Try unemployment maybe?

I've always been curious how they tally ridership, especially knowing that on very crowded trains (on the Penn line at least) the conductors are unable to collect tickets before passengers disembark.

1-3% drop on a packed train is what, 25-75 passengers? (maybe a few more on Camden which doesn't run as many double-decker cars). That's negligible in my opinion when rush hours trains are filled to capacity (even more so when running 1 car short as was the case for several rush hour trains during the last heat wave - no explain from MTA about this either).

Reporting these numbers are OK I guess, but I'd really like to see them report specifically on rush hour trains. This is where you really get a feel for regular ridership.

This is actually plausible. 2009 was an incredibly bad year for the hotel industry, and the summer of 2010 has seen a solid bounce back in hotel occupancy rates. It is quite likely that more people took vacation this July than did so last summer.

http://www.calculatedriskblog.com/2010/08/hotel-occupancy-rate-almost-back-to.html

I drive to the Greenbelt Metro Station in the summer, and take the Marc in the cooler months. The Marc trains either break down or are delayed too often in the summer.

I would actually say that the Penn line has been more crowded when I've been riding it. They've started including anywhere from 1-5 single decker cars, where it used to be all double-deckers. Maybe people have shifted to different trains to try and avoid the epecially touble-prone trains? It would be interesting to see a fully comprehensive ridership census, over a period of time, and see how ridership shifts. I don't know if it would ever be feasible to count everyone on a train, but maybe hand out flyers at MARC stations for a few days and ask people to volunteer to keep a diary of their MARC travels.

I think Bill has a good point though, when trains get slammed, no tickets get collected, and with all the headaches this summer, there were a lot of slammed trains.

Additionally, as uncontrolled as their ticket collection methods are, I wouldn't be surprised if their margin of error was 5-6%, which would render these numbers almost moot.

When is the Sun going to do a story on train conductor Spivey who saves seats for his himself and his pals and makes people sit down or calls the cops on them?

As always, it's helpful to present information regarding the methodology used to divine these figures.

I've never been on a train where the conductors used a counter to log the number of passengers / tickets collected or inspected. Is this based solely on the number of ticket sales? If so, any number of factors could account for the drop (or increase, depending on which line).

Show me a change in the number of monthly tickets sold month to month, a change in number of weeklies and one-ways.

Penn Line increase? I dunno, I saw plenty of long-haul daytrippers this summer, folks from NE MD, DE, and SE PA who rather than go on a destination vacation instead went to Washington to check out the attractions. I'd be curious to know how much of the increase were Perryville or Aberdeen to DC tickets.

Camden Line decrease? You ever wonder how much of the Camden Line ridership was heading to the stadium for the baseball game? With a team as lousy as the Orioles have been for most of the season, coupled with the recent prohibition on public transit agencies offering special game-day transportation, it'd be interesting to track the Camden ridership with the fluctuations in stadium attendance over the same time (but even then the stadium attendence is number of tix sold, not number of people who actaully show up).

@Pete: I don't know about Penn and Camden, but on the Brunswick Line, the conductors go from car to car with a hand-held counter counting the passengers on the train after the trains leave Silver Spring. Then, from each stop after that, they count the number of people who board.