Many issues with study hailing Singapore as best in class in public transport

On 2nd and 3rd of Jun 2014, Straits Times alerted Singaporeans to yet another international study, this time by Credo Business Consulting LLP and sponsored by Siemens AG Singapore, honoring Singapore for having the best transport network in its class in terms of annual cost of commuting.

It reminds us of another accolade, the Best Metro Asia Pacific award Singapore won for two consecutive years in 2009 and 2010 before the most catastrophic MRT disruptions in the history of our nation the following year in 2011. The sharp contrast between the glamour of 2009, 2010 and the sorry episode of 2011 should serve as a warning to all about the trustworthiness of such awards or studies.

Hong Kong has done better

Credo’s study shows that Hong Kong has done better than Singapore as far as metrics relevant to current public transport performance are concerned:

Credo metrics (higher means better) Hong Kong Singapore
M1 Current capacity & crowding 10 10
M4 User functionality 9 9
M5 Reliability & onboard quality 5 6
M6 Affordability 5.5 3
M10 Density of network 10 4
M11 Average commuting time 5 5
Overall 7.4 6.2

Yet Credo deems Singapore as having done better than Hong Kong presumably because Credo divides annual cost of commuting by per capita GDP. Singapore’s higher per capita GDP compared to Hong Kong’s will result in a lower annual cost of commuting for us despite our inferior metrics. Credo should consider dividing annual cost of commuting by wages instead to eliminate the unique problem in Singapore where per capita GDP is high but wages are low. As far as Singapore and Hong Kong are concerned, public transport is fully paid for by wages and so should be normalized by wages, not per capita GDP. Credo can introduce a different metric to reflect differences in fare box recovery across nations.


Credo’s affordability metrics contradicts LTA’s 2013 average bus and MRT fares:

Credo affordability metric (higher is better) LTA average MRT fare (S$/pax trip) LTA average bus fare (S$/pax trip)
Hong Kong 5.5 $1.60 $1.20
Singapore 3 $0.90 $0.70
New York 5 $1.50 $1.30
Tokyo 5 $1.20 $1.60
London 1.5 $2.60 $0.90

Credo lists Hong Kong’s, New York’s and Tokyo’s fares as being more affordable than ours whereas LTA claims, year in year out, that Singapore has the cheapest fares amongst the five cities. This is despite Credo dividing fares by per capita GDP which tends to favor Singapore due to our high per capita GDP. Again for the same reason of fairness, Credo should consider dividing the affordability metric by wages instead.

LTA’s figures can be obtained by aggregating fare revenues for SBS and SMRT and then dividing by their combined ridership:

Revenue (million $) Ridership (millions) Fare ($)
SMRT bus $223.90 335.4
SBS bus $644.90 973.5
Total bus $868.80 1,308.90 $0.66
SMRT train $607.90 654.4
SBS train $148.10 175.6
Total train $756.00 830 $0.91

The average bus fare of $0.66 is lower than the minimum bus fare and obviously comprises lots of rebated fares, which means transfers from trains, from other buses, from SMRT to SBS and vice versa. Herein lies the problem with the LTA approach, it gives the average price for all part journeys that we make every day but it does not give the combined price of a complete journey from home to destination or vice versa. It doesn’t reflect our true fares.

Congestion and overcrowding

Like Hong Kong, Singapore scored a perfect 10 out of 10 for congestion and overcrowding. Credo calculated congestion and overcrowding by dividing public transport capacity by peak AM demand weighted by modal share (page 57 of preview report). An attempt to recalculate the figures using Credo’s methodology arrives at a completely different picture below:

Hong Kong Singapore
Current congestion and crowding (Credo) 10 10
Peak AM volume of commuters (Credo) 1,555,000 997,000
Train percent share (LTA) [1] 55% 41%
Peak AM volume of train commuters
(Peak AM volume × train % share)
853,975 406,185
Total train capacity (passengers / hour) [2] 599,200 147,000
Calculated train capacity / AM volume 70% 36%
Bus percent share (LTA) [1] 45% 59%
Peak AM volume of bus commuters
(Peak AM volume × bus % share)
701,025 590,815
Total bus capacity (Number of buses × seating capacity) [3] 766,630 347,400
Calculated bus capacity / AM volume 109% 59%
Calculated total capacity / AM volume 88% 50%
What the congestion ratios should be 10 5.6

The table above shows that if Hong Kong scores a perfect 10 for congestion and crowding, then comparatively, Singapore should have barely passed only. That would have greatly impacted our annual cost of commuting and our ranking in the Credo study.

Reliability and on-board quality

Credo used the reliability and on-board quality metric as proxy for waiting time when calculating cost of commuting (page 56 to 58). This metric considers age of fleet, service reliability, air conditioning and Wi-Fi connectivity.

• Credo didn’t take into account Singapore’s hub and spoke model of public transportation that forces most journeys into multiple part journeys that compounds and exacerbates our waiting time
• It is unfair to include air conditioning as a criterion because air conditioning is largely not required for the cooler climates of most Western nations which are tolerable even in summer given their lower humidity.
• The inclusion of Wi-Fi connectivity as a criterion is problematic because while our MRT stations may claim to have Wi-Fi connectivity, the connection breaks once we enter the tunnel. People can read books, play pre-installed games, watch pre-downloaded movies, pre-downloaded news on buses and trains and don’t have to rely on Wi-Fi.


Credo should divide annual cost of commuting and the affordability metric by wages instead of per capita GDP for better fairness. It can introduce a new metric to account for differences in fare box recovery across nations.

Credo’s study debunks LTA’s yearly statistics showing Singapore having the lowest fares amongst five cities.

Credo’s congestion and overcrowding metric for Singapore seems way off which may significantly impact our annual cost of commuting and ranking in the Credo study.

Credo’s classification of Singapore, Hong Kong and Tokyo as high density compact centres is problematic as other cities in the “well established” category may be no less dense or compact than them. Credo defines high density compact centres as:
• Modern cities that have experienced recent or ongoing expansion, with high population density centres. Transport networks may be less developed than in well-established cities.
It is not clear why Singapore, Hong Kong and Tokyo are considered less developed or not well-established.

Credo should have accounted for the fact that Singapore being smaller than many other cities will, all else being equal, have lower fares.

[1] Calculation of train percent share and bus percent share for Hong Kong and Singapore

Singapore trips / person / day Hong Kong trips / person / day
Train (from LTA’s Statistics in Brief 2013) 0.44 0.67
Bus (from LTA’s Statistics in Brief 2013) 0.64 0.55
Total 1.08 1.22
Train % 41% 55%
Bus % 59% 45%

[2] Calculation of total train capacity for Hong Kong and Singapore

Capacity per train arrival Interval between trains (seconds) Peak hour capacity (number passengers per hour)
North South line 1,400 120 42,000
East West line 1,400 120 42,000
North East line 1,400 120 42,000
Circle line 700 120 21,000
Total capacity 147,000

Data for Singapore trains
• A six-carriage train can carry about 1,400 commuters during peak hours, said SMRT rail strategic planning director Lui Weng Chee.
[Straits Times, Passenger capacity to increase up to 50%]
• This leads to gaps of at least 120 seconds between a train leaving and the next one entering the station.
[Straits Times, NEW SIGNALLING SYSTEM Work to ease congestion on busiest train lines]

Hong Kong train lines Peak hour capacity (number passengers per hour)
East Rail Line 101,000
Island Line 85,000
Kwun Tong Line 85,000
Ma On Shan Line 32,000
Tseung Kwan O Line 85,000
Tsuen Wan Line 85,000
Tung Chung Line 45,000
West Rail Line 64,000
Airport Express 6,400
Disneyland Resort Line 10,800
Total capacity 599,200

Data for Hong Kong trains
• Legislative Council Panel on Transport Subcommittee on Matters Relating to Railways, Capacity and Loading of MTR Trains, CB(1)980/13-14(03), Annex

[3] Calculation of total bus capacity for Hong Kong and Singapore

Singapore buses Number of buses Bus seating capacity Number of buses × seating capacity
SBS double decker buses 1,330 130 172,900
SBS single deck buses 1996 80 159,680
SMRT Bendy buses 114 130 14,820
SMRT single deck buses 1026 80 82,080
Total 4,466 347,400
Hong Kong buses Number of buses Bus seating capacity Number of buses × seating capacity
Double deckers 5,123 130 665,990
Single deckers 388 80 31,040
Minibuses 4,350 16 69,600
Total 766,630

Data for Singapore buses
• All 3,326 buses in our fleet are air-conditioned, with four in 10 being double decks.
[SBS annual report, page 4]
• Assumed 10% of SMRT buses are Bendy buses (welcome any information on this for update)

Data for Hong Kong buses
• Currently Hong Kong’s franchised bus network is made up of about 600 routes, operated under six franchises by five franchised bus companies with a total fleet of 5,511 licensed buses (at end of 2013), comprising 5,123 double-deck and 388 single-deck vehicles.
• With a fleet of 3777 buses, mostly double-deckers, KMB is one of the largest road passenger transport operators in the southeast Asia.
[Information Services Department publication “Hong Kong: The Facts” dated Sept 2013]
• Public Light Buses (PLBs) are minibuses with not more than 16 seats. Their number is fixed at a maximum of 4350 vehicles.
[Information Services Department publication “Hong Kong: The Facts” dated Sept 2013]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: