Singpore Public Transport Revisited

1. Back in 2008, I wrote an article on how to improve our public transport, especially the buses. (http://elitocracy.blogspot.com/2008/07/land-transport-in-singapore-bus-service.html). With the recent debate on public transport hike, it is time to revisit this issue again.

2. At the core of the contention is the unhappiness of the commuter over the quality of service (QoS) of Singapore’s public transport. The request for a fare hike recently only serves to increase the level of indignation felt amongst commuters.

3. The Government is right to point out that nationalization is not the way to go as it will lead to less efficiency, resulting in higher costs. In terms of correctness of argument, that is as far as it goes. The Government is dead wrong to assert that just because nationalization is not the way, the current model is therefore correct. It is not.

4. As many people have pointed out, for the private operator model to work one needs to introduce competition to drive the operators to improve their efficiency and QoS. That is exactly where the Government has not only failed to achieve but has in fact done the opposite and helped the operators to reduce the competition they face. I will give an example

5. When the North East Line was opened, then Transport Minister Yeo Cheow Tong (YCT) decided to let SBS operate the North East Line (NEL) instead of SMRT. Next, he allowed many of the SBS bus services to be eliminated to avoid “duplication”. (Source http://www.spug.sg/forums/archive/index.php/t-45373.html). With a $5B price tag, NEL was not just expensive to build but potentially hard to break even operationally as well. The Government did not want to subsidize the running of NEL on a continual basis. So what YCT did was to effectively create a public transport monopoly in the North East region by putting NEL and the bus service under SBS and letting SBS did its job of maximizing profits without the worry of competition. And that was just what SBS did; many trunk bus services were removed so that a cheaper transport mode other than MRT would be unavailable to commuters. As a former resident in North East, my transport cost to travel from my house to city along the North East corridor went from around S$1.10 to S$1.80, an effective increase of more than 50%.

6. It should be clear from this example that the Government has been an active participant in manipulating public transport pricing to achieve its own agenda, including creating monopolies so as to pass the bulk of the cost to commuters. We should stop buying the Government’s argument wholesale and instead ask critically how we can improve the efficiency of our public transport system. We can begin by examining the 2 modes of public transport, MRT and buses to find ways in which QoS and efficiency can be raised.

7. MRT trains have large passenger carrying capacity that buses cannot match. A 6 cabin train has probably the carrying capacity of about 9 buses. At one train every 3 min, the MRT is equivalent to the carrying capacity of about 180 buses per hour. It is critical in moving the city around. MRT however is a natural monopoly and it is difficult to inject competition by having 2 or more operators running the trains on a single line. This is problematic as there is less pressure to keep cost down (The Government cannot afford to see the MRT operator go bankrupt and when push comes to shove the operator will get the fare increases it asks for). So even with fare regulation the costs of MRT is likely to go up. Devising a scheme to help to improve QoS and contain cost for MRT is a complicated matter and suggestions on how to go about doing that will be discussed later.

8. Bus on the other hand does not suffer from the natural monopoly syndrome and it is relatively easy to introduce competition for bus services. I have previously proposed a scheme to introduce competition for buses and it is time to re-look at the mechanics of how the scheme works.

9. The core objective of the scheme is to introduce competition into bus services to keep the cost down while ensuring QoS targets are met. We can achieve this by

a. Having 3 or more bus companies operating in Singapore.

b. Inviting the bus companies to bid for bus routes via tender. Tenders will help to maximize the efficiency of the scheme.

c. Specifying within the tender the required QoS for the bus routes and the associated punishment if those QoS are not met.

10. The first step in coming up with the tender is to determine the usage pattern. Turns out the raw data is there for us all along to find the usage patter: Ez-link card. Each time a passenger taps his Ez-link card, a computer log record is generated. Everyday millions of data points are being generated this way.

11. Data analysis of Ez-link log data will enable us to find out the expected traffic volume for a particular bus route at a particular time period and the expected variation. It will also enable us to find out how over crowded the bus is.

12. Armed with this information, it is possible for the Government to specify the QoS that is expected of a bus route. The QoS should specify the expected passenger carrying capacity, the bus frequency for each time period. The tender should also specify the associated punishment if the QoS are not met. Verifying whether the bus company has met the QoS is easy; data-crunch the Ez-link log data to find out. The tender should also detail the usage pattern so as to enable the bus companies to predict the revenue stream from the bus route.

13. The bus companies will take the data from the tender, do their own number crunching and come back with a bid value. If the bus route is lucrative, the bus companies will come back with a positive bid value i.e. the bus company is willing to pay for the right to operate the bus route. If the bus route is money losing, the bus companies will come back with a negative bid value i.e. the bus company will operate the bus route only if the Government is willing to subsidize the bus route.

14. Note that how the Government specifies the QoS will determine the amount of subsidies needed. If the Government set a very high QoS by requiring high bus frequency or large passenger capacity at peak hour, the majority of bus routes will turn out to be money losing. This in turn will raise the amount of subsidies. Given that the current QoS has caused many complaints, it is likely that the Government will have to raise the QoS and in the process subsidize the commuters.

15. Should we have subsidies? We know our Government is ideologically opposed to subsidy. But perhaps this is one of the sacred cows that we ought to slaughter. The duopoly SMRT and SBS do not have incentives to raise QoS beyond the minimum required. In fact, raising QoS (e.g. by running buses at higher frequency) will increase cost and lead to lower earnings. The misalignment of incentives means that to rely on market forces to improve QoS is a non-starter. Instead to raise the QoS beyond the current standard the Government must step in.

16. The subsidy not only benefits the public transport commuters but also the indirectly private transport users as well. One of the stated aims of the Transport Ministry is to raise public transport usage to 70%. To achieve that public transport must become relatively more attractive and raising the QoS is crucial. The increase use of public transport will also have other knock on effects, including less pollution, lesser demands on new road space, lower COE and ERP prices etc.

17. Given that the Government stated aim for its slew of transport related taxes (ERP, COE) are not for revenue generation purposes, the subsidy should come from these taxes. This is a win-win as the subsidy will improve the QoS, leading to better commuter experiences and more public transport usage, which will in turn help to lower the pressure on road usage, leading to lower ERP and COE etc.

18. What about MRT? As mentioned it is much harder to introduce competition into the MRT system. But that does not mean we cannot introduce competition indirectly. The first thing to do is to delink the MRT fare from bus fares as MRT is faster and therefore ought to command a higher fare than buses. It is also difficult to sustain the unified fares if the duopoly model is abandoned and cross subsidies between buses and MRT disappears, which will be if we are to introduce bus route tenders. With the fares delinked, allow buses to ply along the same route as the MRT. This will introduce competition for MRT and prevent its fare from increasing incessantly without losing commuters to bus services.