Since 1992, taxpayers have spent $364 billion (in 2016 dollars) on transit capital improvements. More than $257 billion of this went to rail transit, while $94 billion went to bus transit. The Antiplanner calculated this information on the Federal Transit Administration’s historic time series capital costs spreadsheet.
The official data show that transit ridership peaked in 2014 at 10.5 billion trips and by 2016 had declined 2.5 percent to 10.2 billion trips. This ridership includes urban, rural, and tribal transit agencies, but rural and tribal together add up to only about a million trips per year. The Antiplanner calculated this information on the Federal Transit Administration’s operations spreadsheet.
Tuesday’s post about the 2016 National Transit Database mentioned that the Federal Transit Administration has also posted the 2016 update to its historic time series, which has operating and ridership data back to 1991, capital costs back to 1992, and fares back to 2002 broken down by transit agency and mode. Except for the capital costs, which are in a separate file, all of the information is on worksheets that can be sorted in the same order, allowing users to make such calculations as operating cost per trip or fare per passenger mile.
I’ve modified the two files to add totals for each agency and urban area. I’ve also added GDP price deflators allowing for conversion of historic dollars to 2016 dollars.
The capital costs file is 2.5 megabytes and has raw data in rows 1 through 2515. Row 2517 has totals, 2518 has the GDP deflators, and 2519 has totals converted to 2016 dollars. Column AO also has totals in 2016 dollars for all agencies and modes.
Rows 2521 through 2539 has totals for each mode: commuter bus, cable car, commuter rail, etc. Row 2541 sums the principal kinds of rail (commuter, heavy, light, streetcar, and hybrid) and 2542 sums the various kinds of buses (commuter, motor, rapid, and trolley). Rows 2551 through 3503 provide totals for each agency and rows 3511 through 3999 has totals by urban area.
The operations spreadsheet (13.7-MB) has a dozen individual worksheets:
- Total operating costs;
- Vehicle operating costs;
- Vehicle maintenance costs;
- Other maintenance costs
- Overhead costs;
- Directional route miles of rail lines (DRM);
- Vehicles operated in maximum service (VOMS);
- Vehicle revenue miles (VRM);
- Vehicle revenue hours (VRH);
- Trips; and
- Passenger miles.
The raw data are in rows 1 through 2966. For the sake of completeness, I added the GDP deflators to row 2969 of the total operating expenses worksheet, but did not perform any calculations with them. Rows 2971 through 3923 have agency totals and rows 3931 through 4419 have urban area totals. To these rows I added a column showing the change in ridership between 2014 and 2016 and a column showing the change between the peak year, whatever it was, from 2009 to 2015 and 2016.
All of these calculations make for a very large and slow spreadsheet. If you make one change, it can take a minute or so to recalculate the entire spreadsheet. If you plan to make changes, it would be easiest to change to manual calculation and then recalculate only after you have made all the changes you want.
To sum the data by urban areas, I had to go through and correct some of the urban area identification numbers. The Federal Transit Administration assigns new numbers to each urban area after each decennial census, in rank order by population. New York has always been 1, Los Angeles has always been 2, and Chicago has always been 3. After that it gets confusing because Philadelphia was once 4 but after 2000 was bumped to 5 by Miami, which had previously been 16.
The problem is that when an agency stops operating, the FTA stops updating the urban area numbers for that agency with each new census. So an agency that operated in Miami only during the 1990s would still have the number 16 in the database while an agency that operated in Philadelphia only in that decade would still have the number 4. I went through and corrected all of these numbers in both the operations and capital cost spreadsheets. I may have missed one or two, but it should be reliable for at least the first couple of hundred urban areas.
I’ll probably refer to these data in many posts in the upcoming year. I hope you find them useful as well.