A construction site in Washington, DC. Image by Jordan Barab used with permission.

This is the sixth in a series of posts about how affordable housing works, by an affordable housing developer. This post explores the mismatch between the timelines for accessing affordable housing financing and acquiring land, which creates unintended barriers to development, especially in areas where land is more expensive. Read the rest of the series here.

As housing costs across the country continue to surge, one question frequently arises: Why is so little of the new housing built in American cities targeted toward renters and buyers with middle-class incomes? If the demand for middle-income housing is so strong, why aren’t developers rushing to meet that demand?

In other words, why is our new construction housing market producing so many Cadillacs and so few Model Ts? The answer lies in the economics of land valuation. In expensive cities, middle-income housing developments—new homes and apartments that residents earning about the median family income can afford—simply cannot win the competition for land.

Exclusivity drives land prices

What makes land unique compared to other goods and services is that it is exclusive. Any given piece of land has a set and limited amount of buildable space. As a result, there is a finite number of uses that can fit on the land, and the land owner or developer must choose which use to prioritize. A regular parcel can be a farm, a factory, or an office building, but it usually cannot be all three at once. Building a factory on a site generally precludes you from turning that exact same site into an office or apartment building, and vice-versa.

As a result, different uses of land are in direct competition with each other. If a vacant piece of land is up for sale, one buyer may envision building an apartment complex, while another plans to build an office building, and the third proposes a warehouse. Assuming no zoning or other regulatory barriers, the determination of who—and therefore what—prevails is made by the seller of the property.

Usually, as in most marketplaces, the seller will select the highest-price offer. The land, then, will be sold to the buyer whose planned use is able to offer the highest price. The value of anything is equal to whatever a buyer is willing to pay for it, but the implication of what a buyer of land has in mind is key. It means that only the proposed projects that are able to offer the highest price for the land will be viable.

Let’s imagine that the three hypothetical developers above want to bid on a vacant piece of land in a city. Each developer, as a potential buyer, will estimate the cost to design and construct their proposed building, as well as financing costs and projected income. The developers will treat these numbers as inputs in order to calculate a realistic price they can afford to pay for the land. The land value is the output of the developers’ calculations. The developer will then make an offer to the seller to purchase the property, based on the price they have determined will cover their costs and attract investors.

Imagine that the apartment developer calculates that, based on projected design and construction costs, rents, and operating expenses, they can pay $10 million to purchase the site and still generate sufficient returns to attract equity investment. The office developer calculates that they can pay $6 million. The warehouse developer calculates that they can pay $4 million.

When all three developers put in their bids, the landowner will, most likely, select the $10 million bid, as it offers them the most money. The property will therefore be worth $10 million. Because neither the office-building developer nor the warehouse developer can afford to pay $10 million—at that price, their projects will not generate sufficient economic returns to be financially viable—they cannot compete for this land. Thus, the land will be sold to become an apartment building.

The use of land that is projected to offer the highest price is generally referred to as the “highest and best use” for the site. This is a term used by property assessors when they are estimating the value of land in the absence of a sale (for example, during a mortgage refinancing). The term reflects the reality that in an open competition for land, there is generally a specific use that optimizes the price for the seller. Developers of projects that are not the “highest and best use” for a site generally cannot afford to buy the land.

For the apartment developer, the ideal outcome is to not offer the seller of the land $10 million but, rather, just over $6 million. At $6.1 million, the apartment developer is still making the top offer, meaning an apartment building is still the highest and best use for the site. But at this price, the financial return to investors is much greater. Yet, in a competitive market, the apartment developer knows that offering $6.1 million for the land carries a heavy risk: Another apartment developer making identical financial assumptions could outbid them for the land. This incentivizes the developer to offer $10 million to the seller and reduces the potential of financial arbitrage by acquiring land below its fair market value.

The layout of urban areas reflects this market competition

We can often see the principle of highest and best use at work in the layout of urban areas. In many city centers, high demand for office space meant office building developers could offer the highest price for land, so downtowns became predominantly office buildings. Residential projects could not compete in downtowns but did constitute the highest and best use for many adjacent neighborhoods, while industrial, distribution, agriculture, and other uses were generally pushed further out from the urban core as they struggled to compete on land value with office, residential, and retail uses in dense areas.

These land use patterns form the basis of the “monocentric city model” widely used in urban economics—for more, see Urban Economics by Arthur O’Sullivan.

There is, of course, far more to this story. Zoning laws, transportation networks, the advent and dominance of private vehicles, and many other factors affect the geospatial layout of cities. And falling demand for downtown office space could lead to profound shifts in these patterns over time. But the overall principle of market competition for exclusive land helps to explain the natural clustering patterns of similar uses that prevails in so many cities.

Identical uses also compete for land – and, for housing, this affects rents

Competition for land occurs between the same uses, too. Let’s consider three additional developers competing for the vacant piece of land. Rather than an apartment developer, an office developer, and a warehouse developer competing, imagine three separate apartment developers seeking to build apartments that, while functionally the same structure, would have different rents. As with the earlier scenario, each developer will estimate construction costs, financing costs, and income projections in order to calculate the land value they can afford to bid for the parcel.

Let’s assume the apartment building is 100 units, approximately two-thirds one-bedroom units (about 650 square feet each) and one-third two-bedroom units (about 850 square feet each). For consistency, we can assume the same design, permitting, and construction costs—as well as the same financing and operating costs, outlined in my earlier article on housing finance—with one exception: We can adjust the mortgage interest rate from 6.5% to 6% based on two presumed rate cuts from the Federal Reserve in 2024. Treating the costs to build and operate the building as fixed inputs, we can then compare the land value resulting from three different rent scenarios.

In Scenario 1, our “high-rent” scenario, let’s assume monthly rent of $5 per square foot, or $3,250 for a one-bedroom unit and $4,250 for a two-bedroom unit. In Scenario 2, our “medium-rent” scenario, let’s assume monthly rent of $4 per square foot, or $2,600 for a one-bedroom unit and $3,400 for a two-bedroom unit. Finally in Scenario 3, our “lower-rent” scenario, let’s assume monthly rent of $3 per square foot, or $1,950 for a one-bedroom unit and $2,550 for a two-bedroom unit. (No, $2,550 is not actually substantially lower-rent, but unfortunately everything is relative in high-cost cities these days.)

Based on these rents, financial calculations show that each developer can pay the following amount for the land:

  • Scenario 1, “high-rent”: $12.55 million
  • Scenario 2, “medium-rent”: $350,000
  • Scenario 3, “lower-rent”: -$11.85 million

These differences in rent cause enormous differences in land value. The developer seeking to build Scenario 1 can offer a robust land price to the seller. But the developer of Scenario 2 can only afford to offer a much lower price, and the developer of Scenario 3 cannot afford to build even if the land were free. The project in Scenario 3 would need not only free land, but also subsidized financing, to build units at the proposed “lower-rent” level.

Therefore, Scenario 1 is the “highest and best” use of the site. In an open competition to purchase the land, Scenario 1 will win out. To compete with the developer of Scenario 1, the developers of Scenario 2 and Scenario 3 would have to offer the seller the equivalent price, $12.55 million, for the property. But if they pay that for the land, their projects are not financially viable. Thus, the developer of Scenario 1 will be able to build their project, and Scenario 2 and Scenario 3 won’t happen.

The highest bidder wins, but housing shouldn’t be a zero-sum game

This “highest and best use” dynamic largely explains why developers are not able to produce more middle-income housing in urban areas. Even relatively small differences in rents can cause huge differences in land value. High demand for housing in desirable neighborhoods leads to higher rents, and higher rents make it impossible for middle-income housing developers to compete for land. This essentially guarantees that developers building in those neighborhoods will predominantly produce high-rent housing. So it is no surprise that cities like DC that have seen huge rent surges in recent years are mostly only building relatively high-end new housing.

Giving up on new construction, however, would only exacerbate this dynamic. Creative policy solutions, like land-value taxation, can certainly play a role. But increasing the supply of housing is the most important way to curtail rent growth, as increasing supply to meet demand helps to keep prices in check and reduces the scarcity of housing that drives increases in land value.

With more available homes, developers would be able to build more lower-rent projects, since developers of lower-rent projects could compete more successfully to buy land. This would begin a virtuous cycle in which increasing supply lowers rents, which would drive down land values and enable more middle-income housing projects to move forward, which would further drive down housing costs for people seeking to rent or buy homes.

When residents oppose new construction in their neighborhoods to fight “luxury housing,” they are fostering the environment in which those very high-rent projects thrive.

In debates about housing policy, casual observers often assume that income-restricted, subsidized housing and market-rate housing are in competition, and that decision-makers must choose to prioritize one over the other. But the dynamics of land prices show this is not the case. Yes, when looking at a single parcel, usually only one use can be selected for any given site. But if cities like DC enact land use reforms to significantly increase the number of available homes, that will create downward pressure on land values, helping both middle-income and lower-income housing projects to move forward more often.

While individual land parcels may be exclusive, there is no need to approach housing policy as a zero-sum game.