Out of Control Policy Blog

Phoenix Light Rail Ridership Hits Forecast, Sort Of

The New York Times recently reported on the success of the Phoenix light-rail line exceeding its forecast ridership. Of course, it's telling that the fact the light-rail line exceeded its forecast is newsworthy. Most don't. In fact, it's now considered a "success" if actual ridership falls 20% short of the forecasts used to justify the investment.

A Federal Transit Administration report published in 2007 titled "Contractor Performance Assessment" tracked 19 fixed guideway (mostly rail) transit projects. Based on ridership forecasts from the initial planning documents (Alternatives Analysis), just three (the San Diego El Cajun line, the St. Louis Initial system, and St. Louis Clair Extension) exceeded their forecasted ridership. Three others were within 20 percent of their forecasted ridership: BART Colma line, Denver's SW light rail, and the Salt Lake City south light rail line. 

Notably, the FTA report lauds the improvement in the forecasts, even though systemic biases in over estimating ridership are clear:

"Overall, almost 60 percent of the projects studied [in the FTA report] can be expected to achieve at least two-thirds of their forecast ridership by their forecast year. Comparing these results to the projects analyzed in the [earlier] 1990 study reveals that more than half of ridership forecasts for the recent projects performed better than the best forecast of the older study."

In other words, the record of forecasted ridership is still dismal--few, if any are even expected to meet their original ridership forecasts. The benchmark, in fact, appears to be whether these forecasts are within 20 percent (on fifth) of the forecast used to justify a public commitment to the project. (So, as a rule of thumb, take projected ridership and discount it by 20 percent for new projects.)

But, still, give Phoenix it's due. Average weekday boardings are running at 33,000, higher than the forecast of 26,000. Unfortunately, a few cautionary flags are evident in the actual ridership numbers. According to the New York Times:

"The light rail here, which opened in December, has been a greater success than its proponents thought it would be, but not quite the way they envisioned. Unlike the rest of the country’s public transportation systems, which are used principally by commuters, the 20 miles of light rail here stretching from central Phoenix to Mesa and Tempe is used largely by people going to restaurants, bars, ball games and cultural events downtown.

"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. Only 27 percent use the train for work, according to its operator, compared with 60 percent of other public transit users on average nationwide."

This should be troubling for light-rail supporters. Fixed guideway transit works best when it moves people from one high density place to another. That's why they tend to work better on hub-and-spoke transportation systems that key in around a downtown or Central Business District. With such a low commuter share of travel, Phoenex's light rail system doesn't appear to be competitive with its more important travel market: downtown commuters.

Samuel Staley is Research Fellow

« Tuesday Economics Stories | Main | Milwaukee County Exec Proposes Zoo… »

Comments to "Phoenix Light Rail Ridership Hits Forecast, Sort Of":

Danny L. Newton | September 22, 2009, 10:34pm | #

At 1.75 per boarding, 365.25 days per year and 40 years of income discounted at 5%, the daily fare would only pay for an initial investment of 329 million dollars. This is below the input from the local governments and about one fifth of the estimated costs. I can't figure out how they plan to pay for this at the current ticket price without socializing the costs or passing it to property owners and car related taxes. Even though the estimates were better, they still were not at the break even level. I don't see why there is such a celebration.??

Randall Scott | September 26, 2009, 3:28am | #

Compare the amount of riders & the distance to freeway lanes.

The LRT ridership is listed above as 33,000 per day.
Two freeway lanes (1 each way) can handle about 4 times that traffic.

The length of the LRT route is about 21 miles. Few passengers ride the whole distance. Let's assume that the avg trip length is 7 miles. So the avg ridership has 11,000 trips passing thru any point.

For illustration, consider the riders in relay fashion, with only 4 stops. At 1st stop, 11,000 riders get on, then 7 miles later, get off & another 11,000 get on & so on. Each 7 miles has 11,000 riders, for a total ridership of 33,000. Of course there are many more stops & the rider-miles are uneven. But still the avg ridership per route-mile is 11,000, based upon an avg 7 mile trip. If the trip avg is 10 miles, then the avg ridership per route-mile would be 16,000.

What can 2 freeway-lanes do? Maximum capacity is considered to be 2,000 vehicles per lane per hour. Consider 16 hours of major driving per day (6AM -10PM); 32 lane-hours/day. For 2 lanes, that's 12 lane-hours during commute time (6-9 & 4-7); that's 24,000 vehicles. Let's use an avg of 1,000 vehicles/lane-hour for non-rush hours (the remaining 20 lane-hours); that's 20,000 vehicles.

So, bringing that all together, 2 freeway-lanes, can comfortably handle 44,000 vehicles (continuously, at any point) in a day. If each vehicle trip was 7 miles & the total distance is 21 miles, then the total vehicles over that would be 132,000.

LRT could handle more than that, at max capacity, with 6 full cars, every 5 minutes, but that will not even come close to happening.

BTW, the whole LRT route takes over 70 minutes. Driving that distance would take less than 20 minutes. Of course the route isn't parallel, the whole way, to a freeway.

Also, the route bypasses the airport. For another ~$200 million, the route could have gone through/under the airport, w/2 stops there. It would have avoided about 4 other stops, along 2 miles of arteries, but there is far less need there.

Now, for triple that amount, a "people mover" (tax drainer) is being planned, which will take an extra ~20 minutes.

Food for thought.
Further analysis is up to you.
(This is becoming a whole article.)

DJ | October 12, 2009, 8:53pm | #

As occurs in most Reason Foundation articles, you're manipulating the facts for your anti-transit, pro Climate Change viewpoint. It's considered a success if the ridership projection comes in +-20% of what ridership was projected to be, which means that the some of the "failures" you point to such as the Salt Lake City light rail actually EXCEEDED ridership expectations by greater than 20%. Next time you should check your facts and not make assumptions from looking at pretty pictures.

Out of Control Policy Archives