The Cost of Lowballing Light Rail Ridership Projections

The Overhead Wire has picked up on a piece in Saturday’s New York Times
about how light rail ridership in Phoenix has exceeded expectations.
The post points out that this isn’t the first time the Federal
Transportation Administration has underestimated demand for similar
projects, a pattern that has the potential for real consequences:

3149451036_b6bac905f1.jpgLight rail in Phoenix: The demand is real. Photo by phxwebguy via Flickr.

Yay, FTA Models. You totally rule at figuring out ridership in new light rail cities. You did a bang-up job in Minneapolis (24K in 2020, current 26k), really got those Houston numbers right for 2020 (33k in 2020, current 38k), and Charlotte was right on target (9k
opening, current 14k). Note: the APTA daily numbers are a bit wonky. I
don’t know if I completely trust them to the rider but they make the
point.

Now we can add Phoenix to the list of FTA model lowballing: "The
rail was projected to attract 26,000 riders per day, but the number is
closer to 33,000, boosted in large part by weekend riders."

What
kills me about all this lowballing is what the cost-effectiveness
number was — and what it SHOULD have been. Ultimately, that is what
decides projects. And it’s a little messed up that the FTA keeps
getting it wrong, especially when they can kill a project because of a
CE below Medium.

More good stuff from around the network: Sprawled Out looks at how Panera Bread has found success by creating public spaces in suburban communities that have few or none. Kaid Benfield on NRDC Switchboard writes about how San Francisco lots left vacant by the recession might be repurposed. And The Transport Politic has the news on proposals for high-speed rail in the US from SNCF, the French national railroad operator.

  • I disagree with the conclusion. A generation of transit forecasts (mostly from the 90’s) were overestimating ridership (BART to SFO extension), which was easy fodder for transit skeptics.

    Forecasting is more of an issue of consistency then accuracy. We can always calibrate models to actual ridership, but it’s important that the models are appropriately sensitive to land use, demographic, and transit route characteristics. What we need is a consistent set of forecasting tools that will help us evaluate the benefits between two competing corridors in two different cities.

    Perhaps the formula can account for “seed” first corridors in metro areas which don’t have LRT systems. It’s easier to pivot transit forecasts for a new corridor in the same city based on existing LRT ridership than to “estimate” the theoretical user benefit in LRT-less cities such as Nashville or Columbus.

ALSO ON STREETSBLOG

Conservative Tea Party Movement Targets Florida Rail Plan

|
The conservative "tea party" movement, last seen complaining about the government-funded local transit system that they took during an anti-government march in Washington D.C, is veering back to form in Florida with an organized protest against the state’s proposal for broad new investments in rail transit. The "Tea Party" is now a registered political party […]

Living Up to the Nation’s Transit Potential

|
Today on the Streetsblog Network, we’ve got a post about unfulfilled potential — American cities that could do a lot better at getting people to ride transit. The Transit Pass compares the cities with the top ridership to those with the top population: Riders boarding light rail in Phoenix. Photo by Michael Ruiz via Flickr. […]

Will Muni Ever Consider Catenary-Free Light Rail?

|
It’s hard not to get excited about the new urban rail technology announced this week by Bombardier. Could we ever see something like this in San Francisco? Similar technology has been on Muni’s radar but the future of light rail in the city doesn’t seem to be too high on the MTA’s priority list, at […]