How the New York Times Got It Wrong

All the ways that more sprawl is not the answer to America's housing crisis.

Last week the New York Times published an article titled, Why America Should Sprawl, by Conor Dougherty, in which the argument is made that the only way we will be able to build an adequate supply of new housing to meet demand is if cities across North America continue to expand outward. The problem is that he’s suggesting we should double down on the very same approach that has contributed to the crisis we’re in today.

Too Few Options

Halfway through his article, Dougherty claims “Even if all the regulatory restraints were removed tomorrow, developers couldn’t find enough land to satisfy America’s housing needs inside established areas.”

We did an analysis of “established areas” in one county in Michigan and found that there was a shocking surplus of vacant and underutilized land, with access to existing sewer and water infrastructure, and along major corridors. Here’s what we found:

This map illustrates all of the vacant land that is currently zoned for residential use across Kent County, Michigan (one of the fastest growing counties in the State and home to Grand Rapids). Kent County has a defined need for more than 35,000 new housing units right now - this is the gap between what is available and what would be needed to provide an adequate housing option for every potential household in the County. A total of 70,000 new housing units are projected to be needed by 2050 to accommodate the continuing growth of the region.

There are currently more than 129,000 acres of vacant land in the county, most of which is zoned to require large minimum lot sizes (1 - 2 acre minimums or greater). Most of that land is outside of the areas that are served by existing sewer and water infrastructure. So, if we were going to meet our housing needs using that land, Kent County would have to sprawl beyond its existing urban footprint.

We wanted to fully account for the cost of that sprawl, so we analyzed the likely land use patterns that would be required and the amount of infrastructure that would have to be installed.

We learned from local engineers and the County Road Commission that the combined cost of running new water and sewer lines and building a road on top of them is roughly $1,300 per linear foot (assuming a two-lane road). The average required street frontage per home on all of the vacant residential land in the county is 78 feet. So, the likely cost of new infrastructure per home under the sprawl model is a little over $100,000. That’s not including the cost of the land and before we have started to build a house.

Using the above costs, we created some back-of-the-napkin calculations to estimate the total amount of additional infrastructure that would be required to support pushing all of our new housing out to the fringes. The rough, and likely very conservative estimate, is $7.8 billion of new infrastructure. This analysis does not account for the cost of losing ALL of the agriculture, forests, and wetlands in these spaces.

In the next stage of our analysis, we reviewed the land use patterns along 32 major corridors in the county that were well served by sewer and water infrastructure. These places could easily accommodate more growth if there was vacant or underutilized land that was appropriately zoned. What we found was shocking.

Along just 32 corridors, we were able to find thousands of acres of vacant and underutilized land that were already served by sewer, water and roads. The idea being that we could create a fairly simple plan to allow for unused parking lots and functionally obsolete buildings to be redeveloped into something more productive for the community.

The above is a pretty typical suburban corridor. There are remnants of what was once a Main Street environment on the left side of Lake Michigan Drive. On the right there is an oil change spot, a self-serve car wash, and a strip mall that has been 70% vacant for almost a decade.

Below is an illustration of how this same street segment on one corridor could be transformed.

The next image illustrates the economic impact of this transformation.

By enabling the redevelopment of a dying strip mall, this suburban community could increase its tax revenue in this one spot by a factor of 70x while creating at least 800 new homes on existing infrastructure.

Replicable Models

The above model is highly replicable across hundreds of places in this one County. It doesn’t just provide for more housing options in place-based neighborhoods while creating exponentially more tax revenue for the local community. It also significantly reduces pressure on outlying regions of the county that do not want growth, but instead want to preserve a rural way of life. This model also offers a much more environmentally sustainable development pattern that, if repeated several times along a single corridor, supports viable public transit options. Attached housing is also generally 40% more energy efficient than detached housing.

Finally, and perhaps most importantly, this pattern of development actually solves for the existing mismatch in the market. We know from countless studies that nearly 80% of renters are looking for walkable and amenity-rich neighborhoods, and about 50% of homebuyers want the same. Yet, only about 3% of all neighborhoods in our region are actually walkable places with local amenities nearby and these are quickly becoming the most expensive places to live due to their scarcity.

Maybe we should simply be building more walkable places on existing infrastructure and offer the choices that people say they want.

In an upcoming post I’ll dive into some of the systems change that will be necessary to build these types of neighborhoods at-scale.