Spoiler alert: it’s SLOW.
Over the last few days, I put together a Microsoft Excel file listing all Amtrak routes; for each route, I listed the scheduled end-to-end travel time (in hours), travel time in motion (also in hours; this is the end-to-end travel time minus all scheduled layovers), average end-to-end speed (in MPH), and the average speed after discounting the layovers (also in MPH). I used the latest schedules for each route, as of May 31st. (The Pacific Surfliner has a new schedule going into effect June 6th; I used that schedule for those trains.)
Click to download the chart! (.xlsx file; 47 kB.)
The fastest Amtrak route, to the shock of no one, is its flagship Acela Express; several trains operating from New York to Washington average just over 78 MPH end-to-end (discounting very brief layovers at Philly, this average rises slightly to about 79.4 MPH). The Acelas from Boston, however, have far lower average speeds (max average end-to-end speed just under 69 MPH; about 71.6 MPH after discounting layovers).
Compared with other rail lines the world over – many of which have average speeds in excess of 100 MPH – the Acela “Express” is a slowpoke.
And if Amtrak’s Acelas are slow, then much slower are the rest!
Indeed, the average end-to-end speed of all Amtrak trains across the nation is a meager 53.24 MPH (standard deviation 9.87 MPH); the average after discounting layovers is 54.9 MPH (standard deviation 10.04 MPH).
To emphasize this point, let’s consider the averages without the Acelas. For end-to-end speed, we get an average of 50.65 MPH (standard deviation 7.39 MPH); after discounting layovers, we get an average of 52.29 MPH (standard deviation 7.75 MPH).
Put another way, the Acelas add a negligible 2.6 MPH to the averages!
If this isn’t reason enough to invest in passenger rail infrastructure in America, I don’t know what is; after all, time is money – and the longer the powers that be (and the freight rail companies) put off much-needed infrastructural upgrades, the more money America leaves on the table.