An August 2007 study by the Federal Transit Administration entitled “Contractor Performance Assessment Report” compared average weekday boardings for completed projects with the predictions made during the EIR process. Of 19 New Starts projects (mostly light rail), 16 had boardings below the forecasts, with some as low as 20-30 percent of forecast figures.First off, the busways in this study were either in freeway right of ways like Houston or built on an existing freight right of way as a new road like in Pittsburgh. These bear no resemblance at all to the arterial running Oakland BRT plan and should not be compared to them as much as I think these first generation busway projects show that buses are no replacement for rail.
Ridership forecasts for busways performed even more poorly, according to the report, where “none of the available busway forecasts proved to be accurate. It appears from the limited sample that forecasts of ridership on busway projects . . . will not exceed 41 percent of the forecasts.”
This is a document that was done a number of years ago but recently cleaned up and released by the Bush administration folks under Ma Peters. It was a follow up to the famous Pickrell Report which was used by wingers and libertarians alike to say that transit was worthless. But as Todd Litman notes, you can't take the start number and compare it to the end when you have design changes in the middle of the plan among other things.
Studies by Pickrell (1992) and Flyvbjerg (2005) suggested that many earlier rail transit projects exceeded projected costs and failed to achieve first-year ridership predictions. But much of what Pickrell classified as cost overruns where actually adjustments due to design changes...The FEIS numbers are also from the late 80's early 90's when ridership models for transit were still being honed due to the fact that we had just started building transit projects again after a long time off with a few metro subway lines in between. I'm not going to say things were perfect and there were some mistakes made, but I feel like now the FTA is starting to overcompensate for that. Recently ridership has been going over estimates like Charlotte, Denver, Minneapolis etc etc. In this report, the ridership estimates are extrapolated which make it look a bit worse if you look at the numbers without looking at the year they were forcast for. I'll also note that building automated guideways was a bad idea back then. 6% of ridership is really bad.
Ridership forecasts are developed to reflect trips in a particular year. For eleven of the twentyone projects included in this study, the ridership forecast year remains in the future (as of this writing).
The Capital Costs aren't anything different from what you would find with major freeway projects. Some over and some under the final estimate. But my main problem is using this report against transit at all, especially since the processes are completely different now. I know this report will get used again against transit at some point in the future, but I really wish it wouldn't, because without context, its worthless.