There’s more data than ever, so there are more ways than ever to draw brightly colored maps of supposed transit facts. But that means it’s also easier than ever to take common types of confusion about transit and make them look like the outcomes of analysis. Your results will always reflect your assumptions, and a lot of transit analysis is still built on common mistakes that are completely obvious if you stop to think about them.
Case in point today, US Department of Transportation wants to undertake a National Transit Map project. This seems to mean drawing all the nation’s transit data feeds into a national database, which is certainly a good thing. But everything depends on the assumptions being made, and the initial video — recommended on Twitter by its narrator, Dan Morgan of USDOT — is not encouraging. The big mistakes can all be found in a 3-minute stretch starting at 5:00. Here’s the video.
The three big mistakes are:
- Implicitly confusing land area with population in visual representations. Starting at 5:00, the video presents a map of intercity access by car and train. The “discovery” from this analysis is that Amtrak doesn’t stop for several hundred miles as it crosses West Texas. It looks like a gap on the map, but it’s not a gap in reality because there are almost no people there. Of course, people draw maps like this all the time (we’ll see a lot of them during election season) but good analysis provides some visual cue to caution the viewer that land area, which is what jumps out on maps, has nothing to do with people. For example, this map could have been superimposed on a map of population density.
- Assuming that having transit nearby is more important than transit being useful or liberating. At 6:00 we see a map of the part of Salt Lake City in walking distance to transit, showing obvious gaps.This makes the previous mistake (there’s no indication of whether anyone lives or works in those gaps) but more importantly, it gives the impression that the primary problem with transit is that it doesn’t cover more area. In actual transit systems with fixed budgets, the area you cover will be inversely related to the frequency, speed, and reliability you can offer, which means that a transit agency that spreads itself thin tends to offer services that are useless to almost everyone. This geometric fact is the basis of the ridership-coverage tradeoff problem. When we see analyses that imply that transit’s problem is that it doesn’t go enough places, we need to recognize this as implying an advocacy of coverage over ridership, and more generally an advocacy of spreading service so thin that none of it is useful to most people’s lives.
- Focusing on peak hour service when discussing the access needs of poor people, even though most low-income people need to travel at all times of the day, evening and weekend. Starting at 6:55, we get an analysis that identifies poor people who do not have good access to rush hour transit. Poor people are rarely rush-hour commuters and they go many places other than the downtowns on which most peak service focuses.
I’m sure the analysts behind these examples thought that they were simplifying in a useful way. In my consulting work we simplify all the time. But we are careful to simplify in the direction of clarity about reality, rather than in the direction of helping people feel good about their often-false assumptions. The simplifications in this video — and of so much transit analysis still — are of the latter kind.