Sorry for the longwinded (hopefully) non-rant:
There seems to be much discussion on this board about the need for an increase in the number of subway stations, whether they be part of line extensions further out from the city?s core, new lines or spurs serving urban areas left out of the current system, or infill stations on existing lines. Frequently the rationale is that additional stations will help foster increased development and density around the stations. In the US, I believe the DC Metro is the best example of this synergy between station and development (although I?m by no means totally won over by the type of redevelopment I?ve seen there).
A couple of questions I try to pose when people push for new stations are?why do we need to build new stations for this purpose, when plenty of the ones we already have are not fully developed, if at all? And, in the cases where stations not located at an historic node are suggested (such as Edgeworth, Assembly Square, etc), why have we had almost zero success transforming other non-nodal station locations into our own Clarendons and Bethesdas?
I undertook a quick analysis of the 39 non-central city (defined as all stations not located within Mass Ave on the Shawmut peninsula) subway stations, including only the two controlled-entry Green Line stations (Kenmore and Lechmere). I noted whether or not the stations were located at historic nodes. 23 of them were, ranging in importance from Beachmont and Wollaston to Harvard and Kenmore. The remaining 16 were located outside of traditional nodes (Shawmut, Stonybrook, Community College, etc).
I also looked at aerial photos to determine whether vacant lots or non-station oriented surface parking were present in large enough parcels for a moderately sized development within ? mile of each station. Only 5 of the 39 did not have sizable vacant parcels within a short walk (Ashmont, Shawmut, Green, Stonybrook, and Oak Grove).
Finally, I noted whether or not a station had seen some recent (10-15 years) sizable redevelopment of vacant land. 17 of the 39 stations did have some TOD redevelopment.
Findings: Of the 17 with recent TOD, 13 were stations located at historic nodes: Quincy Ctr, N Quincy, Ashmont, Broadway, Kendall through Davis, Malden, Revere Beach, Lechmere and Kenmore. Of the other 4, 2 of them probably have as much to owe to adjacent highway infrastructure as they do to subway access (Wellington and Alewife). Ruggles, while not a historic node, is probably the most urban location of the 16 non-nodal stations, and the redevelopment at Savin Hill is fairly modest.
Other than Ashmont, the 13 nodal stations where TOD has happened still have room to grow, because of the presence of underutilized parcels close to the station.
There are still 10 nodal subway stations where almost no vacant parcels have been filled in the past decade or more: Wollaston, Fields Corner, Andrew, Forest Hills, Jackson, Roxbury Crossing, Sullivan, Beachmont, Orient Heights, Maverick.
Conclusion: Even if there was conclusive evidence that the presence of a subway station in this metro area would attract the type of high density development we all like, isn?t it much more cost effective to do a little bit better with the dozens of opportunities we already have than to spend millions developing new stations. And in the absence of conclusive evidence, why do we keep convincing ourselves that redevelopment would be any more successful at new stations?