Mapping the Subtle Science of Parking Demand
In the absence of nuanced data about how people drive and live in different contexts, city regulations and building developers generally assume that people need way more parking than they really do. A new downtown condo winds up with too much parking. A multifamily apartment near a subway stop gets way too much. Even a low-rise suburban development likely doesn't need quite such a sea of asphalt.
But we overbuild parking because that's the way we've been doing it for years, and because the people who finance new developments fear what sounds like a risky investment: the transit-oriented development with hardly any parking at all.
The alternative isn't simply to err on the other end of the spectrum; underestimate parking demand, and you wind up with equally real complications (and angry neighbors). So how do cities and developers get this right? How could they treat parking demand as a subtle science – with overlapping variables based on land use, transit access, demographics, jobs, rent pricing – instead of as an assumption?
The King County Metro Transit agency in the Seattle region has been working on this, with the help of the Center for Neighborhood Technology and the Urban Land Institute Northwest. They've spent the past year trying to measure exactly which factors dictate residential parking demand around the region, in downtown Seattle, in urban neighborhoods, in the suburbs and even farther out. The result of their efforts is this Right Size Parking Calculator web application that can estimate parking demand down to a single parcel of land (and that should be replicated in other cities):
That map is based on a model developed through a painstaking survey of 220 representative multifamily buildings from across the region. Researchers went into each of them to count cars in the middle of the night (peak residential demand, as it's defined by the industry, occurs between midnight and 5 a.m.).
"On average, we saw what we assumed, which was that parking was overbuilt," says Daniel Rowe, a transportation planner with the King County Department of Transportation. "Before this project started, we had heard from a lot of different stakeholders in the area that parking was being overbuilt, but we didn’t want to make that assumption. That's the whole reason we did all this data collection – we wanted to verify it."
On average, these buildings were supplying about 1.4 parking stalls per housing unit; residents were only using about 1 stall per unit. And that oversupply extended across the region, from the central business district to urban neighborhoods to the suburbs. The project also collected information from each of these buildings on how the parking was priced, how the rental units were priced, and whether those two costs were bundled together. All of that information from this building survey was then used, alongside data on land use, demographics, job locations, and transit to hone a model capable of estimating the parking demand on a given property, accounting for factors like its proximity to transit and the price of parking relative to rent.
A $100 monthly parking fee doesn't deter parking in a building full of $2,000 a month luxury apartments quite like it does in a complex intended for affordable housing. Unbundle the cost of parking anywhere, and it also looks more expensive than it does when it's nominally free (but drives up the price of your rent). All of these subtle cues, in addition to the more obvious factors like transit access, influence demand.
The web tool enables developers or curious residents to change the specifications on a particular parcel (adjust the rent, the square footage and the number of units), as well as other characteristics about the neighborhood, and track how parking demand changes as a result.
The idea is that developers might use the tool when planning a project, but also that local governments might consider this data in updating their parking regulations. King Country is agnostic about what they come up with.
"The name 'Right Size' is very conscious because we don’t want to just say 'in all cases there should be less parking,'" says Ron Posthuma, the assistant director of the King County Department of Transportation. "In some cases, maybe there should be more."
Zoom out on this map to the entire region, and the tool also has the effect of painting larger patterns in parking demand. If the county can update the underlying data in two years, in five years, in 10 years, it may also be able to track how perceptions of parking will inevitably change with time.
"Those macro-scale trends are not surprising to us at all," Rowe says. "The real power in this model is to be at a parcel level, to be able to show it’s really not 'urban' and 'suburban.' It’s not that simple. It really matters at a parcel level, what is the transit access in walking distance to that development? What is the pricing scenario?"
What might influence a family of four more than cultural assumptions that they keep two cars in the garage?