A nocturnal habit

Trains don’t run at night. But what we do here might also be useful in the day.

Despite the fact that almost nobody rides them, night buses have a pretty interesting history in Singapore. Well, part of it may be due to the higher fares (even higher than express buses!), part of it may be due to the low frequencies and limited operating hours, with departures every 20–30 minutes and the last ones at 2am, and part of it may also be the fact that they only operate unidirectionally towards the housing estates.

Is there hope, though? Although I’d throw back the question and ask if one wants to believe. If you want to believe, then there is hope. If you don’t want to believe, then there won’t be.

EDIT: This post aged like milk. With the KTV cluster out there, it might be hard for the nightlife sector to regain public support, so there may not be much of a point in talking about restarting, let alone expanding night service. That said, the lessons for bridging buses and reducing engineering hours through technology still stand — lesser ECLO help late shift workers, too.

Pleading the case

We all know “Disneyland with the Death Penalty” and its implications — a clean, sanitized (did I say “squeaky-clean”?) place where all vice is behind closed doors, if not totally eradicated.

The nightlife community has struggled a fair bit under current public health restrictions; some favourite haunts have even shuttered for good. If and when the coast is clear, providing public transport at night can help their recovery; even more so since, on a technicality, nightlife also becomes more accessible to those who don’t own a car, or who may not want to be alone at night with a driver they don’t know.

But on the other hand, unlike in America where advocates constantly demand ever-later service — if not 24/7 — due to employers requiring employees to report at odd hours of the day, in Singapore many shift-work places (at least from observations) typically keep to a three-shift system with changes of shift taking place at around 7am, 3pm and 11pm — all within MRT and regular bus operating hours; which means that the social benefits of late-night service for the less well to do aren’t that great here.

The best argument that I can possibly come up with at this point, with time pressures meaning that many more economic indicators cannot be explored, is that perhaps, and only perhaps, we can keep the bars open later, and help people get home safer. An additional argument made by London, though, is that this can help drive tourism — they think the dynamism lent by a beefed-up nightlife can attract more tourists, and perhaps make it easier to get in and out of the airport at night.

Of course, it would require policymakers to shift from thinking that private hire cars and taxis will save the day (the 50% night surcharge being motivating enough), to thinking that you want these people to take public transport home. It could reduce accidents, since you wouldn’t have to take the car out to go clubbing in the first place, or leave a friend out of the fun because they’re the designated driver; it could also mean less temptation to do foolish things with the car while drunk. At the same time, with improved night public transportation, popular nightlife hotspots can also be declared car-free or car-lite precincts. Even the HK MTR has provided overnight service on major public holidays, unlike here where last trains leave at 2am.

I’m not sure how far that will get us, personally, but it’s still worth a thought.

The state of play

The relatively good news is that this may not necessarily be such a big jump, considering what already operates at night. Before the government stepped in to all but take over the industry, bus routes were a free for all — that meant that like in the daytime, SBS and SMRT Buses had their own night routes that somewhat competed against each other. Worse still, the low frequency (30–40 minutes) means these aren’t time competitive compared to private hires or driving anyway.

And much like a fair bit of the daytime network, not much has been done about the duplication inherent in this — though, as explored below, that might not be a bad thing. Although I will also argue that the current public health measures also mean night buses aren’t running today anyway. As such, with how long the routes have been hibernated, driver training would have to start afresh from zero; it would then be a good time to reboot the entire network and start from a clean slate.

Except it’s not entirely a clean slate. What still runs at night are employee bus services for employees of the bus operators. Little is known about this — and for obvious reasons, since they’re internal affairs of the bus companies, run in order to get their staff to work. And since they’re not covered under government contracting, the bus companies are free to design routes passing by where a majority of staff live, even if they wouldn’t make sense if serving the general public. The newer operators even attempted to bring back on-demand bus service for their employees.

Could we then start from here and strengthen the network such that it is actually usable for fare paying passengers? Night operation of feeder buses may be doable to get around the housing estates at night. The various workers’ transport services could then be replaced with a combination of ferry services from major transport hubs, and using these feeders which are free for public transport workers anyway.

It is easy to claim there is no demand, but this is one of the rare edge cases where it might make sense to throw more resources and hope it sticks. It could even be done with net zero deductions anyway — since most of these services might already have to have someone at the wheel. Perhaps 10 to 15 minute service all the way from midnight to 6am might be a bit too much to ask for, but maybe we can compromise and have the last trips leave at 2am or 3am, becoming either night feeders or company transports as they arrive in the housing estates.

I’ve left something out, haven’t I?

Amidst all this I’ve never really talked about the trains.

It’s not like we’ve never run the MRT at night before, but it’s not so palatable in the current-day moral panic on reliability and making every engineering hour count. Furthermore, thinking that buses can solve the problem despite their high cost (not just in operational dollars, but in manpower too) is very much against what this blog has stood for for a very long time, isn’t it? The fact of the matter is that operating the trains costs more. For now.

First, we need to reduce the need for engineering hours first. Steps are already taken in this direction, by equipping trains and the trackside with predictive condition monitoring tools that allow for monitoring to take place in service. Only after this has proven to increase engineering productivity can we consider to reduce the total amount of engineering hours available, be it through later last trains, or overnight service.

The main problem with overnight train service, is that the relatively inflexible track layout of the MRT makes it difficult for single track service to be run nightly as a long term measure. Remember the lack of diamond crossings? What this means is that in order to save space, many emergency crossovers on the MRT are facing a single direction. And should these crossings have to be taken down for maintenance work, it may also mean that the whole area has to be shut down anyway.

Long sections of single track may also limit the frequency that can be offered, to 20–30 minutes at best. While not a problem by itself, since the status quo is equally bad, you now need more people to staff the stations at night. Even with automated trains, this can be a higher cost than the bus drivers needed to operate a bus service. And then there are noise concerns to worry about from operating trains in open-air sections in the dead of night.

Learn the right lessons

Alternatively, if one really wanted to throw the baby out with the bath water, a combination of feeder routes and rail replacement to somewhat mimic MRT service could be done in the night hours. But I think this is a bad idea, since the entire concept of rail replacement, as is implemented here, has its pitfalls.

It’s important, I think, to be time-competitive. In the day, the point of travelling by train is to be time-competitive. But if we don’t want to operate the trains in the night, then the current night bus system with generous use of expressways to shorten travel time is more than fit for purpose. And if the need to stop rail service and provide replacement is there, that means people are already inconvenienced, and there should also be a priority to minimize the pain too.

The lesson here for bridging bus planners is not only to have a single route trundling along the affected section of line, calling at every single station. At least these have been seen in planned closures, where extra routes are operated to bring people to nearby stations on other lines. This is a step in the right direction, but since they’re not part of the formal bus network, they don’t pick up anywhere apart from train stations.

Night buses have gotten this right, somehow. The relative variety of night bus services mean that some estates even have a choice of a slow option passing through another town, or a fast one that bypasses more. This can be retained, to create a rail replacement bus strategy much like what is practiced in Sydney — where approximately 100k train trips are to be replaced with a variety of bus routes. Total operations? 180 buses, and this is supposed to be a school holiday period there with lesser travel. I shudder to think what resource usage would be if only all-stop service was run, as we previously have.

But this is okay at night, with the significantly lower clientele and thus needed capacity. With our nightlife scene starting from an even worse position than cities in the West, improved service on most of the existing network would help to kickstart things. Other options such as night train service, either on weekends only or throughout the week, can be explored once things are in a better shape, through the self-reinforcing loop of supply and demand.