After a great phone call yesterday afternoon, I realized that this is a good time to write a little more about park and rides.
In general, free parking is bad. Don Shoup’s paper (linked) is excellent and hasn’t been successfully challenged. His work has influenced parking policies in many cities, leading to improved traffic and improved economic activity with reduced emissions – basically, great stuff.
However, for transit stations outside Seattle, there seems to be a disconnect. While cities are implementing priced parking and increasingly re-purposing street parking for bicycle, transit and pedestrian infrastructure, transit agencies are still catching up. In 2008, Sound Transit had many parking garage and surface lot expansions in the Sound Transit 2 measure. Today, they’re looking at other station access options, but it’s still taken as a given by many transit supporters that park and rides are good. I used to be one of these, but now I’m not, and I want to explain why.
We often view ridership as an end goal of a transit system – it’s the metric we measure in the short term, and it’s a good indicator we’re on the right track, but getting riders on a train isn’t why we build transit. We build it for its positive impact on our economy, and because it increases quality of life, mostly as a direct result of more people walking to their destinations and fewer people driving. Sound Transit was created because the Puget Sound Regional Council – then the Puget Sound Council of Governments – adopted a policy for linking growth to transportation. Their 1990 Vision 2020, which said we needed a regional transportation body, and now their Vision 2040, are about managing growth – we’re building transit to create compact development, not to its own end.
So are we meeting those goals with park and rides?
When park and rides are built in areas where there isn’t much within walking distance, people start driving to the stations. Not all the park and ride trips are trips that were previously taken on the highway, but most of them are, hence why there are so many ready park and ride users when a new transit station opens. The day before the station opens, most were driving all the way to the city center – the parking lot fills almost immediately.
At first blush, this looks great! All these people are now on transit. But we missed something important: elasticity.
Basically, the more you charge for something, the less people buy. We deal with this in tolling all the time – if a toll goes up, more people take alternate routes. If there’s a parallel route that many people will take – enough that we wouldn’t make any more total money with a higher toll – we say that demand for that road is highly elastic. With something like healthcare, where you either spend your life savings on cancer treatment or you die, you’ll likely spend no matter what – so we say demand for healthcare is very inelastic. This goes the other way too – the less you charge for something, the more people will buy. There is, for any good or service, a price at which you make the most total money – a sweet spot.
Demand for a free highway is subject to a cost that’s like a toll – the cost of delay. People take this into account when choosing a mode of travel. This is more complex than time taken – if the cost of being late to work is high, you have to plan for the maximum likely delay, not just the average or best case travel time. Likely delay on a highway is far higher than that on a mostly grade separated train, for instance. Likely delay on Link is a couple of minutes, and on a Shinkansen it’s near zero. On a highway, the number of users is, at any given time, based on the likely delay as part of the cost of travel. Lowering the delay increases the number of users, and increasing the delay decreases the number of users – demand is fairly elastic.
When you take a thousand cars off the highway and put them in a parking lot, it decreases the delay on the highway, decreasing the cost. And because demand for that highway was relatively elastic, it increases the number of trips on that highway back to the sweet spot – back to the amount of delay most people are willing to deal with before picking another option (like moving). Getting back to the sweet spot can take months (although around here it’s nearly immediate), but the car trips we terminated at a park and ride are replaced with new car trips on the highway.
It gets worse. The trips that are brought to the highway are trips that were previously too long to take – meaning they tend to be from farther away than the trips they replace. And now, not only do you have these new, longer highway trips, but you have these trips going to the park and ride as well. In total, miles traveled by car increases, you lower local air quality, increase the demand for suburban arterials, and encourage sprawl. For every parking space we build at a transit station, we’re encouraging a new, car-oriented, suburban housing unit, demand for suburban shopping, and suburban road expansion to serve them.
There’s one more negative impact. When Sound Transit builds a $20 million park and ride, that $20 million comes with an opportunity cost of other transit capital projects. This isn’t highway money we’re spending. For instance, if South King dollars hadn’t been spent on park and rides, Sound Transit might have enough money today to build light rail to Federal Way. In East King, we might have a better, more central tunnel for East Link in Bellevue.
So what if we didn’t have these park and rides? We’d have fewer transit users at first – those who can walk to a station will, and people will still park nearby, as we deal with in Seattle near stations (and mitigate with neighborhood permits). But we cause demand for better land use! Because we’re not encouraging car-oriented sprawl to satisfy regional growth, the pressure to build housing, retail, and offices next to stations – is higher, and it happens faster. If there’s a LOT of demand, landowners nearby will offer paid parking, but we won’t have lost transit money to do it.
There’s a political cost, of course. As far as I’m aware, Sound Transit has never tried A-B testing a ballot measure with polling to compare options with and without park and rides, and some people would surely be less likely to support a transit expansion without parking. But with the same money, Sound Transit could enter public-private partnerships to develop housing or even retail next to stations (like Japanese transit agencies do), or provide better access for cycling and walking.
I believe it’s clear that park and rides are in direct opposition to the regional growth policies that led to the creation of Sound Transit, they work against the state’s climate goals, they exclude poor, young, older, and mobility impaired residents, and they encourage car ownership by the suburban poor.
It would be harder to win future ballot measures without park and rides, but all our fights for a better world are hard to win. This one is worth having.