only what a tax can actually do. ELVT hasn't been tried anywhere yet, so none of these can be answered from experience — but the questions themselves already tell you where the pressure points will be. Assessment currency gets harder, not easier, under ELVT, because now a municipality has to keep not just land value current but two additional, much fuzzier scores — ecological potential and ecological value — current as well, and Pittsburgh's eighty-eight-year run shows exactly how a single overdue reassessment cycle can be enough to end a tax nobody actually stopped believing in. Public understanding gets harder too, in the specific way Altoona's failure illustrates: a tax that's genuinely revenue-neutral in aggregate, but whose bill for any one household depends on a nonobvious ecological-potential-minus-ecological-value formula, is a much harder thing to explain at a town hall than "we're taxing the dirt instead of the house." And the green-gentrification risk is precisely the Harrisburg lesson in a new coat — a tax that succeeds completely at what it's built to do, densifying and de-sprawling a region while genuinely improving its ecology, can still coincide with, or even cause, a real harm the tax itself was never built to prevent.

None of that is a reason to dismiss ELVT. It's a reason to read this paper the way its own authors clearly want it read: as a serious, well-specified argument that the mechanism can work, sitting a full empirical validation short of anyone being able to say it does — and as a genuine widening of what Georgist land value taxation could eventually be asked to price, if the people implementing it are honest about the three conditions the whole family of these taxes has always lived or died by.

---

Sources: Walker, D.B.; Mertan, A.; Farley, J.; Rizzo, D.; Reynolds, T. "Land Stewardship and Development Behaviors Under an Ecological-Impact-Weighted Land Value Tax Scheme: A Proof-of-Concept Agent-Based Model," Land 2024, 13, 1795 (docs/land-evlt.pdf); model source code repository.