Unleash the Network

One of the fatal flaws of mass transit planning and development is its project-oriented nature. Projects provide a well defined scope and are an efficient way of getting things done, but when every line is its own creature, the transit network as a whole can be overlooked. For example, if you are single-line-oriented, you'll probably focus on direct coverage and may allow more turns and curves than you should. However, if you are system-oriented you'll prioritize a shorter and clearer route with many connections to other lines, which will take care of the areas you "missed".

How broad is the problem? As Jarrett Walker has pointed out in an important blog post back in 2021, federal funding in the US is biased towards single-line-oriented projects (and if I may say, in other words, it is biased towards flashy ribbon cutting). He quotes from the funding criteria:

"The total employment served includes employment along the entire line on which a no-transfer ride from the proposed project’s stations can be reached."

He adds:

So all destinations that require a connection are excluded, while all destinations on the same line, even if they are an hour away, are included.

The post is great and I recommend reading it (as well as reading Jarret Walker in general).

I don't know if this is still the wording of the criterion (I hope not), but it could help explain some problems with American mass transit: with most lines leading downtown and no overall network (except for maybe in the downtown area itself where all the lines meet in a tangled mess often without a coherent hub).

The problem is that unless you want to go downtown, the system is of no use to you. Most trip destinations are usually not downtown, so it's crucial to provide service to other places as well. But you can't provide a particular solution for each person, for each potential origin and destination (like a private car or ride). Our only solution is to provide a broad transit network with efficient lines which people can transfer between to cover their particular O-D.

So what should be done? Use the full scope of transit metrics to compare alternatives and check them against the system as a whole. It is not easy and takes a lot of work, but it is the right thing to do (I'll elaborate in future posts).

In the mean time, I'll leave you with a useful technique which you can use as a rule of thumb or in case the more complex tools aren't available:

Plot your proposed transit line and stations along a simple straight line and mark the connections to other lines at each station. Then count the connections to get a feel for the connectivity the line will achieve.

How many stations have connections as a percentage of all stations along the line? How many different lines in total does the line connect to? If you have coverage data for the other lines, how much single-transfer coverage (people, jobs, etc.) does the line provide? This last question is often more important than how much coverage the line's stations themselves provide.

You can use this as a simple metric for comparison, even though it is only an indicator for the things that really interest us - travel times, reachable destinations, etc. - an not an exact calculation.

Notes:

• If you have alternatives, do the same for them and compare the results.
• If there are more planned/proposed lines, create development scenarios and take them into account.
• Different service types should be counted separately, or not at all. For example, a bus is hardly comparable to a metro line.

For example:

From this transfer scheme it is easy to tell that the Asakusa metro line in Tokyo is very well connected. Almost all the stations have at least one connection if not more.

The results:

• Stations: 20
• Stations with Transfer: 15 / 75%
• Total Transfers: 34 / 170%
• Total Unique Transfers: 20 / 100% (some lines connect at more than one station)

These metrics are amazing and can only be achieved in a highly developed network in a proper city without many suburban enclaves. But the physical circumstances (or "excuses") don't really matter for this metric, because no matter what, in any given transit system you always want high connectivity, and that should be a long term goal for any network.

[Note: I counted all rail service and didn't differentiate between metro and commuter rail because they are similar enough in terms of the service structure (in this case).]

Finally, let's see how Tel Aviv's brand new Red Line compares. I'll take into account all planned mass transit lines. The Red Line is a partly underground LRT line, 2 more LRT lines are under construction, and 3 Metro lines are to due start construction soon. I'll count connections to other LRT lines and to Metro lines, but not to heavy rail lines which are great connections but in this case should be counted seperately because they provide a completely different service.

The results:

• Stations: 35
• Transfers: 7 / 20%
• Total Transfers: 9 / 26%

Not great. Yes, there will be bus transfers, some with dedicated lanes, but it's not mass transit, not even BRT, and so it doesn't provide comparable capacity, reliability and speed. I think it's fair to say that poor connectivity is the Tel Aviv's planned system's major flaw. And it is not helped by the problematic urban development and sprawl in the region. I may go into some proposed solutions in the future if that will be made possible (I can't share the work without permission).

That is all for now! Thanks a lot for reading and please let me know your thoughts, insights, and questions.