When we looked at some other metrics typically used in measuring housing markets, we found that they didn’t zero in, specifically, on zoning effects. Our preferred metric is land value per unit, which we’ll call Land Value Ratio.
From our discussion of construction types, recall that construction types that use less land per unit cost more to construct, while those that cost less to construct use relatively more land per unit. Another way of thinking about this: as land gets more expensive, it becomes more efficient to use more expensive construction techniques in order to use less land per unit. Cheap land: go ahead and use a lot of it. Expensive land: it’s worth it to shell out to build some elevators so you don’t have to use so much land.
Because of this relationship, if we look at the ratio of land value to number of units (or put another way, the cost of land per unit), we can expect it to stay in a fairly narrow range. It only makes sense to build more densely (i.e. use less land per unit) when land is so expensive that paying more for denser construction techniques saves enough money on land. If land costs per unit balloon very high, it’s a good sign that there’s regulatory, rather than market, reason preventing the use of denser construction techniques, because the market will typically respond to high land costs by building denser.
Let’s walk through an example. I selected four properties from Austin on sale for similar cost per unit. While we aren’t proposing a mechanism based on individual property values, these are pretty typical for their neighborhoods so you should be able to extrapolate.
|Riverside houses||Zillker houses||Eastside duplex||Downtown hi-rise|
|Land Value Ratio||$193K||$400K||$100K||$38K|
|Total / du||$486K||$516K||$547K||$635K|
The two highest “total cost per unit” in this comparison are the high-amenity, recently-built units with less land per unit. The high costs of housing are a combination of the high demand and high construction costs for high-rises. The reason the Zilker house is relatively inexpensive is that, to bring it to the same quality level as the others, a new owner would have to either heavily renovate or replace the house, adding ~$250-300K to costs. Looking strictly at total costs, it is easy to deduce that duplexes and high-rises are necessarily more expensive.
But another way of looking at this is that an old delapidated house in Zilker as expensive as a brand-new duplex or hi-rise! The Land Value Ratio isolates the happenstance of the parcels, that some were built recently and some built a long time ago) and shows specifically where zoning is driving up prices. The high-rise and new-build duplex flip from being the most expensive to least expensive, with the high-rise considerably cheaper. The Zilker house really stands out. All three of the non-highrise condos could probably stand a little bit more density (indeed, the Riverside House is actually zoned to allow greater density), but the problem area is clear. Indeed, the Land Value Ratio for the high-rise and the single-family home are more than ten times apart, leaving lots of room for measurement error while maintaining a strong signal.
This one metric is at the heart of pretty much every one of our policy proposals. This is not because all policy must stem from this. There is plenty of room in zoning codes for better and worse design guidelines, flooding controls, environmental safeguards, etc.
But on the specific question of what densities to zone for to achieve the most economical housing, this ratio has extraordinary ability to cut through the complications and give simple but sophisticated answers.