Downs-Thomson and rapid transit in Boston

Revisiting last week's post about the Downs-Thomson Paradox in Boston, here's the charts for several rapid transit park-n-ride stations. Once again, the X-axis represents a time point in the peak morning hours, ranging from approximately 6am to 9am. The Y-axis is the number of minutes it takes to commute from the area of the named station to a spot in the Financial District; in blue by transit, and in red by driving. One big, immediately noticeable difference with the rapid transit charts is that there are many more blue dots: the reason being that the frequency of service ranges from 5-9 minutes consistently, instead of being sporadic and sparsely served like the commuter rail.

Users of the non-Green Line stations are probably not surprised to learn that the T generally gets them to work faster than driving at peak congestion. The only chart which stands out is the last one, for Riverside station in Newton. Although the "D" line is the fastest of the four branches, it still takes upwards of 50 minutes to commute from Riverside to the Financial District, but only 35 minutes by using the Mass Pike. Of those 50 minutes, the breakdown is approximately 30 minutes to Kenmore station, 15 minutes to Park Street, and then the rest for connecting or walking.

According to Blue Book 2010, Riverside saw approximately 2192 boardings per weekday in 2009. The parking lot is reported to be 93% full on average during the day. So clearly, it's being used, though not as heavily as the other stations. Could it be due to the express bus, the 500, that departs from Riverside station during rush hours and is time-competitive with driving? However, the 500 bus only saw about 47 inbound and 48 outbound boardings on a typical weekday as measured in Spring 2009. This actually makes the 500 one of the least cost-effective bus services provided by the T, at a subsidy of $6.88 per rider, despite it being fast. Strangely, even as recently as 2006, the 500 was seeing much higher ridership, getting 746 passengers on a typical weekday. Perhaps the recession damaged its customer base permanently, or else those riders have shifted to other modes.

So, while Riverside is getting customers, they don't seem to be the kind that are interested in going to the Financial District. Luckily, we can find out more specific information because a few years ago the CTPS collected detailed passenger surveys. The Origin-Destination cross-tabulation matrix for Riverside station reports that of passengers surveyed, 23.9% were going to Longwood Medical Area, 13.5% were going to Fenway, and only 7.3% were going to the Financial District. This stands in contrast with most of the other stations in this data-set, with which Financial District commuting topped all (or nearly so).

This chart shows that Green Line travel from Riverside to Longwood Medical Area is approximately equal to driving time during rush hour peak. Overall, it seems that Riverside is being used as a park-n-ride for a secondary business district much more than the CBD.  In related news, a couple years ago, a developer purchased land from the MBTA around Riverside and they are about to commence building a brand new "transit-oriented development" on that site, as this blogger has been covering nicely. It will be interesting to see who eventually moves into the new apartments, and whether it turns out to be folks working in Longwood, or a potential new customer-base for the 500 bus. (Update: The 500 bus has been cancelled as of July 2012).

Turning back to the other stations, including the commuter rail, I looked at the relative proportion of Financial District commuters compared to the quality of the public transportation trip to the same place, as measured by relative travel time. The following chart shows that distribution: trips that are faster by driving are on the left, and trips that are faster by train are on the right. The higher the point appears on the chart, the higher percentage of the riders are headed to the Financial District according to the passenger surveys. The R-squared value is 0.35 and the slope is approximately 0.37% per minute.

Despite some messiness, there does seem to be a correlation between attracting a higher share of Financial District commuters and public transportation travel time along those radial lines. This is probably in part because people choose to live where they can access convenient transportation options, and if their local station is that bad, then they will either drive to another one or all the way to work. On the other hand, some MBTA expansions have been extended to places which suffered from long driving commutes: the Old Colony lines are a prime exhibit, with the Kingston and Middleborough data points near the top of the chart.