Jonathan Lamb (PhD candidate, Pardee RAND Graduate School)
Motivations
Can land use incentives stimulate more livable, equitable, and sustainable cities, and how do they compare to “top-down” design and planning?
Land value taxation (LVT) has enjoyed renewed mainstream attention in recent years, through discussions in the popular press like The Economist, Financial Times, and The New York Times, as well as proposals to enact or study some form of LVT in Michigan, Minnesota, Oregon, and California. LVT is often pitched as a tool to combat unaffordable housing crises by incentivizing new and more efficient residential construction, and as a way to curb urban sprawl by spurring development of vacant or underutilized urban lots. However, despite its popularity among theorists, only a few places have partially implemented a LVT, and voters and policy makers alike are skeptical of its overall long run costs and benefits, unintended effects, and fairness.
With support from the Robert Schalkenbach Foundation and as a PhD candidate in public policy at the Pardee RAND Graduate School, I launched this project to better understand the holistic effects of LVT on the livability, equity, and environmental footprint of cities, and how it could work with other incentives to stimulate sustainable development. Since one concern of denser cities is loss of urban green space—trees, gardens, grass, and other foliage that provide benefits like recreation, cooling, cleaner air, and aesthetic appeal—I specifically looked at how a type of policy historically used for rural and wilderness conservation called payments for ecosystem services (PES) could encourage private land owners to preserve and incorporate these important resources in new developments.
Methodology
To do so, I created the foundations of a digital “policy sandbox” to run virtual experiments in a simulated city and compare “what-if” scenarios with different policies in place. In the simulation, individual households (all renters in this version) compete for housing constructed by profit-maximizing landlords. Households are willing to pay more for homes that are larger, closer to the city center, have more nearby green space, and offer more opportunities to form social ties that provide social capital. In particular, based on some of my other forthcoming research, households see green space and social capital as complements: more of one makes the other even more valuable. Conversely, congestion from traffic to and from the city center makes an area less desirable. Each year, property and income taxes are collected and redistributed equally to all households as a bundle of public services representing things like utilities, public safety, and so on.
For my experiments, I ran my simulation many times with different policies and compared outcomes based on household well-being from different sources (green space, social capital, public services, and budget left over for other spending after housing and commuting), equity across income groups, and indicators for environmental impact like how much land was made impermeable by buildings, total commuting by all residents, and overall construction material used to support the final population.
My “control” scenario applied a traditional uniform property tax and no incentives for green space. I then tested a shift to a pure LVT (100% on land and 0% on structures), where a parcel’s land value was calculated as the net present value of revenue under its “best use” minus expected construction cost. Next, both separately and in combination with LVT, I introduced several levels of PES that paid landlords up to the full value of increased neighborhood rents resulting from any green space they provided. I analyzed these against two designs for public park space: a “green belt” around the city center and a grid of small, neighborhood parks. Figure 1 shows top-down visualizations of a city after 50 years under several policy settings.
Findings
In my simulations I found that by eliminating the tax on buildings a pure land value tax by itself did lead to more intense development as predicted by theory, but did so everywhere rather than in a more compact form. After adding a modest construction cost “elasticity,” where each project within a year gets a little more expensive due to short-term local labor and material limits, the tax surprisingly had a reverse effect of slowing new housing construction. This is because of how the tax was incorporated into the expected lifetime payoff of a project—the capitalization effect—resulting in less “slack” to overcome temporary market changes.
On the other hand, the LVT recaptured significant profits that landlords otherwise received due to desirable “spillovers” from neighboring green space and social capital. Importantly, even though land values dropped due to the capitalization effect, they did not go to zero as sometimes argued and instead settled to a new level that supported a stable and substantial increase in public services. This most benefited lower income households, for whom public services make up a larger overall share of effective wealth.
I also found that PES appears to be a plausible strategy for preserving and expanding urban green space, but the right level of subsidy is not straightforward: landlords may not respond at all below a certain cutoff (if the subsidy does not cover revenue lost from less rentable area), while too high a level can exhaust public funds on a few concentrated areas.
A more surprising result is that PES incidentally acted as demand side or “pull” force that was more effective than the LVT (as a “push”) for more compact development. This was due to the complementary effect between green space and social capital that drew residents closer together in a self-reinforcing pattern, and by acting as a “soft” growth boundary since, until housing demand reached a high enough level, landlords at the city’s edge earned more from preserving (or improving) green space than new construction.
I found that combining a pure LVT with PES created a virtuous cycle that balanced widely accessible green space, growth, and higher public service levels (these differences in amenities are plotted in Figure 2). This came at the cost of some leapfrog development and consequently a higher environmental footprint from commuting. Households also had less budget for general consumption since the better perks translated to higher rents paid. However, these downsides are partially offset by the more permeable built up area, and by options for how LVT proceeds could be redistributed.
In comparison, the public “green belt” design also performed well for median households and was less commuting intensive than my incentive-based experiments, but it led to more non-permeable development in most of the city’s extent. It was also less equitable between non-median households, and in my simulation resulted in a large displacement of low income tenants by higher earners when the belt was added, an effect referred to as environmental or green gentrification.
How should we balance these trade-offs? Figure 3 visualizes my experiments on two axes that combine several measures of well-being and environmental impact. Scenarios further to the right tend to make households personally better off, while scenarios further up have a lower environmental impact in terms of total commuting, permeability, and construction material efficiency. The plot shows that scenarios with moderate PES subsidies, a combination of LVT and PES, and the public green belt all perform best overall, but the preferred approach would in practice depend on the priorities of stakeholders. Note that experiments in Figure 3 that combined LVT with the two examples of public green space highlights how a LVT may work better when paired with other policies.
Recommendations & Next Steps
This version of the sandbox makes some simplifications common in economics research that limit the ability to test other nuances of land value taxation. For example, features like a more detailed land and materials market for landlords and deterioration of housing over time would give a better view into the role of land speculation and blight. Additionally, letting households own homes (or become landlords) and pass savings on to descendants would open exploration to the role of land value in systemic wealth inequality, a topic colleagues and I at RAND have explored in the context of the racial wealth gap.
Within these limitations, my findings support a mix of criticism and support for land value taxation. Practically speaking, the potential to unlock more efficient development is likely sensitive to local supply chain conditions and similar factors—and in the worst case, desired development could be discouraged, calling for deeper analysis of these contexts. The experiments also suggest that compact growth might be better achieved by other strategies or in combination with an LVT. However, the results re-center land value taxation as a tool to address inequities, and demonstrate the plausibility of recapturing socially-created wealth without unraveling public finance or private markets.
PES appears to be a plausible alternative to direct urban planning for providing amenities like green space, but more research is needed on how subsidies might be estimated and allocated. Future virtual experiments could distinguish types of green space (like recreational areas, tree canopy, protected habitats), mixed strategies that combine PES with public parks, and even subsidies for other kinds of urban ecosystem services such as “third places”—easily accessible venues outside of work and home where social tie formation and interactions occur.
From a broader perspective, the results invite more discussion around [1] land value taxation as a tool for addressing inequality versus shaping urban form, [2] publicly framing the distinction between socially and privately created value, and [3] how to balance and cultivate different aspects of urban well-being in a way suited to specific places and communities. I am excited to explore these and other threads more deeply in future work.
