Who rides expressway buses?
Let’s see if the claims really hold water.
One thing that’s been controversial on the blog is how I’ve been saying that many of the expressway public buses shouldn’t be much of a thing anymore with an expanding MRT network and a reduced emphasis on the peak hour.
Well, I better bring the receipts, then. Extraordinary claims require extraordinary evidence, and with the LTA DataMall now being a thing, it may be a good idea to show some numerical proof. There are three main corridors of expressway service to look at, from the north, east, and west to the CBD — leaving out true expresses because chances are they’re a dying breed anyway.
I’m also largely leaving out the TPE sector, since in the foreseeable future, the TPE doesn’t have an acceptable MRT equivalent, but some further optimization can be done there too. Sure, there might be the option of the CRL at Ang Mo Kio, Bright Hill or Hougang, but the jury’s still out on whether that’s good enough in place of the North Shore Line as a long term solution.
This will be a long post full of statistics. It can and will get boring; so buckle up. In the name of traceability and accountability, all datasets and code used to generate these reports are on Github and Google Drive.
The North remembers
The first group we look at is the northern group of expressway services, mostly fanning out from the last bus stops along Bukit Panjang Road.
There are two subgroups here — the first group is 187, 963, and 966, which use the BKE northbound as a shortcut between Woodlands and inner Bukit Panjang. 187 takes a direct route along Bukit Panjang Road; 963 serves Petir Road, and 966 is a special case which we will look at later. 171 also takes this route but exits at Mandai, so it still has a role to play.
In the case of 187, 963, and the northern half of 966, this may yet still be acceptable. The current alternative is a long trek by LRT out to Choa Chu Kang to access the NSL. Traffic conditions willing, it may still be time-competitive to use northbound BKE buses and then change to the MRT at Marsiling. Thus, most of these may not need to be touched for now.
This could change with the completion of the DTL2 extension to Sungei Kadut. Even if the same idea of having to take a detour still remains, it’s likely to be time-competitive depending on train speeds, with the transfer just shifting from Marsiling to Sungei Kadut. In the case of 171 it’s also quite possible that it could serve a DTL2e station as well, then perhaps on to Yew Tee or Sungei Kadut instead of going to Bukit Panjang, but I’m not sure how well this would work.
966 is the strangest route. It’s the very definition of a spoke in the hub and spoke system, where it takes the shortest direct route out of Woodlands to the BKE, takes a brief detour to Pending/Petir Roads to pick up Bukit Panjang residents, then expressways straight to Jalan Toa Payoh.
A single driver shift on 966 is also long enough that special arrangements need to be made for drivers to use the toilet (or at least, the bus company has to look the other way in the name of employee welfare). Both ends of 966 will also be served by the upcoming TEL. It thus seems obvious to state that many end-to-end 966 riders will now have the option of taking the TEL instead with a bus transfer at the Marine Parade end, making 966 a prime candidate for shortening.
Again, it’s a shame that little if any bus facilities were developed alongside the TEL — it might have been possible to send 966 into Mount Pleasant for passengers to change to the TEL there had there been such facilities.
Place your bets
The second group is what we actually want to examine — 190, 960, 972, which serve as an express corridor between the Choa Chu Kang/Bukit Panjang area and Stevens Road onward. There used to be only two, but then 972 was brought in in order to relieve 190 congestion, its meandering route making local stops in Bukit Panjang before hitting the highways. This part is important. Why?
Like 966 in Woodlands, 190 already takes as direct a route as possible from Choa Chu Kang to the BKE, so chances are people are already transferring from other services to reach 190 bus stops. While it has its own catchment along Teck Whye Lane and Bukit Panjang Road, that isn’t a lot compared to the whole Choa Chu Kang-Bukit Panjang area. I must thus assume that by providing a direct route and thus removing that need to transfer, policymakers must have thought that demand could be diverted from 190 to 972.
And they might have been right. There’s a prior example — 960 plays a similar role as 972 in providing an express bypass of Bukit Timah Road from the CBD area to Woodlands Road, with direct stops along Woodlands Road.
In a similar vein, Service 857 takes the most direct route out of Yishun to get onto the CTE ASAP — Lentor Avenue, with nothing but NSE construction sites along it. Chances are, like with 190, passengers may already be transferring from other services to get to 857 at Yishun or Khatib. It’s the only bus plying the CTE, but that’s understandable considering the mayhem that is peak-hour CTE, and that we will soon have two rail lines from the northern area to town.
You would be kidding yourself if you didn’t think these formed spokes of a hub and spoke transfer system — the same spokes as recently-developed rail services. And as I’ve mentioned several times, this forced modal split doesn’t make good use of both the rail and bus services.
So who uses these express bypasses? Here’s a route profile of Service 857, consisting of aggregates of all trip pairs that are made on the expressway section of Service 857. I’ve removed all trip pairs within the CBD/Serangoon Road areas as well as within Yishun.
That works out to a few hundred an hour. Not great, not terrible, especially when you consider that there are some industries along Yio Chu Kang Road that 857 must serve before entering the expressway proper.
What about the Bukit Timah expresses? Sure, the numbers seem high at first glance, but there’s a parallel MRT line. It might be easier to eke out the additional staffing for a marginal MRT service increase to make up the difference; albeit that’s still quite unnecessary considering our investments in automation should allow for a decreased manpower presence without compromising too much on system reliability.
The East
Even compared to some of the other services described above, Service 23 is the strangest of them all. It previously made no stops between Temasek Polytechnic/Tropica condominium and the bus stop near Kallang Basin Swimming Complex. The opening of DTL3 added a new stop at Bedok North station, but between Tampines and Bedok North it still bypasses the various private and public housing estates along Bedok Reservoir Road. By all means and purposes, that last point alone should have made service 23 an express service with the accompanying 60 cents surcharge. Yet, it is not.
As we see in the above profile, 23 fares worse than 857 with less than 400 trips per hour taken to and from town on the express sector between Tampines, Bedok, and Boon Keng. Considering that some TP students may choose to use 23 to get to Temasek Poly from the EWL, like 857 the challenge may be to find a good place to turn short the bus service. Though I guess for 23 it’s a lot easier with the upcoming Tampines North interchange.
Poor links to the Bedok Town Park stations might actually mean enough residents along Bedok Reservoir Road might actually prefer to take Service 5 to Kaki Bukit station — but with the apparent issue of poor links along DTL3, they may choose to remain on Service 5 straight to town. To me, though, even if the Rochor link were to be built, for those wanting to get to the NSL and TEL it might be an additional transfer. For that reason, there’s a case for 5 to stay. Like 966, it could also benefit from some form of transfer facility at Mount Pleasant.
While 518 also does a special job, its low frequencies (15–20 mins throughout the day) make it much less useful. The best case scenario here is that it could also see more cutbacks or other changes to focus mainly in providing bidirectional connections between Pasir Ris and Orchard Road. The Bayfront detour should also be removed with passengers going there expected to travel by train. But I wouldn’t be surprised if the whole thing went; some shifting of bus stops in the Bencoolen/SMU area could help in terms of providing connections to Orchard Road, if they’re not building low level underpasses.
36 is a bit iffy, but its time has not yet come. When the time comes, however, Changi Airport must ask hard questions about whether it still expects arriving passengers to depart the airport by private-hire vehicle or taxi. This needs to be thought of as a matter of transportation policy as we’ve seen in the past decades, and not just short-term public health considerations.
Why the need for introspection? These two maps show just how accessible the TEL is compared to the bus stops served by service 36. 36 and its various short-trip variants are the hardest-hit example of how excessive bus capacity could be caused by the opening of TEL4, but with the amount of routes passing through Marine Parade Road there’s probably going to be more. The Census named Marine Parade as the most bus-reliant planning area; TEL will change that. It’s wise to be prepared.
The greater problem
When you think about it, the numbers presented here are a drop in the ocean compared to the approximately 3 million individual bus rides taken every day. There are singular residential feeder services with much more demand than this.
One of the impacts of the bus driver shortage is that SBS Transit has likely found itself having to overstretch its workforce. This results in bus drivers not getting enough rest, sleeping at the wheel, and getting into accidents. Luckier ones that didn’t get into accidents have also realized that they’re being stretched thin by rostering practices. They thus quit, and sue for their rights as workers. SBST’s practices may be legal by the book, but do they truly ensure the welfare of the worker?
Apart from the increased chances of making fatal mistakes and wrong calls, poor worker welfare leads to high attrition rates. Twelve separate lawsuits against SBST for the same grievances should say something about working conditions. It creates a vicious cycle where the public transport industry develops a bad name as an employer and people are more willing to drive Grab instead. Likewise, with these practices leaving a bad taste in the mouth of foreign bus drivers, will we be able to attract them to come and work in Singapore? Safety, additionally, is compromised — who wants a sleep deprived driver at the wheel?
The solutions are not only on the supply side in encouraging more to join the industry by dangling financial perks. There’s probably no job loyalty anyway as people get the bonus, work a while and then leave. A final solution would be up to Bus Planning and the LTA in general to solve this on the demand side as well, by making the most out of current and upcoming rail projects. It will be quite a leap of faith to take especially with the TEL, but SBST did it in 2003; we can do it again.
The issue with data is that it tells you about the current situation now. Sure, there are ways to predict how things will change, but reality doesn’t always match predictions. The origin-destination matrix doesn’t tell you how people are affected by things like being given a choice of bus stops. I don’t think it might be possible to model the impact of providing a replacement at a nearby bus stop, or whether someone might walk somewhere else instead. Yes, I know it’s a bit inappropriate for me to say it, but this takes courage. Instead of waiting for certain parameters to be reached as one slides down the frequency-ridership spiral, it might be better to just do it all at one go instead.
Within the profiles presented, there is also some strange but explainable phenomena — lesser direct bus trips towards town, but more when heading away from town. People want a seat on a bus after a long day from home. Are there better ways to do this, though? Sure, part of it could be to surrender this sector to the private sector as well.
After all, the purpose of public transport is to remain affordable to the masses; much like supermarkets and their house-brand products. Luxuries can be provided but not at the expense of everyone else; likewise in public transport, fare increases should not be the first resort if there are other means to optimize the system and avoid unnecessary modal competition.
