Applying some smoothing to the data, and comparing the “1999-2001 annual average” to the “2009-2011 annual average”, I’m getting Michigan-systemwide growth in ridership of 57%. Ann Arbor is right about there, at 51%. For some reason, the Oakland County stops on Wolverine are seeing big growth in that decade: Pontiac +92%, Birmingham +144%, Royal Oak +262%. (Though the three of those together have only half as many riders as Ann Arbor.) The west side, not so much growth: Jackson +9%, Battle Creek +2%, Niles -16% — Kalamazoo posts +48% (their train station was rebuilt in 2006 as a multimodal transit center, which probably helped its appeal — New Buffalo’s rebuilt and relocated station probably also helped their 477% increase in the smoothed data.)

MDOTs rail stats site doesn’t give trip-pairs, so we can’t easily see where people are going from any given station, but since Chicago accounts for 43% of the boardings/alightings in the entire Michigan system, we can assume most trips are to/from Chicago. I’ll speculate that Ann Arbor & Oakland County’s relatively higher growth is because gas prices (and airline annoyances) are a bigger motive for seeking alternatives from Royal Oak to Chicago than from Niles to Chicago, especially as the single biggest source of delays in the system is on the south side of Chicago — 2-3 hours of delay on a 90-minute trip is a much bigger problem than on a 5 hour trip.

Chicago’s “Englewood Flyover” project, started this past fall, should clear up much of that delay for the entire Michigan system — if my guess above is correct, we’d then see more ridership growth in the west michigan Wolverine stops. Michigan’s HSR work speeding up the line east of Kalamazoo will be a bonus on top of that. (If you really wanted to dig in, you could use Amtrak’s “on-time performance” stats to get better resolution on the delays, but both the south-side-of-Chicago and the Kzoo-Dearborn segments are owned by Norfolk-Southern, with track ownership the variable that Amtrak breaks down their stats by, so you’d have to do some work to disambiguate (comparing the Pere Marquette, which uses the Chicago NS track but not Kzoo-Dearborn?), which is more work than a Saturday morning comment merits…)

I’m curious how the rise of Megabus (sounds like a Godzilla film, doesn’t it?), which caters heavily to students, has affected Ann Arbor Amtrak ridership. It’s interesting to note that the sharp increase in ridership beginning in ’02 leveled off after 2006, the year the Megabus Detroit-Chicago route began, but correlation doesn’t necessarily mean causation. As Eric and Vivienne suggest, the upcoming improvements will increase the train’s relative appeal.

I used to be a rail commuter (over an hour each way, So. California). We had many adventures, including washed-out bridges, people on the tracks (a dead body really, really slows you down), people injured trying to run through crossings, engine trouble. But it was still better than driving. (Though not cheap, it consumed about 1/4 of my salary.)

The train takes 4-8 hours (unpredictable), is not cheap, but allows me to work.

The car takes 4 hours (usually, with rare exceptions), is not cheap, but does not allow me to work.

Even if the actual train speed is not improved (and it should go up some), changing the reliability from 4-8 hours to usually 4 hours will create an entirely different dynamic and is likely to result in a real spike in usage.

Of course, all my opinion and only time will tell.