Jarrett Walker is an international consultant in public transit planning and policy. He writes the blog Human Transit and is author of the book Human Transit: How Clearer Thinking about Public Transit Can Enrich Our Communities and Our Lives.
When it comes to mobility, what does “adequate” service mean?
The Chicago-based nonprofit Center for Neighborhood Technology recently unveiled its AllTransit Gap Finder—an online mapping tool designed to point out areas with “inadequate” transit service. It’s a good effort, and it’s certainly good that we have more tools for understanding transit demand.
But the concept of “transit gaps” (or even worse, “transit deserts”) is less enlightening than it seems, for two reasons. First, it ignores the cost of providing transit, which has to be considered when actually doing anything about a transit gap. Second, it presents values, goals, and priorities as though they could be deduced purely from the data, which is never true.
A transit gap is some kind of difference between transit service and transit need or demand. But need and demand are different things. A need means that there are people whose lives would be better if they had transit. A demand is an indication that transit service, if it were provided, would achieve high ridership.
These terms correspond to the two opposing goals of transit service. If the goal of service is ridership, then it should provide excellent service where there is demand. On the other hand, many people who need transit wouldn’t be served if transit agencies ran only high-ridership service. So transit agencies run a certain amount of service for the non-ridership goal of coverage, which responds to need. In other words, they spread service out so that everyone has a little bit, even though low ridership is the predictable outcome. This critical distinction is explained more fully here. It’s a difficult budgetary choice about dividing resources between competing goals, one that local governments need to think about.
The notion of a transit gap confuses these two things. It will help people feel outraged about their transit service, but it won’t help local governments think clearly about what their priorities should be. As a result, it’s not likely to have much impact on the problem it claims to describe.
Let’s look at New York City to see how this works:
That’s Manhattan on the left, and the Long Island City district of Queens on the right. Blue means “adequate service” and red means “strongest transit market with inadequate transit service.”
What do they mean by “strongest transit market?” They mean density, and some demographics, but they don’t mean demand. Demand would mean that transit would get high ridership for the dollar invested. And that’s not necessarily true here.
The tool picks out the waterfronts on both sides, and Roosevelt Island at the top of the picture. These are all physically hard places for fixed transit to serve, because they do not provide the linear paths that help us combine the demands of many people onto a single line that feels reasonably direct. It’s easy to say that the south tip of Roosevelt Island has “inadequate” transit service—but it’s also a cul-de-sac. Any bus route through there will be circuitous, which means it won’t be useful to anyone but the people right there. That’s the “be on the way“ problem, and it’s a basic indicator of a transit market that will be expensive, per customer, to serve.
Just because a place shows up as a transit gap says nothing about what it would cost for a transit agency to serve it, and whether that’s a good thing to ask taxpayers to pay for compared to other areas where more people could be served at less cost to the taxpayer. Maybe it is, because transit agencies have goals other than ridership, but weighing those goals is not something any algorithm can do.
Meanwhile, many people will be shocked to discover that AllTransit thinks they have “adequate transit service.” Here’s a slice of my hometown, Portland, Oregon. Again, “adequate transit” is blue:
Downtown Portland is just off to the left, so the west end of the map is the increasingly dense and gentrified inner city. The red patch is an area with very good transit by Portland standards, and isn’t obviously different from similar areas right around it.
Meanwhile, the east end of this map is a historically neglected area of significant suburban poverty, where the city has been concentrating infrastructure investments to address issues of inequality. As a recent new bus service upgrade proved, this area produces high ridership if given frequent service, but much of it still has only half-hourly bus routes.
In short, nobody who knows Portland transit issues will find this map very helpful, and some may find it offensive. Partly this is because the blue category includes such a huge range of cases that it misses all the local distinctions that matter. But more importantly, the map is ignorant of Portland’s values, as any national algorithm will be. Portland as a whole is very concerned about its disadvantaged eastern area, much more than some other cities would be, and according to those values the transit service out there isn’t adequate at all.
No algorithm should be making claims about adequacy, because adequate is not an objective term: It depends on what a community’s goals are for transit. There are places in the U.S. where terrible transit service is considered adequate. On the other hand, by Canadian standards, and certainly by European ones, almost all Americans receive “inadequate” transit service. This is one of those times when a value judgment—what are your goals for transit? how much transit should there be? what kind of city do you want?—is being disguised as a purely scientific finding. It’s the same mistake that gives us traffic studies that “conclude” we must widen a road, as though there were no place for a full discussion of values and impacts.
Although AllTransit’s claims are framed in misleading terms, the idea of being able to accurately see exactly how well any given neighborhood is served by transit is a laudable one. Over the years I’ve written about other efforts to get this right. An especially important idea, buried deep in the overly complex methodology, is that a transit quality index should be about where you can get to in a given amount of time, rather than what transit is available. In my own work I routinely use this measure to describe the human benefits of transit service changes, because getting to destinations, and having a choice of more destinations, is what makes for a great life.
In any case, we urgently need a tool that helps people see transit quality clearly, but in a way that also helps them see their choices. This one brings us one closer, but we’re not there yet.